Load Bearing Glass

Budapest University of Technology and Economics
Faculty of Civil Engineering

Department of Construction Materials and Engineering Geology
LOAD BEARING GLASSES
Doctoral (PhD) Thesis
Kinga PANKHARDT, MSc (CE)
Supervisor:
Prof. György L. BALÁZS, PhD
2010.
Budapest University of Technology and Economics
Faculty of Civil Engineering
LOAD BEARING GLASSES
Doctoral (PhD) Thesis
by
Kinga PANKHARDT, MSc (CE)
Supervisor:
Prof. György L. BALÁZS, PhD
2010.
"I want to erase architecture, that's what I've always wanted to do
and it's unlikely I'll never change my mind." *
writes Kengo Kuma (1954) architect in Tokyo, Japan
Erasing architecture means making it transparent itself as well as

preventing the appearance of any object.
Glass is the building material which can make this dream come true.
But what do engineers think about it?
Kinga Pankhardt, 2009
This dissertation is dedicated to Dr. Salem Georges Nehme.
*
http://www.archilab.org/public/2000/catalog/kengo/kengoen.htm
© K. Pankhardt, 2010 for this Doctoral Thesis “Load bearing glasses”

K. Pankhardt Load bearing glasses Acknowledgements
Acknowledgements
I would like to express my thanks to Prof. György L. BALÁZS, my supervisor for his
continual support and guidance throughout this work.
I would also like to thank the members of the review committee for their time, advice and
expertise.
I would like to express my thanks to RÁKOSY GLASS Ltd., especially to Eszter
RÁKOSY for providing the specimens, as well as SK Engineering for the financial support.
I would also like to thank all of my colleagues at the BME Department of Construction
Materials and Engineering Geology, especially Dr. Zsuzsanna JÓZSA for the support at the
beginning of my PhD studies and Dr. Attila ERDÉLYI, Dr. Piroska ARANY for the
intellectual support.
I would also like to thank for the support with SEM measurement Dr. Katalin
KOPECSKÓ and Dr. Béla KOCZKA at the Department of Inorganic and Analytical
Chemistry.
I would also like to thank for the support with DSC analysis of plastics Dr. Viktória
VARGHA and Ramóna Carmen BENDE, Szilvia MÁTÉ, Department of Physical Chemistry
and Materials Science.
I would also like to thank Dr. Dezső L. KOVÁCS, Dr. Sándor FEHÉRVÁRI, Márk
HATALA, András EIPL, András BÁCSI and Mátyás VARGA and Attila CSICSELY, Péter
TISZA, Dávid DIRICZI, Gábor KOVÁCS for the technical support.
I would also like to thank all of my colleagues at the University of Debrecen, Department
of Civil Engineering, especially to Dr. Imre KOVÁCS Head of the Department of Civil
Engineering and József KOVÁCS who helped me, so I could spend more time on my PhD
research.
I would also like to thank Andrew SALAMIN as reviewer of the English text.
I would also like to thank my former colleagues at GLASMETAL Ltd., Pál GORDON,
Zoltán BUN, Erzsébet SZUHAY with whom I worked and realized lots of special glazing
projects in Hungary.
Last but not least, I would like to thank my fiancé Dr. Salem Georges NEHME for
providing plenty of advice and diversions, unwavering support and love and sacrifice, and to
my grandparents, parents and brothers and friends for their support and love.
This work is connected to the scientific program of the “Development of quality-oriented

and harmonized R+D+I strategy and functional model at BME” project. This project is
supported by the New Hungary Development Plan (Project ID: TÁMOP-4.2.1/B-09/1/KMR-
2010-0002).
Kinga PANKHARDT
iii
K. Pankhardt Load bearing glasses Contents
Contents
Acknowledgements iii
Contents iv
Preface viii
Notations and abbreviations ix
Chapter 1 1
1. Research significance and aims of Thesis 1
Chapter 2 2
2. Properties of glass 2
2.1 Introduction, glass in nature 2
2.1.1 Brief history of glass 2
2.1.1.1 Early manufacture of flat glass 3
2.1.1.2 Modern flat glass manufacture 3
2.2 General properties of glass 4
2.2.1 Glass types by ingredient of glass 4
2.2.2 Glass products by manufacturers 4
2.3 Conclusions 4
Chapter 3 5
3. Float glass as building material 5
3.1 Introduction 5
3.2 Physical properties of float glasses 5
3.2.1 Viscosity of glass 5
3.2.2 Basic physical properties of glass 6
3.3 Strengthening of flat glass 6
3.3.1 Wired glass 7
3.3.2 Heat-treated glass products 7
3.3.3 Chemically strengthened glass 8
3.3.4 Edge strength of heat-strengthened and tempered glasses 8
3.4 Recycled glass 9
3.5 Conclusions 9
Chapter 4 10
4. Plastic interlayer materials of laminated glass 10
4.1 Introduction 10
4.2 Influencing parameters of laminated glass 10
4.2.1 Type of binder materials 10
4.2.2 Resins as interlayer materials 10
4.2.3 Foils as interlayer materials 11
4.3 Laminated glass as composite material 12
4.3.1 Physical properties of interlayer materials 12
4.3.2 Thermo-mechanical properties of polymers 12
4.3.3 Shear strength of the interlayer 13
4.3.4 Time and temperature superposition 14
4.4 Conclusions 15
iv
Chapter 5 16
5. Strength and durability of glass 16
5.1 Introduction 16
5.2 Corrosion and weathering of glass 16
5.2.1 Effect of surface condition 16
5.2.2 Leaching of glass 16
5.3 Chemical resistance of glass 17
5.3.1 Influence of sharp impactor 17
5.3.2 Loss of strength by aging 18
5.3.3 Strength recovery by aging 18
5.4 Crack development during fracture of glass 19
5.4.1 Hardness 19
5.4.2 Crack growth in glass 19
5.4.3 Crack direction in glass 20
5.5 Conclusions 20
Chapter 6 21
6. Bending strength of float and laminated glasses 21
6.1 Introduction 21
6.2 Influencing factors on bending strength of single or laminated glasses 21
6.2.1 Single float glass pane 21
6.2.2 Laminated float glass pane 21
6.2.2.1 Laminate without bonded layers 22
6.2.2.2 Laminate with bonded layers 23
6.2.2.3 Edge strength of laminate 25
6.3 Standardisation 25
6.3.1 Standards, technical guidelines for glazing that support human loads 25
6.3.2 Calculation of bending strength in four-point bending according to
EN 1288-3:2000 26
6.3.3 ASTM E 1300-04:2004 27
6.3.4 Technical guidelines given by glass manufacturers 27
6.4 Conclusions 27
Chapter 7 28
7. Experimental programme 28
7.1 Introduction 28
7.2 Test parameters and test programme 28
7.2.1 Test parameters 28
7.2.2 Test programme 29
7.3 Method of testing (four-point bending) 30
7.3.1 Test specimens 30
7.3.2 Experimental set up 30
7.3.2.1 Force measurement 30
7.3.2.2 Rate of loading 31
7.3.2.3 Strain measurement 31
7.3.3 Scanning electron microscopic analysis (SEM) 32
7.3.4 Differential scanning calorimetric analysis (DSC) of cured resin 32
7.3.5 Physical properties of interlayer used during test programme 32
7.4 Conclusions 33
v
Chapter 8 34
8. Effect of tempering on load bearing capacity of single glass pane in bending 34
8.1 Introduction 34
8.2 Experimental results and test observations 34
8.2.1 Force vs. mid-point deflection relationship 34
8.2.2 Stress-strain relationship 35
8.2.2.1 Strain in mid-region (Region 1) of pane 35
8.2.2.2 Strain in edge region (Region 2) of pane 36
8.2.2.3 Stresses in mid-region (Region 1) of pane 37
8.2.2.4 Stresses in edge region (Region 2) of pane 39
8.3 Conclusions 42
Chapter 9 43
9. Effect of lamination on bending characteristics 43
9.1 Introduction 43
9.2.1 Stages of fracture process of laminated glass 43
9.2.2 Force vs. mid-point deflection relationship 44
9.2.2.1 Effect of number of glass layers on load bearing capacity
of laminated glass pane 44
9.2.2.2 Effect of interlayer material on load bearing capacity
9.2.3 Stress-strain relationship 49
9.2.3.3 Stresses in mid-region (Region 1) of pane 51
9.2.3.4 Stresses in edge region (Region 2) of pane 52
9.3 Conclusions 52
Chapter 10 53
10. Effect of temperature on bending characteristics 53
10.1 Introduction 53
10.2.1 Effect of temperature on load bearing capacity of single glass layer 54
10.2.1.1 Load – deflection relationship 54
10.2.1.2 Ultimate strain in mid-region (Region 1) and
edge Region (Region 2) of pane 54
10.2.2 Effect of temperature on load bearing capacity of laminated glass 55
10.2.2.1 Load – deflection relationship 55
10.3 Conclusions 63
Chapter 11 64
11. Flexural stiffness of laminated glasses 64
11.2.1 Flexural stiffness 64
11.2.2 Calculations of temperature dependent flexural stiffness
of laminated glasses 67
11.3 Exposure classes of laminated glasses 70
11.4 Conclusions 71
vi
Chapter 12 72
12. Residual load bearing capacity of laminated glass panes 72
12.2 Fracture process of laminated glass in four-point bending 72
12.2.1 Force during fracture of laminated glass 72
12.2.1.1 Effect of number of glass layers 72
12.2.1.2 Effect of tempering on residual load bearing capacity
12.2.1.3 Effect of temperature sensitivity of interlayer material
on residual load bearing capacity 73
12.2.2 External work of laminated glasses 73
12.3 Conclusions 74
Chapter 13 75
13. Future work 75
Summary of Scientific New results S1–S3
References xi
Appendix A – Literature review A 1 – A 26
Appendix B – Test parameters and test results B 1 – B 36
Appendix C – Properties of interlayer materials C 1 – C 13
Appendix D – Definitions D1–D5
vii
K. Pankhardt Load bearing glasses Preface
Preface
Would buildings be acceptable without windows?
Glass appears everywhere in modern life: lighting, household wares, cars, fibres, etc. Glass is
used nowadays in architecture, not only in windows, but also in load bearing structures.
Architects prefer to design buildings with high transparency or sometimes prefer full
transparency, like Water and Glass house by Kengo Kuma, Atami, Japan, 1995. With the
increasing transparency of buildings the engineers can meet with special design and
construction problems for glazing and load bearing glass structures.
Water and Glass house, architects: Kengo Kuma and Assoc., Atami, Shizuoka (1992-1995)
Glass used in building exteriors must resist environmental loads and impacts. Therefore, the
appropriate type of glass must be carefully selected. Laminated glass is considered by many
people the best solution to cover commercial buildings while guaranteeing high impact
resistance. Laminated glass can also significantly reduce noise from the street as well as
control solar energy transmittance. Laminated safety glass (LSG) is a composite of two or
more glass panes connected by an interlayer of resin or different types of plastic foils e.g.
polyvinyl butyral (PVB) or ethyl-vinyl-acetate (EVA).
The bond between the glass panes depends mainly on the quality of the bond layer (interlayer
material), load duration, temperature and can be affected by aging of the interlayer material. If
one or more glass layers fracture, the residual load bearing capacity and structural safety
depends on the nature of the fracture and on the adhesion between the interlayer and the glass
as well as on the properties of the interlayer material (tear strength, elongation, etc.).
Design and calculation of load bearing glass structures are special areas of civil engineering.
Reasonable solutions and models are necessary to calculate glass structures. Engineers are
still afraid of the fragile nature of glass, therefore, glass constructions are mostly secondary or
tertiary structural elements (in terms of the structural hierarchy of a building) (Pankhardt,
2004, Pankhardt 2010e). There is a lack of useful glass technical documents and standards,
especially for glass engineers in Hungary.
Therefore, this Thesis “Load bearing glasses” is focused on the testing of single and
laminated glasses and also demonstrates the influencing factors by creating load bearing glass
structures e.g. glass beams, glass columns, glass slabs and roofs, where the principal loads can
be carried by glass elements.
Bold letters are used for the text of the new scientific results.
Kinga PANKHARDT
viii
K. Pankhardt Load bearing glasses Notations and abbreviations
Notations and abbreviations

Upper case letters
A Area [mm2]
A, B Parameters by the chemical composition of the glass (Lakatos et al.,
1972)
C(t) Compliance (1/E) of plastic materials [1/(N/mm2)]
Dfl Flexural stiffness (EI) [Nmm2]
Dfl -1 Flexural flexibility (1/EI) [1/(Nmm2)]
E Modulus of elasticity (Young's modulus) [N/mm2]
Ea Activation energy [kJ/mol]
Efl Flexural Young's modulus [N/mm2]
Fmax Maximum force [kN, N]
Fu Ultimate force [kN, N]
G Shear modulus (complex) [N/mm 2]
G’ Elastic component of G-modulus, dynamic storage modulus [N/mm 2]
G” Viscous component of G-modulus, dynamic loss modulus [N/mm 2]
HK0,1/20 Hardness Knoop (P=0.1kg, t=20 sec) [GPa]
HM Hardness (Mohs’) -
I Moment of inertia [mm4]
Js Effective section modulus (Wölfel, 1987) [mm3]
KIc Critical value of stress intensity factor, toughness (mode I) [MPa·m1/2]
L Length [mm]
Lb Distance between the centre lines of the bending rollers [mm]
Ls Distance between the centre lines of the supporting rollers (span) [mm]
M Moment (bending) [Nmm; kNcm]
Pc Threshold of indentation load [N]
R 8.34, Ideal gas constant [J/mol⋅°K]
Td Delamination temperature [°C]
Tg Glass transition temperature [°C]
Tm Melting temperature [°C]
Tref Reference temperature [°K ,°C]
Tservice Service temperature [°C]
Weff Effective section modulus (Norville, 1997) [mm 3, cm 3]
Lower case letters
a Crack length [mm]
aT Temperature shift factor -
b Specimen width [mm]
c Specific heat capacity [J/(kg⋅°K)]
ft,fl,k Bending strength (characteristic) [N/mm2]
fcu Compression strength (ultimate) [N/mm2]
g Acceleration due to gravity [m/s2]
h Thickness [mm]
heff Effective thickness [mm]
hint Thickness of interlayer material (in laminated glass in one [mm]
lamination plane)
hint,tot Total thickness of interlayer in laminated glass [mm]
hnom Nominal thickness [mm]
htot Total thickness of laminated glass [mm]
k Dimensionless factor (EN 1288-1 Ch. 6.2) -
k1 Dimensionless multiplying factor for laminated glass -
ke Dimensionless multiplying factor for single layer glasses -
kservice Dimensionless multiplying factor for laminated glass at service -
(application) temperature
n Number of glass layers [pcs]
ix
K. Pankhardt Load bearing glasses Notations and abbreviations
nR Mean refractive index -

q Shear transfer coefficient (Norville, 1997) -
t Time [s]
u Shift [mm]
y Central deflection of the specimen relative to the supporting rollers [mm]
v Velocity [m/s]
Greek letters
α Coefficient of thermal expansion [1/°C]
β Decay factor -
ε Strain [%; µm/m]
ϕ Creep parameter -
η Coefficient of viscosity [Pa⋅s]
ηT Temperature dependent coefficient of viscosity [Pa⋅s]
κ Coupling parameter -
γ Shear angle [-, °]
Γ Transfer coefficient (Wölfel, 1987) -
λ Thermal conductivity [W/(m⋅°K)]
µ Poisson coefficient -
ρ Density of the specimen [g/cm3] [kg/m3]
σ Surface strength [N/mm2]
σa Applied stress [N/mm2]
σb Bending stress in the surface area defined by the bending rollers [N/mm2]
σbB Bending strength [N/mm2]
σbB,edge Bending strength in edge region (Region 2) [N/mm2]
σb,eff Effective bending stress [N/mm2]
σbG Bending stress imposed by the self-weight of the specimen [N/mm2]
tan δ = G”/G’ -
ξ Reduced time [s]
Abbreviations
AFM Atomic Force Microscope
DSC Differential Scanning Calorimetric analysis
EVA Ethyl Vinyl Acetate
GTF Glass Type Factors (tables of ASTM E 1300-04:2004)
HDT Heat Distortion Temperature
HSG Heat Strengthened Glass
HST Heat-soak tested
IG Insulating Glass
IRHD International Rubber Hardness Degrees (~ Shore A) (ISO 48 : 1994),
LEFM Linear Elastic Fracture Mechanics
LG Laminated Glass
LR Load Resistance
LS Load Share factor
LSG Laminated Safety Glass
NFL Non-Factored Load (charts of ASTM E 1300-04:2004)
NiS Nickel Sulphide
PDF Probability Density Function
PVB Poly Vinyl Butyral
SEM Scanning Electron Microscopy
SF Glass Strength Factor
SG Strengthened Glass
TTS Time-Temperature Superposition principle
UP Unsaturated polyester
VA Vinyl Acetate
WLF Eq. Willians-Landel-Ferry Equation
x
K. Pankhardt Load bearing glasses Chapter 1
Research significance and aims of Thesis
Chapter 1
1. Research significance and aims of Thesis
Nowadays glass strongly influences modern architectural design. Glass is a rigid material and
for a long time its brittleness was a well-known property besides its transparency. With the
development of glass strengthening methods, glass becomes an often used building material
also in load bearing structures. Requirements are increasing on the safety of glazing in public
buildings e.g. used in balconies, balustrades, staircases, overhead glazing (App. A, Figs. A.1.1
a-l, Fig. A.1.2) (Pankhardt, 2004).
DT
Heat-treatment procedure is the most often used glass strengthening method for building
glasses. Two different types of so-called pre-stressed glasses are produced with the heat
strengthening method: tempered and heat strengthened glasses. Tempered glass can be
considered as safety glass owing to the fracture pattern with tiny blunt edged fragments.
AR
With lamination of more glass panes, multi-layered (hereinafter laminated) glasses are
produced. EN ISO 12543-1:2000 differentiates between laminated and laminated safety
glasses. Laminated safety glass is generally used with an interlayer of foil. For larger sizes
KH
and curved shapes, laminated glass is usually available with a resin interlayer. The interlayer
material serves two purposes: (i) to keep glass splinters in place during the fracture process to
reduce the risk of injury and (ii) to increase residual load bearing capacity. Until recently,
some national standards explicitly stipulated the composition of the interlayer, usually within
AN
the standard definition of laminated glass. The new European standards (EN 356:1999) also
focus on performance rather than composition. As a result, interlayer materials are not
explicitly specified for fields of application e.g. use in commercial or public buildings in
outdoor or indoor conditions. Therefore, those foil and resin interlayer materials or others are
K.P
acceptable, which are determined to be acceptable for use in laminated glass by the
manufacturer of interlayer materials. Further investigations are required especially in those
areas, where glass is used as a load bearing element.
Since glass is used in buildings, the most important property for load resistance is the bending
t by
characteristics. The aims of this Thesis were to study the bending characteristics of single
soda lime silicate glasses and those of laminated glasses, including:
§ force and deflections (Chapters 8 to 10),
§ strains in two different regions of glass surface (Chapters 8 to 10),
§ flexural stiffness (Chapter 11),
h
§ temperature dependent residual load bearing capacity (Chapter 12).

yrig
The following main influencing factors on bending characteristics were studied:

§ effect of thickness of glass pane (Chapter 8),
§ effect of tempering (Chapter 8),
§ effect of edgework (Chapter 8),
cop
§ effects of interlayer materials (Chapter 9),

§ effect of temperature of glass pane (Chapter 10),
Interlayer materials used in laminated glasses were resin (UP) or EVA (ethyl-vinyl-acetate)
foil. While PVB (polyvinyl-butyral) foil is widely used in laminated glasses, EVA foil is a
new generation of foils. EVA interlayer was first used in Hungary in 2005 by Rákosy Glass
Ltd. Glasses are used not only in the interior but also in the exterior. Therefore, the effect of
temperatures of -20 °C, +23 °C and +60 °C were investigated on the bending characteristics
of glasses. Bending strength with the well-known formulas was also evaluated in the case of
non heat-treated float and tempered glasses and was compared with the measured surface
stresses. Are the strains measured in the centre of the pane and near the edge equal? Strain
measurements of the surface of loaded specimens – at the middle of the pane and near the
edge – were also investigated.
1
Properties of glass
Chapter 2
2. Properties of glass
2.1 Introduction, glass in nature
Glass is one of the most important building materials of architecture, which is nowadays also
a construction material for engineers. Glass exists in nature like volcanic glasses, for example
basaltic glass of volcanic activity. The ratio of silica (in latin silex) among the eight main
minerals (O, Si, Al, Fe, Ca, K, Na, Mg) is 31 %, and its ratio among the minerals containing
silica and oxygen is 78 %, in the Earth’s crust. The minerals can be in crystalline or non-
DT
crystalline (amorphous) forms (Török, 2007). Vulcanite amorphous glasses can be agates,
(App. A, Fig. A.2.1.a). Deep-seated magmatic crystalline silica is for example quartz, (Fig.
A.2.1.b).
AR
Another example of amorphous glasses is the stone-age arrow head made of obsidian as a
natural glass. The arrow heads are found to be still razor sharp after tens of thousands of
years (silica and alumino-silicates) (Gibbs, 1996). Composition of amorphous natural glass is
similar to manufactured glass. Both are produced at similar temperatures (temperature of
KH
magma is about 650 to 1250 °C, under high pressure it can reach even 1400 °C). Natural glass
is usually quenched more abruptly then manufactured glass. Magma contains liquid, solid and
gas phases including silica 37 to 75 %. Its viscosity and liquid-like behaviour depends on its
composition. The more alkaline the magma, the more viscous it is. This characteristic is
AN
similar to manufactured glass containing soda. Like artificial glass, it forms fibres, drops and
sometimes contains iron impurities. The fractured surface of mineral opal (Fig. A.2.2.a) looks
like non heat-treated glass (shell-like fracture, Fig. A.2.2.b and Fig. A.2.3).
K.P
Glass fibres are used also in nature as a building material by sea animals (Fig. A.2.4).
Some sea animals (sea-sponge Euplectella aspergillum, Rosella racovitzea) ingest dissolved
silica from seawater to form their skeletons (Sarikaya et al., 2001; Aizenberg, 2005; Woesz et
al., 2006). This natural glass fibre is very flexible and resistant. Fig. A.2.4.a indicates the
natural glass fibre system of a sea-sponge. Fig. A.2.4.b is enlarged from Fig. A.2.4.a and Fig.
t by
A.2.4.c is enlarged from Fig. A.2.b as a section of one fibre of about 0.05 mm. This small
diameter fibre can reach much higher strength than architectural glasses (Figs. A.2.5 a) and b)
and A.2.6). Strength of glass can be more influenced by the number and distribution, etc. of
cracks; its durability is influenced by the environmental conditions (e.g. humidity), see
h
Chapter 5.
yrig
2.1.1 Brief history of glass

Phoenician merchants actually discovered glass (or rather became aware of its existence
accidentally) by transporting stone in the region of Syria around 5000 BC (Fig. A.2.7). It was
the Romans who began to use glass for architectural purposes with the discovery of clear
cop
glass (through the introduction of manganese oxide) in Alexandria around 100 AD. Cast glass
windows with poor optical qualities then began to appear in the most important buildings in
Rome and the most luxurious villas of Herculaneum and Pompeii (Schittich, 1999).
An interesting demonstration of annealed glass was introduced to England in the 17th
century by Prince Rupert of Bavaria (1619-1682), grandson of James I of England. Prince
Rupert brought an object called “Prince Rupert's Drop” (Fig. A.2.8) to the attention of the
King, which was used as a joke. This solid glass object had a bulbous end with a thin curved
tail. It is formed by dropping a small gob of hot, molten glass into cold water, and leaving it to
cool. This creates stress between the cooled outside layer and the inside, which is warm.
When the thin fragile tail is broken, the glass releases the internal stress and the entire object
shatters into fine particles and powder. It was harmless fun though, as the glass shatters into
tiny pieces, not into shards. However, de la Bastie (1874) granted a patent (British Patent
2
Properties of glass
2783) on the process of tempering, yet the effect of tempering on glass was not completely
understood. First Littleton and Preston (1929) described the temperature gradients of glass on
cooling. Tempering was industrially realized about 1950 based on the former researches.
A further interesting aspect is, for as long as glass has existed, experiments with glasses have
produced sound. The tuned metal cups or bowls of Asia were transformed in Europe into
tuned glasses “Musical glasses” (see App. D) and are first seen in the Musica theoretica
(1492) (Encyclopaedia Britannica Online, 2009).
2.1.1.1 Early manufacture of flat glass
Flat glass was manufactured in the 11th century by blowing a hollow glass sphere and
DT
swinging it vertically. Gravity pulls the glass into a cylinder measuring about 3 metres long
with a width of up to 45 cm. While it was still hot, the ends of the cylinder were cut off and
the resulting cylinder cut lengthways and laid flat.
Another type of flat glass was crown glass (also known as "bullions"), relatively common
AR
across Western Europe. With this technique a glass ball was blown, then opened outwards at
the opposite side to the blowing pipe. Further spinning of the semi-molten ball caused it to
flatten and increase in size, but only up to a limited diameter. The panes created in this way
KH
were joined with lead strips to create windows (Fig. 2.1). Stained glass windows reached their
peak at the end of the Middle Ages with an increasing number of public buildings, that had
clear or coloured glass (see App. D) decorated with historical scenes (Schittich et al., 1999).
AN
K.P
t by
Fig. 2.1 Flat glass making techniques a) glass blowing, b) spinning of the semi-molten glass
(Schittich et al., 1999), c) „bullion window” (National Geographic Channel, 2002)
2.1.1.2 Modern flat glass manufacture
h
A key figure of modern glass research was the German scientist Otto Schott (1851-1935), who
used scientific methods to study the effects of numerous chemical elements on the optical and
yrig
thermal properties of glass (Schittich et al., 1999). In the 19th century, the main available
glass products were pressed glass, cast glass and rolled glass (Fig. A.2.9). American engineer
Michael Owens (1859-1923) invented (due to increasing automation at the end of the 19th
century) an automatic bottle blowing machine which only arrived on the continental part of
cop
Europe after the turn of the century.

In 1905 Fourcault invented the vertical drawing of glass from a tank into continuous
sheets with constant width. Around the end of the First World War, Emil Bicheroux
developed a process whereby the molten glass was poured from a pot directly through two
rollers. Like the Fourcault method, this results in an even thickness of glass and made
grinding and polishing easier and more economical.
The float process (see App. D) was developed and introduced in 1959 by Pilkington Brothers
with high optical qualities of flat glass. Molten glass poured across the surface of a bath of
molten tin (Fig. A.2.10) spreads and flattens before being drawn horizontally in a continuous
ribbon into the annealing lehr (oven). Nowadays, flat glasses used in architecture are
produced in the float process. The four main float glass manufacturers in Europe are: AGC,
Guardian, Asai Glass and Pilkington.
3
Properties of glass
2.2 General properties of glass
2.2.1 Glass types by ingredient of glass
Glasses are made by melting a mixture of batch (see App. D) ingredients to a homogeneous
liquid. Their compositions are presented in Table A.2.1: 1 typical modern soda lime silicate
glass (used to make bottles and windows); 2 container glass, (ancient Roman soda lime
silicate glass); 3 glass fibres; 4 low expansion borosilicate glass Pyrex (laboratory and some
baking ware); 5 high lead crystal (optical); 6 96% silicate glass (can withstand very high
temperatures) (Balázs 1984, Hrma 2006).
In soda lime silicate glasses (see App. D) (of about 70% SiO2 content), the viscosity
DT
decreasing effect of Na2O is about twice that of K2O. Lime, CaO, is added into the mixture to
improve the chemical resistance; alumina, Al2O3, can also be used for this purpose. CaO and
MgO decrease only the melting temperature and CaO increases the annealing temperature.
AR
K2O, MgO, BaO, ZnO and PbO have nearly the same viscosity decreasing effect at melting
temperature. Alkali oxides and PbO decrease the viscosity even at higher viscosity levels and,
therefore, an increase of these oxides leads to “long” glasses. The viscosity decreasing effects
of CaO and B2O3 are higher at low viscosity levels and decrease quickly at higher viscosities.
KH
These components make “short” glasses (Lakatos, 1976). When a liquid is cooled down to
below its melting point, it solidifies. Sometimes it can become supercooled and remain liquid
below its melting point because there are no nucleation sites available to initiate the
crystallisation. The molecules then have a disordered arrangement and their cohesion
AN
maintains some rigidity. In this state it is often called an amorphous solid or glass. At the
level of molecular physics, there are three main types of molecular arrangement (Gibbs,
1996): crystalline solids: molecules are ordered in a regular lattice, fluids: molecules are
K.P
disordered and are not rigidly bound, glasses: molecules are disordered but are rigidly bound.
During batch mixing, it is important to eliminate certain compounds, like nickel sulphide,
which can be dangerous in tempered glass. By melting of a mixture, nickel and sulphide can
form (NiS) inclusions (Fig. A.2.11) and will cause spontaneous breakage. To prevent the
formation of NiS, the manufacturer must take care that the raw materials are inspected for
t by
contaminants. Also, the contact of molten glass with nickel-bearing materials of instruments
or machines, such as stainless steel should be limited (Loughran, 2003). To eliminate the
spontaneous breakage of tempered glass it should be heat-soak tested (see App. D).
2.2.2 Glass products by manufacturers
h
Glass products are grouped by manufacturers with standardised requirements (see, Table
yrig
A.2.2). Base products are: products obtained upon leaving the furnace and undergoing no
further treatment. Processed products are: products obtained by processing of base products.
Further distinctions: (i) primary processing are e.g. large sizes (“Jumbo”, Fig. A.2.12) or,
where necessary, standard sizes (Fig. A.2.13) and (ii) secondary processing are e.g. tempering
cop
of glass panes (see, Chapter 3).

2.3 Conclusions
Glass is an ancient material which exists also in nature, in the form of minerals and in the
skeletons of sea animals. The history of glass manufacture dates back to ancient times. Due to
the development of different methods of glass manufacture, a wide range of glass types is
produced. Although at the beginning flat glass was available with poor optical qualities and in
limited size, glass was commonly used in windows. Significant change happened due to the
increase of industrial manufacture of glass. Architects began to design buildings with more
glass surface due to the increase of transparency and larger available glass pane sizes. With
the discovery of glass strengthening methods, in the last few years glass began to also be a
load bearing material for engineers, which raises many questions too.
4
Float glass as building material
Chapter 3
3. Float glass as building material
3.1 Introduction
Float glass is widely used in architecture because of the good optical quality of the surface.
The float process produces glass sheets with a uniform thickness and perfectly smooth
surfaces that need no further polishing. The resulting glass will then be further treated in
various ways. Soda lime silicate glass is used mainly for architectural purposes. The advanced
treatment technologies are applied to float glass products depending on the end-products and
DT
on the application. When float glass is used as a load bearing element, safety glass is needed.
3.2 Physical properties of float glasses
AR
3.2.1 Viscosity of glass
The viscosity is the most important property technologically as well as for understanding the
nature of glass transition, stress relaxation, etc. (Graves, 2005). At lower glass temperatures
both viscous flow and stress relaxation are reduced due to higher viscosity. The presence of
KH
temporary tension for longer time increases the probability of fracture, mainly in the edge
region. The edge quality must be improved to minimise fracture. Hence, the balance between
mechanical strength and glass deformation must be optimised by controlling heating rate,
AN
glass temperature, glass thickness, roller spacing and cooling rate (Gulati et al., 2001).
Viscosity,η, is the property of the fluid to resist flow. The law of elastic shear for solids and
Newton’s law of viscosity for classical liquids are linear. The respective proportionality
constants, the elastic shear modulus, G, and viscosity, η, are well determined for most
K.P
common materials. The viscosity of the melt (see App. D) represents the most significant
parameter in glass processing. The dependence on the temperature is determined by the
Vogel-Fulcher-Tammann equation (Zimmermann et al., 2001):
B
log(ηT ) = A + (3.1)
T − T0
t by
where parameters A, B and T0 are determined by the chemical composition of the glass
(Lakatos et al., 1972). Workability is defined by the time used for processing (App. A, Fig.
A.3.1 and Table A.3.1). It can be stated that the shorter the glass melt becomes, the steeper the
h
viscosity-temperature curve runs (Fig. A.3.2). This definition can be restricted because of the
time independent viscosity of glass.
yrig
Taking into consideration the time dependence of the process, it follows that the faster the
glass cools down or heats up, the shorter it is (Zimmermann et al., 2001). Within the concept
of dynamic workability, the temperature, T, has to be considered as a function of time, t. The
viscosity depends indirectly on time:
cop
B
log(ηT (t ) ) = A + (3.2)
T (t ) − T0
So far, the viscosity has been considered as a function of temperature and indirectly as a
function of time, Eq. (3.2).
Because of the heating and cooling process, temperature gradients occur. Therefore, the
spatial dependence of temperature and viscosity cannot be neglected:
B
log(ηT ( t ,r ) ) = A + (3.3)
T (t , r ) − T0
The viscosity can be completely described within the concept of dynamic workability,
Eq.(3.3).
How does viscosity influence the development and reduction of stresses?
5
On the basis of the Maxwell-model of viscoelasticity, Eqs. (3.4) and (3.5) describe the
reduction of mechanical stresses:
σ t = σ 0 ⋅ e −t / τ 0 (3.4)
The relationship between the time-dependant mechanical stress and viscosity is described by
relaxation time, τ0, in Eq. (3.5), where G is the shear modulus of glass:
η
τ0 = (3.5)
G
From Eqs. (3.4) and (3.5) follows Eq. (3.6). With increasing viscosity, the relaxation time also
increases leading to an enhanced time for the reduction of the stress in the glass:
DT
 
 
 − t ⋅G 
 ln( 10 )  A + B 


  T ( t , r ) − T0  
σt = σ 0 ⋅e e   
(3.6)
AR
At room temperature, the structural relaxation time indicates that the glass may flow. This
worries the telecom industry, which is concerned about retaining limits on the geometry of
optical fibres which guarantees their performance many years on (Glaesemann, 1991; Graves,
KH
2005). Glass strengthening methods e.g. tempering, can help to reduce this effect.
3.2.2 Basic physical properties of glass
Single glass is a homogeneous isotropic material having almost perfect linear-elastic
AN
behaviour. Soda lime silicate glass is used mainly for architectural purposes. The basic
physical properties of soda lime silicate glasses are different from other glass types (Table
A.3.2).
Glass has a very high compressive strength and theoretically a very high tensile strength. The
K.P
tensile strength of glass predicted from the molecular structure is around 10000 N/mm2. Glass
usually fails under tensile stresses, because of the very high compressive strength. Glass
normally fails at stresses below the predicted tensile stresses (Table 3.1).
Table 3.1 Mechanical properties of soda lime silicate glass
t by
Properties Symbol Soda lime silicate glass Unit

Young’s modulus (EN 1288-1) E 7.0 × 10 *4
N/mm2
G-modulus G 2.92 × 104 N/mm2
Poisson’s ratio (EN 572-1) µ 0.23 -
h
Hardness (Mohs’) HM 6 -
N/mm2
yrig
Bending strength (characteristic) ft,fl,k

float glass (non heat-treated) 45 (15 design strength) N/mm2
heat-strengthened float glass 70 (35 design strength) N/mm2
tempered float glass 120 (50 design strength) N/mm2
Compression strength (ultimate) fcu 1000 N/mm2
cop
4 2
* In the Appendix of ASTM 1300-04:2004, 7.17 × 10 N/mm is given for Young’s modulus
The surface of the glass has many irregularities which can weaken it, especially when the
glass is subjected to tensile stresses. The irregularities are caused by attack from moisture or
by contact with hard materials, and are continually modified by moisture which is present in
the air (see Chapter 5).
3.3 Strengthening of flat glass
Using glass in load bearing elements usually requires the strengthening of flat glass. All types
of glass (rolled, float, plate, etc.) produced in a flat form are called flat glass (see App. D),
regardless of the method of production. According to the early method of strengthening flat
rolled glass, it was reinforced with wire mesh and was used especially for glass doors and
6
roofing. Wired glass was mainly used in public buildings 20 years before, also in Hungary,
but nowadays it seems to be disappearing. The tendency is to use transparent and large scale
glazing elements in architecture (Fig. A.3.3).
The process of float glass strengthening may usually be physical (thermal) or chemical. Most
of the glass strengthening methods are used to introduce residual compressive stresses into the
outer layers by physical or chemical tempering. The resulting compressed layer helps to close
cracks initiated on the surface (Fig. 3.1), can stop crack propagation and can also increase the
bending strength. Safety glass (see App. D) is a glass which does not fracture into sharp and
potentially dangerous shards when it is broken (Fig. A.3.4). Safety glass may be produced by
heat- or chemical treatment of the surface or by laminating (see Chapter 4).
DT
AR
KH
Fig. 3.1 Vickers diamond pyramid indentation on tempered glass surface a) section view, b)
plan view. Indentation field drives the cracks, residual tempering field opposes them.
(Marshall and Lawn, 1977)
AN
3.3.1 Wired glass
Wire mesh reinforcement in rolled glass helps to hold pieces of broken glass together against
impacting hard objects (Fig. A.3.5). Wired glass was produced by continuously feeding wire
K.P
mesh from a roller into the molten flat glass just before cooling. Nowadays wired glass is
made by laying a wire mesh between two sheets of glass and hot-rolling the sandwich
material.
3.3.2 Heat-treated glass products
t by
Heat-strengthened or heat tempered glasses (see App. D) are produced in a very similar way
using the same processing equipment. Briefly, the glass is heated to just below its softening
point (approximately 650 ºC) and then cooled down by special jets of cold air to create
compressive stresses within the external layers and leaving the core in tension (Fig. A.3.6).
h
The cooling rate defines the glass as heat-strengthened or tempered. While annealing removes
the residual stresses induced during the production process, tempering actually induces a pre-
yrig
determined amount of stress into the glass.

To produce tempered glass, the cooling is much more rapid by creating higher compressive
stresses in the external layers. To produce heat-strengthened glass, cooling is slower and the
resulting compressive stresses in the glass are lower. Because of the resulting compressive
cop
stresses in the external layers, the tempered glass is approximately twice as strong as the heat-
strengthened glass of the same thickness. The tempering process creates a kind of thermal pre-
stressing of the glass pane. The stressed layers produce a birefringent effect (Horváth, Szabó,
2003; Redner, 2000; Pankhardt, 2008a), which is visible in polarised light (Fig. A.3.7).
The stress distribution is different over the thickness of the glass due to the cooling
process. The strength of the glass pane in Region 2 is lower than in Region 1. Maximum
compressive stress will increase from Region 2 to Region 1 (Fig. 3.2).
The annealing process is designed to eliminate or limit internal stresses by strictly
controlled cooling in a special oven or furnace, called lehr. When the glass reaches the
annealing point, the temperature in the lehr is stabilised for a specific length of time
(depending on the glass type, thickness and coefficient of expansion). The stresses present in
7
the glass can relax. The stabilising phase is followed by cooling with a pre-defined
temperature gradient.
compression side
tension side
Fig. 3.2 Distribution of residual stresses over the Fig. 3.3 Distribution of cracks
thickness of tempered glass pane at the middle (R 1) over the thickness of tested float
DT
and near the edges (R 2). Residual compressive glass, (specimen of this Thesis)
stresses near the surface are σ ≈ 120 MPa.
The most dramatic and important difference between heat-strengthened and tempered glass is
AR
the post- failure characteristics of the two products, i.e., their failure pattern, (Figs. 3.3 and
Figs. A.3.8, A.3.9 and A.3.10) (Pankhardt, Balázs, 2006). When tempered glass fractures, it
shatters into tiny pieces with blunt edges. The most common use of tempered glass is in the
automotive industry.
KH
Tempered glass is used in safety glazing applications from 3 to 25 mm thicknesses. It is
required to break into particles of a specified maximum size and number e.g. according to EN
12150-1:2000. Thicker glass (more than 9 mm) cannot be heat-strengthened on current
AN
equipment because of the large residual heat content as it emerges from the furnace
(Loughran, 2003). Tempered and heat-strengthened glass cannot be cut, edged polished, etc.
If surface treatment (sandblasting, acid etching etc., (Fig. A.3.11, App. D) is carried out after
the heat treatment, both will reduce the thickness of the compressed layer and the strength of
K.P
the glass. When heat-strengthened glass fractures, it breaks in shards like non heat-treated
float glass (Fig. A.3.8.b). For this reason, heat-strengthened glass is prohibited by most codes
from use in public areas (in overhead glazing, etc.) unless it is laminated (see Chapter 4).
Single heat-strengthened glass is not a safety glazing material.
t by
3.3.3 Chemically strengthened glass

The chemical process is based on the so-called ion-stuffing technique. Different chemical
elements have different densities. Glass containing sodium (Na) is cooled slowly in a salt bath
of molten potassium (K). The sodium ions will migrate from the glass to the salt, while the
h
potassium ions will move to the surface of the glass where (through ion exchange) they create
a denser and therefore stronger surface layer (not less than 0.1 mm thickness) (Fig. A.3.12).
yrig
Glasses which have been chemically tempered are five to eight times stronger than untreated
glasses. Although chemically strengthened glass can be cut after treatment, the cutting process
causes total loss of the added strength of the glass for 2.5 cm or more on either side of the cut
(Loughran, 2003). The fracture pattern of chemically strengthened glass follows that of heat-
cop
strengthened glass. Single chemically strengthened glass is also not a safety glazing material.
It is used in the ophthalmic and aeronautical industries where glass less than 3 mm thick is
required to have better strength than untreated glasses. It also became popular as a protective
sheet over polycarbonates.
3.3.4 Edge strength of heat-strengthened and tempered glasses
Analysis of the stresses that develop in tempering, bending and lamination were reported by
Gulati (1966), Széwald (2001), Redner (2000), Feingold et. al (2003),
Horváth, Szabó (2003). They strongly recommended stress measuring as a systematic and
necessary part of quality control to eliminate the potential of delayed fractures.
Surface stresses are related to the temperature gradient that results from cooling. Surface
stresses are also related to bending in bent and laminated glass.
8
Fig. 3.4 a) Surface and mid-layer stress b) Tested tempered glass (specimen of this
near the edges (Redner, 2000) Thesis)
DT
As shown in Fig. 3.4, at the surface E compression will develop comparable in size to the
surface F stress, assuming the cooling rate and the heat flow are approximately the same.
Compression on surfaces E and F is balanced by tension in the mid-plane. Under certain
AR
conditions, whenever bending introduces additional surface tension, the tensile region reaches
the glass surface. There is a peak average tensile stress at a small distance from the edge
(typically 12-25 mm). A large positive tensile average is an indication of an undesirable edge
KH
cooling rate and a potential bending problem. Gulati (1966) indicated that also a membrane
tensile stress of 7 MPa could lead to delayed crack growth.
ASTM C1048-85:1985 defines the required surface and also edge compression stresses.
Compressive stress for heat-strengthened glass is at least 24 N/mm2 on the surface (Region 1).
AN
No requirement for edge compression is specified. Required compressive stress for heat
tempered glass is at least 69 N/mm2 on the surface and at least 67 N/mm2 at the edges (Region
2), given in Table A.3.3.
K.P
3.4 Recycled glass

The float glass process theoretically recycles all the glass waste from the melting and cutting
processes. Broken glasses are reintroduced with the raw materials batch mix in the furnace.
The production of glass from waste glasses needs half the amount of energy than from using
only new raw materials. The recycled glass product (App. A, Fig. A.3.13) called Geofil
t by
bubbles was also studied by Nemes (2005) for use as a lightweight aggregate in concrete.
3.5 Conclusions
Requirements on structural glazing in public buildings used in balconies, balustrades,
h
staircases, overhead glazing, etc. are increasing in order to increase safety. In the design of
yrig
glass structures, further additional strength is sometimes required. There are three methods of
strengthening single flat (float) glass: 1) using wired glass, 2) using heat-treatment to produce
tempered or heat-strengthened glasses or 3) chemical treatment for chemically strengthened
glasses. A single glass layer can be considered as safety glass if tempered or reinforced with
wire mesh. Heat- or chemically strengthened glasses can not be considered as safety glass
cop
owing to the fracture pattern (fractures in shards), unless they are laminated.
9
Plastic interlayer materials
Chapter 4
4. Plastic interlayer materials of laminated glass
4.1 Introduction
Laminated glasses are used for automotive and building applications (Fig. A.4.1) because they
offer both safety and strength. When laminated safety glass fractures, the pieces remain
attached to the internal plastic layer, the resultant splinters are held together. Safety of goods
and individuals means limiting the risk of injury in the event of breakage as well as protection
against vandalism and burglary (AGC, 2007) bullets, explosions and fire. Other uses are
DT
sound insulation and decoration (App. A, Figs. A.4.2 and A.4.3).
The discovery of glass lamination also happened accidentally. While Benedictus (1910)
heated a solution of nitrocellulose and accidentally dropped it on a glass pane, he found that,
AR
when the glass fractured it did not shatter. Benedictus patented (British Patent 1.790)
laminated glass in 1910. The production of laminated glass started in 1912 in Great Britain.
Nowadays laminated glass consists of two or more glass layers with one or more plastic
layers between the glass panes. Joining of the glass layers with foil can take place in a
KH
pressurised vessel, called an autoclave (Fig. A.4.4). In the autoclave, under simultaneous
heating of the already processed layers of glass and special plastic, lamination occurs (Fig.
A.4.5, App. D). In the cast in place process, liquid resin (Fig. A.4.6, App. D) is cast between
AN
two glass layers and then is polymerised with UV radiation or by catalysis.
4.2 Influencing parameters of laminated glass
To study the behaviour of laminated glass, it is important to know the physical properties of
K.P
the glass layers and also the physical and chemical properties of commonly used interlayer
materials. The temperature and the rate of loading influence the properties of the polymers
which will also affect the laminate. Water content and aging of the interlayer material can
reduce both the adhesion to the glass surface and bond strength (Van Duser et al., 1999; Glick
et al., 2001; Pankhardt 2008a). Also, the edge strength (Gulati et al., 2001; Pankhardt &
t by
Balázs, 2006) of the laminate is influenced by the interlayer material. In this PhD Thesis, the
applied interlayer materials for laminated glass specimens were resin and EVA foil (ethyl-
vinyl-acetate).
4.2.1 Type of binder materials
h
There are three main kinds of binders: butadiene copolymers, acrylates, and vinyl copolymers
yrig
(Kamath et al., 2004). The chemical composition of the monomer material determines
stiffness properties, strength, water affinity (hydrophilic or hydrophobic balance), elasticity,
durability and aging. Solvent resistance, adhesive characteristics and cross-linking nature are
determined by the type and nature of functional groups of binder. The higher the glass
cop
transition temperature, Tg, the higher will be the dry tensile strength of the binders (Kamath et
al., 2004). Although manufacturers are seeking alternative technologies such as thermal
bonding, chemical bonding still has its advantages and a promising market. Bonding between
the resin and the glass (chemical bond) is extremely strong because of the chemical link
between the resin and the silol (SiOH) groups on the glass surface. These chemical bonds
formed during and after curing are highly stable and resistant. Laminates sometimes offer
better humidity resistance than foil-laminated glasses (Kadri, 2003).
4.2.2 Resins as interlayer materials
In the cast in place process, the adhesive layer thickness is set with the help of a transparent
double-sided adhesive strip (Fig. A.4.6). The major benefits of the cast in place process allows
producing a wide variety of designs and thicknesses while maintaining the benefits of
10
laminated glass, moreover, requires little energy and minimal investment capital. The
manufacturing process operates without pressure and without added heat and there are no
additional costs of autoclaving. As the resin is liquid, it perfectly fills the space between the
glass layers, hence it is ideal for use with imperfectly smooth glass surfaces, such as tempered
and textured glass or non-parallel sheets, such as bent glass. Therefore, the cast in place
process is mostly used for non-standard dimensions of laminated glass. Resins can be also
used for bonding glass to plastic, or plastic to plastic, but PVB is limited only to glass to glass
laminating.
4.2.3 Foils as interlayer materials
DT
4.2.3.1 PVB (polyvinyl butyral) foil
Usually PVB is used in sheet form as a plastic interlayer in the production of laminated glass.
The interlayer for LSG (laminated safety glass) must be made of PVB material with an
AR
ultimate tensile strength of at least 20 N/mm² and ultimate strain of at least 250 % at 23°C.
PVB is an amorphous thermoplastic polymer with highly temperature dependent behaviour.
For the small strains (ε < 0.1 to 0.5 %, γ < 0.5 to 0.8) usually found in building applications,
PVB behaves linearly viscoelastic. The mechanical properties of the interlayer are
KH
furthermore influenced by the water content. Due to its hygroscopic nature, the PVB
interlayer absorbs moisture from the environment. Water content above the optimum value of
around 0.45 % (Van Duser et al., 1999) can reduce both the adhesion to the glass surface and
AN
the rigidity of the bond. The interlayer in LSG has to be protected from the intrusion of water.
4.2.3.2 EVA (ethyl-vinyl acetate) foil
EVA material (Fig. A.4.7) is commonly used in expanded rubber or foam rubber, also used in
K.P
the manufacture of photovoltaic modules, hot-melt adhesive and in sealing. The material has
good resistance to UV radiation, and can have good optical properties (clearness,
transparency). Rheological properties of asphalt including EVA modified binder was studied
by Hussein et al. (2005). The molecular structure of EVA during the laminating process will
change (Fig. A.4.8).
t by
The vinyl acetate (VA) content influences the mechanical properties of ethylene-vinyl-acetate
copolymers. Ethylene-vinyl-acetate copolymers are semi-crystalline if the vinyl-acetate
content is below 50%, but are amorphous above 50% VA content. With the increase of vinyl
acetate (VA) content, the relative Young’s modulus (Ec/Em) of the resulting material will
h
increase, where Ec is the Young’s modulus of the composite and Em is the Young’s modulus
yrig
of the initial (co)polymer matrix. The following theoretical Ec/Em values were calculated by
Peeterbroeck et al. (2007): for EVA 12 it is 4.9, for EVA 19 it is 7.6, as well as for EVA 28 it
is 9.5. Fig A.4.9 illustrates stress versus strain for different EVA compositions.
During degradation, the acetate functional groups are eliminated from the polymer backbone
cop
in the so-called deacetylation leaving an unsaturated polymer backbone. Polyene is formed

after deacetylation of EVA copolymers. Oxidative degradation of polyene is only valid for
EVA copolymers with high mass % of VA (Rimez et al., 2008). There is a lack of systematic
research to understand adhesion issues for EVA. The aging of EVA by storage over ~5 years
considerably reduces the adhesion strength. Lower adhesion strengths were observed for the
blank EVA and non-EVA copolymers, such as poly(ethylene-co-methacrylate) (PEMA) or
poly(ethylene-co-butylacrylate) (PEBA) (Pern et al., 2003). In previous studies (Glick et al.
2001; Tucker, 2002), moisture penetration and retention was observed in laminated glasses.
That moisture condensation occurred in the laminated glasses around the edges (Fig. A.4.10).
In the former laminates, the originally clear and transparent EVA layer would become white
or opaque which gradually can disappear in time by displacing the laminates to dry air
11
conditions. If delamination occurs in the case of resin laminated glass pane, the process is
irreversible (Fig. A.4.11).
4.3 Laminated glass as composite material
Processes like cross-linking change the individual polymeric chains into a three-dimensional
network structure. High molecular weight polymers have a higher amount of molecular
entanglements (Fig. A.4.12) created by molecular intertwining. Chain entanglements improve
such properties as tensile strength, ultimate elongation, etc.
Influencing factors on laminated glass due to interlayer properties are the following: chemical
composition (VA content of EVA); rheological behaviour; thermo mechanical properties;
DT
adhesion properties; time dependent properties (creep, stress-relaxation, breaking of primary
chemical bonds); durability, resistance to humidity, water, UV radiation (storage stability,
hardening due to ageing), etc.
Required properties of interlayer materials, which may also depend on the end-user are the
AR
following: strength (tensile-, shear strength); adhesion (adhesion strength, bond has to be
considered); flexibility (deformation properties); elasticity; resistance to humidity (moisture or
vapour transport through a sealant in the interlayer); resistance to aging (the binder should be
KH
stable and not degraded during storage or use); colour retention (yellowing problems should
be considered); economical (minimising the cost); compatibility with sealants (silicones, see
App. D); other special requirements (such as flame resistance, resistance to chemicals, air,
light, heat, etc.).
AN
4.3.1 Physical properties of interlayer materials
There are three general categories of polymers (Hertz Jr., 1979):
K.P
1. Rubbery: materials with the glass transition temperature, Tg, below room temperature.
2. Rigid: crystalline materials with the Tg above room temperature e.g. polystyrene,
polycarbonate, polymethyl-methacrylate.
3. Partially crystalline: Polymer molecules are often partially crystalline (semi-
crystalline), with crystalline regions dispersed within an amorphous material. Partially
t by
crystalline materials have their glass-transition temperature above or below room

temperature e.g., polyethylene, polypropylene, nylon, PTFE.
A polymer matrix can be described as numerous chains of equivalent length swimming
around and entangling with each other (Fig. A.4.13). Each of these chains rotate, stretch and
h
undergo segmental vibrations. These modes of motion require energy (i.e. heat).
yrig
4.3.2 Thermo-mechanical properties of polymers

The glass transition temperature, Tg, is the temperature above which the materials turn from a
glassy and brittle state to a rubbery state (Fig.A.4.14). G is called the complex shear modulus
and can be separated to G' and G", where G' relates to the energy that is stored in the material
cop
and is called dynamic storage modulus, and G" relates to the energy consumed as heat and
called dynamic loss modulus. The ratio of G”/G’ is noted tan δ (Eq. 4.1 and Fig. A.4.15) and
represents the phase lag between strain and stress during the periodic loading of the material
sample (Brandrup et al., 1989). The angle δ is the tangent vector relationship of G" and G'.
tan δ = G”/G’ (4.1)
Rheological states of an interlayer material as a function of temperature indicated in
Fig.A.4.14 are:
1. Glassy region – In the glassy region, the polymer molecules are completely entangled or
unable to rearrange and are subjected to such fast oscillation or minimal time scales that
they are unable to deform and can fail upon impact. The magnitudes of the values are
different for each polymer, and can also differ within a single polymer depending on
molecular weight and the temperature at which the polymer was tested.
12
2. Transition region – As shown in Fig. A.4.14, tan δ has a local maximum value during the
glass transition temperature range. High tan δ indicates a material that has highly inelastic
or viscous deformation behaviour. Glass transition temperature, Tg, of polymers is the
temperature below which molecules have little relative mobility, it is the mid-point of a
temperature range in which they become rigid and change from being liquid to solid
(Goubard et al., 2009). If a specific volume is measured for an amorphous and semi-
crystalline polymer, the increase of the specific volume as a function of the temperature
appears (Fig. A.4.16).
As the temperature increases (as shown in Fig. A.4.13), the slope of the line changes at the
glass transition temperature, Tg. At this temperature, the amorphous polymer or the
DT
amorphous component of the semi-crystalline polymer changes from a glassy state to a
rubbery state and the material softens considerably. Beyond this temperature, the purely
amorphous polymer does not show further transition and the material is in a liquid state.
AR
However, the semi-crystalline polymer exhibits another transition at a higher temperature
than the glass transition temperature. This is the melting temperature, Tm. (Note that there
is no melting behaviour in the amorphous polymer.) Physical and mechanical properties
of polymers (including thermal expansion coefficient, heat capacity, refractive index etc.)
KH
will change at the glass transition temperature, Tg.
The increased volume above the glass transition temperature is due to the increased space
where polymer chains can freely move around. This space is called free volume. Recent
AN
understanding seems to suggest that glass transition is not a true thermal transition but is a
relaxation of the freezing of the polymer molecules. This phenomenon is universal to all
elastomers and occurs when the fraction of empty space (free volume) in a polymer is
about 2 to 0.5 V% (Hertz Jr., 1979). Crystalline polymers do not show slope change (Fig.
K.P
A.4.16), as the forces maintaining the crystalline state override the increase in molecular
mobility. Earlier, Williams, Landel and Ferry (Ferry et al., 1957), with the WLF equation
(Eq. 4.6) pointed out that all amorphous polymers, irrespective of their chemical structure,
exhibit similar viscoelastic behaviour above their respective glass transition temperatures
(Hertz Jr., 1979).
t by
3. Rubbery plateau – The polymer chains begin to unentangle. This region is called the
rubbery plateau.
4. Flow zone – During melting, both liquid and solid phases exist. Here the chains are
completely or nearly unentangled and the matrix appears liquid-like while G” (liquid-like
h
behaviour) has a greater value than G’ (solid-like behaviour). At extremely low

yrig
frequencies or long time periods, the imposed stress is so slow that their motion can not be
observed on a laboratory time scale, like the flow of a glass pane (Thompson, 2005).
4.3.3 Shear strength of the interlayer
Behaviour of the laminate depends on the shear rigidity and shear strength of the interlayer.
cop
Maximum rigidity supports the behaviour in bending if one or more panes are fractured. The
shear resistance of the laminating material depends on the temperature, rate of loading and
duration of load.
The question arises whether a rather viscous intermediate layer is advantageous for holding
the fragments together. The theoretical shear modulus of the interlayer material by applying
theory of elasticity – Poisson’s ratio of EVA interlayer is 0.32, of PVB 0.5, G = 4.54 N/mm2
for EVA and G = 3.33 N/mm2 for PVB interlayer material. The theoretical value of shear
modulus, G, is 1.5 N/mm2 up to 3.33 N/mm2 in the case of PVB foil interlayer material at
room temperature. That means that EVA foil interlayer material behaves more rigidly at room
temperature (for the same loading rate) than PVB foil interlayer material, which affects the
properties of laminated glasses. With EVA foil, laminated glass panes at room temperature
behave more rigidly than with PVB foil, in the case of appropriately bonded glass layers.
13
The following two questions arise: how does the laminated glass behave at different
temperatures (lower or higher than room temperature)? Will the glass pane laminated with
different types of interlayer materials show the same characteristic behaviour at different
temperatures? One of the aims of this Thesis was to study the temperature behaviour of
laminated glass with different types of interlayer materials.
The behaviour of laminated glass is also affected by the loading rate and by the duration of
loading, Fig. 4.1. For short term load (impact), the interlayer behaves similar to a solid glassy
material and G=G0 (where G0 ≈G’). Over a long duration load such as self weight, the
interlayer tends to creep and G=G∞ (where G ≈G”) (Palotás & Balázs, 1980).. In the previous
studies (Dharani et al., 2005) on LG, the PVB interlayer has been modelled as a linear
DT
viscoelastic material with stress relaxation modulus, G(t). The stress relaxation modulus is
defined by the following equation (Palotás & Balázs, 1980).:
Gt = G∞ + (G0 − G∞ )e ( − βt ) (4.2)
AR
where, G∞= long term shear modulus (rubbery modulus is about 1 MPa), G0= short term
shear modulus (glassy modulus is about 1 GPa), β= decay factor, t=time.
G
(N/mm2)
KH
G0
60 °C 10 °C
AN
G∞
1+E-14 1+E+14 t, (s)

K.P
(s)
Fig. 4.1 Temperature and time dependent shear modulus of polymer materials
4.3.4 Time and temperature superposition
The properties of the interlayer materials, e.g. the shear behaviour measured and determined
at the testing temperature can be different from the properties measured at the service
t by
temperature of the laminate (Fig. A.4.13). A shift of the characteristic curves can be observed
(Fig. 4.1). Shift factors can characterise the viscoelastic properties of polymers at different
temperatures. Shift factors are also used to extrapolate the creep of materials over a range of
temperatures and loading times for a limited set of experiments (Thompson, 2005). The
h
method of reduced time is used to obtain the creep compliance, C(t), at temperatures other
than the reference temperature:
yrig
t
ξ= (4.3)
aT
where, aT= temperature shift factor, t= real time, ξ=reduced time.
cop
Using the elastic modulus E as an example (Brostow et al., 1999), and considering it as a
function of time, the change of temperature is seen to be equivalent to applying the shift factor
aT to the time scale:
E (Ti , t ) = E (Tref , t / aT ) (4.4)
The time-temperature superposition principle (TTS) allows viscoelasic properties to be
predicted without actually performing a long term experiment. A property that takes a long
time to study at room temperature can be realised in much shorter time at an elevated
temperature. When testing many short term properties at various temperatures, these curves
can be superimposed on each other at a reference temperature.
The creep compliance parameter, C(t), is the tensile compliance in Fig. A.4.18 and in Eq. 4.5:
C (t ) = E −1 (t ) = ε(t ) / σ (4.5)
14
where ε(t) is the engineering strain, σ is the engineering stress, and E(t) the tensile Young’s
modulus (Fig. A.4.19). Usually C(t) and t are presented on a double logarithmic scale (Fig.
4.5). From the results at the testing temperature, a master curve can be built, where aT=1:
C (t , T , σ = const ) = C (t / aT , Tref ,σ = const )
The most well-known time-temperature superposition principle is the Willians-Landel-Ferry
(WFL) Equation (Ferry et al., 1957):
η (T ) − C1 (T − Tref )
log at = log = (4.6)
η (Tref ) C2 + T − Tref
where η(Tref) is the viscosity at reference temperature Tref , C1 and C2 are constants and Tg is
DT
the glass transition temperature. The WLF Equation is valid only for temperatures above glass
transition temperatures, for Tg+100°C > Tref > Tg. Below Tg, it is recommended to use the
Arrhenius function. Superposition principle with the Arrhenius function for Tref <Tg is:
AR
Ea  1 1 
log at = − (4.7)
2.303R  T Tref 
where aT is the temperature shift factor, T is the temperature which the data are shifted from,
KH
Tref is the reference temperature the data are shifted to, Ea is the activation energy for flow
below Tref (activation energy is the energy which needs to be overcome for a process to
occur), R is the ideal gas constant 8.34 J/mol⋅°K.
Enough free volume is available for crystalline polymers at temperatures far above Tg or Tm,
AN
where zero shear viscosity can occur. Under these circumstances, the viscosity follows the
Andrade (Andrade, 1914), or Arrhenius equation as a good approximation (Hussein, 2005):
η = Ae Ea / RT (4.8)
K.P
where A is a constant characteristic of the polymer and its molecular weight, Ea is flow
activation energy, R is the universal gas constant and T is temperature.
The TTS is mainly applied to amorphous materials (not well studied for semicrystalline
viscoelastic materials). Schwarzl, Staverman (1952) studied the relationship between
morphology and the TTS principle. They defined two classes of material: class A (thermo-
t by
rheological simple materials), where a change of temperature is equivalent to a shift on the

logarithmic time scale, and class B (thermo-rheological complex materials).
4.4 Conclusions
h
Laminated glass is produced by joining the surfaces of glass panes with polymeric materials
yrig
e.g. resin or foil. Usually PVB is used as an interlayer material for laminated glass. Until now
EVA foil is commonly used in solar cells, but nowadays there are also investigations to use it
as an interlayer material for laminated glass applications. There are numerous influencing
factors on the physical properties of the interlayer material and therefore, on the laminated
glass. The rheological state of an interlayer as a function of temperature can be the following:
cop
glassy region – transition region – rubbery plateau – flow zone. G is called complex shear
modulus and can be separated to G' and G". The tan δ=G”/G’ has a local maximum value
during the glass transition temperature range. The stress transition temperature of the
interlayer is an important factor, where the properties of the interlayer material significantly
change, therefore, affect the behaviour of laminated glass panes. By lowering the temperature
(testing temperature or service temperature of the glazing), the amount of free volume in the
interlayer material will decrease. It is often more important to predict the behaviour of the
laminate at higher temperatures and in long term behaviour. Instead of performing
experiments for a long time at room temperature, we can heat the specimen (free volume in
the interlayer material will increase), and record series of events in hours which otherwise
would take many years.
15
Strength and durability of glass
Chapter 5
5. Strength and durability of glass
5.1 Introduction
Strength of different glass types is specified in standards or technical guides. Strength is
influenced by many factors during the production or lifetime of glass. Specified strength
values give guidance also to the manufacturers, e.g. rate of prestressing in heat-strengthened
or tempered glass panes. There are many affects which can decrease the strength of a glass
pane. The strength of the glass product depends e.g. on edge finishing techniques (App. A,
DT
Fig. A.5.1) as well as the handling of glass panes during and after the manufacturing process.
Transportation of glass can cause flaws on the surface, which can propagate during the
lifetime of the glass. The strength of glass decreases over time and affects the durability
AR
which is an important factor in the design of glass structures, especially load bearing glasses.
Chemical composition of the glass and of the environment are also affecting factors on the
strength and durability. The durability of glass products are also influenced by the following
(EN 572-9:2004): design of glass support (e.g. prevention of direct contact of glass and
KH
supporting members), design of dilatations (expansion of glass support e.g. due to adsorbed
moisture from the air or other sources as well as thermal expansion can cause stresses in the
glass pane), movements or vibrations of the building or construction due to various actions,
AN
deflection and racking of the glass supports, etc.
5.2 Corrosion and weathering of glass
5.2.1 Effect of surface condition
K.P
Tests on the bending strength of glass (e.g. according to EN 1288-1:2000) have shown that
glass behaves almost ideally as a linear elastic and brittle material. The brittleness means that
contact with any hard object can lead to surface damage in the form of ultra-fine,
submicroscopic cracks. Submicroscopic cracks are practically unavoidable during the normal
t by
handling of glass. Surface cracking is a major factor in reducing mechanical strength, whereas
the chemical composition of the glass has only a small and in some cases negligible
significance.
5.2.2 Leaching of glass
h
Water attack, chemical air pollutants and alkali attack as well as physical damage can
yrig
decrease glass strength. Glass is exposed to wet (aqueous) environmental conditions usually
by moisture condensation on the surfaces. Even water alone can damage the surface of glass.
Solid particles from chemical air pollutants can turn into destructive compounds with water
condensation (Bene, Pankhardt, 2010d). Glass in contact with water enters into a series of
cop
chemical reactions which can result in alkaline solutions. A very common cause of such a
problem is when alkalis leach from precast concrete panels by rain.
Leaching is a kind of glass corrosion whereby sodium and potassium ions are dissolved from
the glass structure. During leaching of glass, the pH value will change and water molecules
infiltrate the glass as well as alkali ions go into solution creating a layer which eventually
causes opacity. Initial attack may cause only opacity, whitening of the glass surface due to the
changes in the glassy silicate structure. At this stage, a light polishing or special chemical
treatment would probably restore the glass surface. When the condition worsens over time,
further deterioration of the glass surface can cause cracking (Fig. 5.1) and in this case there is
no practical method of restoring the glass surface. Glasses stored in humid conditions, e.g. job
site, unconditioned places, improperly designed or maintained areas, or subject to frequent
washing, may show signs of leaching (Fig. A.5.2). Also when the relative humidity is high,
16
condensation can occur as the air next to the glass is cooled down. In some climates, this
cycle of wetting and drying may occur frequently during humid seasons.
Fig. 5.1 “Griffith flaws” on a glass surface. Cracks caused by sodium vapour-deposition, due
to abrasion of the surface (Schittich et al., 1999)
5.3 Chemical resistance of glass
DT
5.3.1 Influence of sharp impactor
Flaws during optical observations can behave in the same way as macroscopic cracks behave.
AR
The question now arises: how far we may extrapolate the fracture mechanics theory for
microscopic flaws. In particular, what response might we expect as we enter the subthreshold
region? Systematic studies by Dabbs, Lawn and Kelly (1982) using Vickers indentation
load P (using pyramidal “sharp” indenter, App. D) as a primary test variable provide some
KH
answers to these questions and indicate the resistance to crack extension. The fatigue strength
properties of glass containing Vickers indentation flaws are described by Mould et al. (1959),
Wiedernhorn and Bolz (1970), Marshall et al. (1977), Dabbs et al. (1982), Lawn et al. (1985),
AN
Wiedernhorn et al. (2002), etc. Once a crack is initiated by Vickers indentation, it grows
spontaneously in the radial and lateral directions (Fig. A.5.3). Cracks are produced during and
after contact. The cracks continue to extend with time especially in wet environments. Above
some threshold in the loading, cracks initiate from the deformation zone, where high stress
K.P
concentrations exist. Regardless of the source of crack nucleus the threshold of the
indentation load, Pc, can be determined. For silicate glasses, the threshold is about Pc≈10 N
(Lawn, 1985) under inert test conditions. The threshold of load can be easily attained with
contacts by sharp particles in everyday handling. The threshold load, Pc is found to be (Lawn,
Evans, 1977):
t by
3
 E  K 
P c = A  c  K c (5.1)
 H  H 
where A is constant (different for each crack system), E is Young’s modulus, H is hardness,
h
Kc is toughness. Eq. (5.1) indicates why glasses (and most ceramics) are brittle in comparison
to metals, which are tougher and can withstand relatively high stress concentrations without
yrig
crack generation.
Deformation induced radial cracks (post threshold state) are also present due to the Vickers
indentation load (Figs. A.5.4, A.5.5). In the case of removing the residual stresses (with heat-
treatment), higher strengths can be reached with reduced fatigue susceptibility. When radial
cop
cracks are not present (subthreshold state), in the case of sufficiently small contact loads, the
measured data deviate from the extrapolated predictions of macroscopic crack theory (Fig.
A.5.6). The observed strength is higher than the equivalent post threshold levels, with
increased fatigue susceptibility and greater scatter (Lawn et al., 1985). The sharp-crack
concept of flaws remains valid down to the threshold load for crack initiation, but below this
threshold the crack precursor processes control the failure properties. The results have shown
that the nature of flaws is an important factor in fatigue characterisation. A flaw of given size
may respond in different ways, depending on whether there are residual stresses present or
cracks have been formed. This is also important in the context of lifetime design. By
removing residual stresses from flaws in glass, lifetimes can be increased.
Flaw history has a potential role in the fatigue behaviour. The aging of newly created flaws
should be beneficial and residual stresses will tend to relax over time (especially in reactive
17
environments). If the flaws begin their life in the subthreshold state in a high strength region,
the danger exists that radial cracks may spontaneously initiate during service time. This
danger may dramatically increase under certain conditions, e.g. unfavourable chemical and
thermal environments, and failure may occur. In the case of small scale natural flaws, their
abrupt formation increases with prolonged exposures to aqueous environments (Lawn et al.,
1985). Under such conditions no appropriate strength control exists (including proof testing)
which would be capable of predicting the lifetime.
5.3.2 Loss of strength by aging
Glass mechanically fails by subcritical crack growth. When glass is subjected to an applied
DT
stress, cracks located in the glass surface grow until they reach a critical size for failure. The
crack velocity then accelerates and failure occurs almost instantaneously. Water or water
vapour is the main chemical agent for subcritical crack growth in glass (Wiederhorn, Dretzke
and Rödel, 2002). Without water, cracks either do not grow or require higher stresses than in
AR
the presence of water. Previous studies (Marshall, Lawn, 1985) have already demonstrated
similarities in the general damage configurations associated with indentation, abrasion,
scratching, and particle impact (Fig. A.5.7), all of which are caused by penetration of the
KH
surface by hard, sharp particles. Before indentation fracture mechanics can be extended to
natural flaws (caused during surface finishing and service), correlations in the responses to
applied stresses need to be established.
Loading threshold depends on the loading rate and time (Lawn, 1985). The influence of
AN
chemical agents of the environment is strong on the development of critical crack initiation.
Water can lead to the reduction of loading threshold, which decreases further when the time
of contact is prolonged. The presence of water strongly reduces the delay times (Fig. A.5.8).
K.P
5.3.3 Strength recovery by aging

5.3.3.1 Surface flaws
The nature of surface flaws in glass can be highlighted by indentation fracture theory. The
mechanical response of appropriately developed indentation cracks to applied stresses can be
t by
described. In direct observations of the crack evolution, residual elastic-plastic stresses have
an influence on stabilising effects. It has been demonstrated from acoustic scattering
experiments that machining flaws have essentially the same characteristic response as the real
microcracks. Aging experiments on indented, machined and abraded surfaces, in aqueous
h
environments, have been described by Wiederhorn (1967), Michalske (1977), Ito and
yrig
Tomozawa (1982), Marshall and Lawn (1985), Wiederhorn, Dretzke and Rödel, (2002), etc.
Strength recovery with aging time was observed (Table A.5.1).
The aging study of Lawn et al. (1985) for indented soda-lime glass specimens has shown
(Fig. A.5.9) that in the case of silicone oil aging environment (with trace quantities of water) a
strength increase of ~25 % is realised after several hours, similar in trend to the result for
cop
abrasion damage. Residual stresses remain active during exposure to an aqueous environment,
which is directly attributable to secondary crack growth effects. The Charles–Hillig theory
(1965) of crack tip corrosion (Fuller et al., 1980; Chuang et al., 1992) attributed the low
stress limit for fracture as the stress at which crack tip blunting occurred, the crack no longer
propagated as a sharp crack. The strength of glass with newly formed flaws tends to increase
with aging in aqueous environments. This strengthening effect has been reported for many
damage configurations, e.g. scratching abrasion and cutting.
5.3.3.2 Effect of heat-treatment on aging
In a classic early study on aging of soda-lime glass in water Mould et al. (1959) published the
following:
18
- The strength of specimens with grit blast abrasion and sharp particle scratch damage, after
immersion in water for one day at room temperature, increased by 30 % to 60 %.
- Annealing (heat-treatment) increased the strength of newly damaged specimens by
approximately the same amount as the aging treatment.
- The strength of specimens with annealed damage did not increase further with more aging.
- Annealed damage was less susceptible to fatigue in water than fresh damage (i.e. time-
dependent reductions in strength).
The same characteristics have been observed in Vickers indentation flaws (Fig. 5.2).
DT
AR
Fig. 5.2 Vickers indentation (P=50 N) in soda-lime glass, a) immediately after indentation,
b) after aging in moist environment for 1 hr. (Note the development of radial cracks in
KH
polarised light reveals strong stress birefringence) (Marshall, Lawn, 1985)
The fracture mechanics approach for analysing the fatigue of glass in reactive environments is
based on the assumption that fatigue failure occurs by subcritical extension of a dominant
microscopic crack to an instability configuration. The aging study of Lawn et al. (1985) for
AN
indented and post-indentation annealed soda-lime glass specimens has shown (Fig. A.5.10)
that the annealing effectively removes all residual driving forces on the indentation cracks,
leading to an increase in strength.
K.P
5.4 Crack development during fracture of glass

5.4.1 Hardness
The objectives of any general mechanical description are the material parameters, such as
elastic (Young’s) modulus E, hardness H, and toughness KIc. Hardness of silicate glass:
t by
Knoop HK0,1/20 =5 - 6 GPa, Mohs H= 6 (-). The greater the ratio H/E, the greater the role of the
elastic component in the contact properties. Glass with a value of H/E approaching 0.09, lies
at the top end of materials, which is attributable to the relatively high rigidity of the silicate
network (Lawn, 1985).
h
5.4.2 Crack growth in glass

yrig
Cracks can propagate in glass over a wide range of values of tensile stresses
(Fig. A.5.11). There exists a lower limit to the stress intensity factor below which cracks do
not propagate. Subcritical crack propagation at higher levels of stress intensity factor is
influenced by humidity, temperature and chemical agents. Above a critical stress intensity
cop
factor, crack propagation is very rapid and leads to (almost) instantaneous failure. The
subcritical crack propagation is also affected, e.g. by the rate of load increase and/or the
duration of static loading.
Wiederhorn (2002) observed that his results could be separated into three qualitatively
different regions indicating three distinct mechanisms controlling crack growth: (I) a slow
growth region in which the crack velocity exponentially depended on the load, (II) a plateau
region in which crack growth was limited by the diffusion rate of water from the crack tip and
(III) a region in which crack growth was rapid and independent of the amount of water in the
environment.
Failure originates from cracks with atomically sharp tips. When strength controlling flaws are
considered to be intrinsically “blunt”, the strength is then related to the radius of curvature of
the tip as well as to the length of the crack. In any failure process, both flaw lengthening
19
(subcritical crack growth) and flaw sharpening (crack initiation from a notch-like defect) can
occur and the slower controls the kinetics. Transitions of flaw lengthening and sharpening
mainly happen in large-scale (double-cantilever beam) cracks in glass. When the stress
intensity factor, KIc, is lowered below ≈ 0.25 MPa∙m1/2, loading is needed to restart the
propagation (Marshall, Lawn, 1985). Wiederhorn and Bolz (1970) also published
experimental data for soda-lime glass on stress intensity factor threshold of 0.25 MPa∙m1/2 in
water.
Studies of Michalske (1977) and Gehrke et al. (1991) as well as Wiederhorn et al., (2002)
indicated that cracks in soda–lime–silicate glass held below the static fatigue limit experience
a time delay in restarting their growth when the stress intensity factor is again raised above
DT
the fatigue limit. When the stress intensity factor was less than about 0.4 MPa∙m1/2 to restart
the crack, the times to restart the crack ranged from approximately 1000 s to greater than
6000 s. At higher restart stress intensity factors (>0.45 MPa∙m1/2) cracks restarted
AR
immediately (Wiederhorn et al., 2002).
5.4.3 Crack direction in glass
AFM (atomic force microscope) results published by Wiederhorn et al. (2002) showed that
KH
below the apparent fatigue limit, cracks change their direction of growth by an angle of about
3° to 5° to the original crack growth direction (Fig. A.5.12). Increasing the stress intensity
factor to a value that is greater than the fatigue limit again changes the direction of crack
growth and causes the crack to grow on a flat plane. Investigations by Wiederhorn, Lawn and
AN
Hockey (1979) on the decrease of strength by sharp-particle impacted glass surfaces
highlighted that a complex analysis is required to modify the theory of elastic/plastic
indentation fracture.
K.P
The effect of tangential loadings (Figs. A.5.13, A.5.14) on both elastic and residual
components of the contact stress field should be taken into account. Design of ceramic
systems for optimum resistance against strength loss in potentially dangerous contact
situations could be possible with the new identified parameters.
The problem mentioned above shows how difficult it could be to make a lifetime
t by
prediction on glazing under service conditions. Also in the case of statistical methods, it is
hard to determine what kind of impacts can be dangerous which can also be affected by
environmental conditions.
5.5 Conclusions
h
Calculations on the strength of glass as a material are influenced by many factors, e.g. surface
yrig
treatment as well as the environmental conditions during service time. To predict the lifetime
of glazing is difficult because of the high number of influencing factors on its durability.
Lifetime design procedures should always be based on the actual surface conditions of glass,
as far as possible. It makes difficulties for those who intend to characterise strength properties
cop
in terms of statistical distribution functions. Appropriate design of glass elements (supports

etc.) also influences the durability. In spite of the careful manufacture and transportation of
glass panes, impacts with sharp particles or environmental impacts can cause defects on the
surface. Aging of newly created flaws should be beneficial and residual stresses will tend to
relax over time (especially in reactive environments). Under such conditions, no appropriate
strength control exists (including proof testing) which would be capable of predicting the
lifetime. Mould et al. (1959) showed that the strength of specimens containing abrasion
microcracks can increase with the aging time (i.e., time between abrasion and testing to
failure). Recent studies (Marshall et al., 1977) have shown that surface strengthening can lead
to substantial improvements in degradation resistance. Therefore, in outdoor conditions, when
the glass surface is exposed to humidity etc., tempered or heat-strengthened glass should be
used.
20
Bending strength of float and laminated glasses
Chapter 6
6. Bending strength of float and laminated glasses
6.1 Introduction
Since glass is used in buildings (App. A, Fig. A.6.1), the most important property for
resistance is the bending strength of single or laminated glass panes. Safety glass should be
used for glazing in public buildings. The physical properties of glass are the basis for
designing float glass in buildings given by standards e.g. EN 1288-1:2000. For special
applications, tests can be carried out to fulfil the requirements. The test methods on bending
DT
strength are specified in e.g. EN 1288-2:2000 and EN 1288-3:2000.
6.2 Influencing factors on bending strength of single or laminated
glasses
AR
6.2.1 Single float glass pane
The bending strength of a single glass is influenced by the following factors (EN 1288-
KH
1:2000, Pankhardt, 2008b):
a) heat treatment, b) surface condition (e.g. non-slip characteristics for stairs), c) rate and
duration of loading, d) area of surface stressed in tension, e) relaxation (Chapter 3),
f) ambient medium, through stress corrosion cracking as well as healing of surface damage
AN
in the glass (Chapter 5), g) age, i.e. time elapsed from the last mechanical surface
treatment or modification to simulate damage, h) ambient temperature, i) edgework.
Some of the influencing factors (e.g. rate of loading) mentioned above can be well defined
K.P
during the service life or testing of a glass pane, but some of them (e.g. change of surface
condition over time) are difficult to define. The presence of irregularities and their
modification by moisture contributes to the change of the properties of glass which can
influence the bending strength. Defects may induce surface cracks. The development and
orientation of cracks continuously changes during the lifetime of the glazing. Some cracks can
t by
be stopped, some can start to propagate in different environmental conditions (Chapter 5).
Therefore, loading strength in not directly predictable just based on the information of surface
condition. Recent studies (Chapter 5) have shown that surface strengthening can lead to
substantial improvements in degradation resistance, therefore, predictions on bending strength
h
of heat tempered or heat-strengthened glasses are easier.

yrig
6.2.2 Laminated float glass pane

The structural behaviour of laminated safety glass lies between two limits. The lower limit of
the load bearing capacity is the so-called layered limit, where the glass panes react without a
shear bond. The upper limit of the load bearing capacity is the monolithic limit where all glass
cop
panes are rigidly connected (Schittich et al., 1999). For both limits, stresses in the glass pane
can be calculated using the known formulas and models. In reality, the maximum stress of the
laminate lies between those two limits. The bending strength of the laminate is influenced by
the properties of the interlayer material (aging, delamination, behaviour at different
temperatures, etc.) (Chapter 4).
When laminated glass began to be used in significant quantities for the architectural
glazing industry in the 1960’s, building codes defined a strength factor of 0.6 relative to
monolithic single glass of the same thickness (Savineau, 1999). The basis to define this factor
numerically is unclear, although in bending tests the load bearing capacity (Fmax) of a layered
two-ply laminate without an interlayer is 0.5 times a monolithic pane of the same thickness
(Pankhardt, 2008b). Nowadays, there is a general consensus to increase the glass strength
factor. Strength factor, SF, of laminated glass is defined by Eq. (6.1):
21
maximal resisting force of laminated glass (6.1)
SF =
maximal resisting force of monolithic single glass
Depending on the shear modulus and thickness of the interlayer, the laminate can exceed the
strength of the equivalent monolith glass having the same total glass thickness (see Chapter
6.2.2.2). However, this effect is only valid as long as the interlayer material remains stiff.
Increase of temperature results in a decrease of the stiffness of the laminate, which also
affects the strength factor.
Usually the glass plies are modelled (Suetomi et al., 2001) as a linear elastic material with
isotropic linear elastic material properties (with elastic constants E and µ) (App. A, Fig.
A.6.2). The interlayer has viscoelastic material behaviour (Dharani et al., 2005; Suetomi et
DT
al., 2001) that can be characterised by time and temperature dependent material properties
(see Chapter 3). Both relaxation and creep can be taken into account with linear viscoelastic
models (Van Duser et al., 1999). These are only valid for small strains and do not include
AR
some observations such as hardening of polymers with large shear deformations (tan γ > 0.5)
at room temperature (Sobek et al., 1999). In the relaxation model (App. A, A.6.2), the time
dependent shear modulus G(t) is given by Eq. (6.2):
 λi  − 
t
KH
n
G (t , T ) = G0 + ∑ Gi   1 − e λi 
(6.2)
i =1  t  

where, Gi, is the shear modulus of the ith spring element and the viscosity ηi of the ith dashpot.
AN
Retardation time is λi=ηi/Gi and varies with temperature, T. Experimental data is required to
evaluate all values of Gi and λi.
When LG is exposed to dynamic loads (for example in the facings of building exteriors,
glass slab, etc.) dynamic or impact tests should be done (App. A, A.6.3-6). In the first stage,
K.P
the outer (impacted) glass ply fractures while the inner glass ply may remain unbroken. In the
second stage, the inner and outer glass plies are also fractured (see Chapters 9 and 12). There
are two different ways in which the glass fails under the impact of a hard missile (Glathart et
al., 1968). If the thickness of the glass is large, a hard missile initiates a crack, the so-called
Hertian cone crack on the upper surface outside the contact zone. When the thickness of the
t by
glass plate is small, the glass breaks on the lower surface under the contact centre (Dharani et
al., 2005).
6.2.2.1 Laminate without bonded layers
h
If the slabs are laid loosely on top of each other (Fig. 6.1.a), then each slab carries its share of
yrig
the total load in proportion to its bending strength, e.g. with continuous lateral support along
the edges of two layers of thicknesses h1 and h2 and unit width, an external uniform
distributed load, q, would split in the ratio of (h1/h2)3. The maximum bending stress would
then be σ bB =1.5q(Ls/h)2, in the case of symmetrical arrangement h1=h2=h/2. Translucent or
cop
opalescent glasses are produced without a bond, only with a paper or gauze interlayer laid
loose between the panes and only the edges are glued.
Fig. 6.1 Deflection behaviour and stress distribution in a) panes laid without bond on top of
each other, b) elastically bonded laminated safety glass, c) rigidly bonded or monolithic
panes (Schittich et al., 1999)
22
6.2.2.2 Laminate with bonded layers
In the case of two layers bonded with a shear-resistant interlayer, the loads can no longer be
split in proportion to the strengths but are carried as a composite. The maximum bending
stress for a rigid bond (Fig. 6.1.c) would be σ bB =0.75q(L/h)2. The strength in the glass panes
is influenced by the shear transfer of the interlayer (Wölfel, 1987). When the bond is efficient
and the strain of the interlayer is small, the composite behaves almost monolithically (App. A,
Fig. A.6.7). Due to the combinations of different materials (for example metal/plastic or
glass/plastic) special influences such as temperature must be taken into account, which have
significant effect on the load bearing performance. In the case of the so-called rigid bond, the
DT
load bearing capacity of the laminated glass can be over-estimated. Norville (1997) published
experimental strength data on laminated glass specimens manufactured with a 2.28 mm
polyvinyl butyral interlayer which exceeded the strength of laminated glass specimens having
the same dimensions but which used a 0.76 mm thick interlayer, as well as the strength of
AR
monolithic glass samples having a comparable thickness (Fig. 6.2).
KH
AN
Fig. 6.2 Results of bending tests under uniform loading at room temperatures, where
specimens were supported along two edges. Squares indicate monolithic specimens with
K.P
thickness of 6 mm, diamonds, two plies of glass with thickness of 3 mm laminated with 0.76
mm PVB interlayer, triangles, specimens laminated with 2.28 mm of PVB interlayer thickness
(Norville, 1997)
Hooper (1973) used four-point bending beam tests for observations concerning the
t by
performance of LG in bending. Researchers (Behr et al., 1984; Vallabhan et al., 1987;

Norville, 1997, etc.) published data of tests concerning the behaviour of monolithic and of
PVB laminated specimens. Test results of LG plies under uniform loading indicated that the
strength of LG specimens increases with increasing PVB thickness at room temperatures. A
theoretical LG beam model provides a basic understanding of how the interlayer works by
h
transferring horizontal shear in bending. In the theoretical beam model glass behaves linear
yrig
elastically until the fracture of the beam (App. A, Fig. A.6.8).

Effective section modulus, Weff, was defined for PVB laminated specimens by Norville
(1997).
cop
a) b)
c) d)
Fig. 6.3 a, b) Monolithic glass beam with thickness of ‘2h’ and c, d) laminated beam with
glass layer thickness of ‘h’ and an interlayer thickness of ‘hint’ (Norville, 1997)
23
The PVB interlayer material maintains the spacing and transfers some portion of the shear
force ‘H’ between the glass plies (Fig. 6.3). Factor q defines the ratio of horizontal shear
forces of the monolithic beam with thickness of 2h and the horizontal shear forces of the LG
beam (consisting of two glass plies each of thickness h) manufactured with PVB. Factor q
was determined from experiments. The ratio of horizontal shear force transferred by PVB
varies principally between 0 ≤ q ≤ 1. In the case of q=0 there is no shear transfer between the
glass plies; q=1 means monolithic behaviour. Effective section modulus, Weff, was defined by
Eq. (6.3) for PVB laminated specimens (where b=1 mm):
M  h q ⋅ h q ⋅ hint 
Weff = = b⋅h⋅ + +  (6.3)
σ 3 3 2 
DT
Krüger (1998) defined a coupling parameter, κ, which represents the bond of the glass
layers to the interlayer, Eqs. (6.4 to 6.6).
AR
KH
Fig. 6.4 Three-point bending tests by Krüger (1998). Concentrated load F=20 N was applied
at the midspan of the beam (span Ls=400 mm, width b=30 mm) which consisted of laminated
glass (4 mm glass / 0.76 mm PVB / 4 mm glass), where h is the thickness of the glass layer,
AN
hint is the thickness of interlayer.
The centre deflection can be calculated by Eq. (6.4):
F ⋅ L3
y=
K.P
(6.4)
48 ⋅ ( E ⋅ I ) beam
where, E and I are the Young's modulus and the moment of inertia of the beam, respectively.
The moments of inertia of the glass layers, Ig, of the interlayer, Iint, and the coupling term, Ic,
contribute to the total amount of the moment of inertia, I. The cross section of the beam is
t by
shown in Fig. 6.4.

( E ⋅ I ) beam = E g ⋅ I g + Eint ⋅ I int + Ec ⋅ I c ≈ E g ⋅ I g + Ec ⋅ I c (6.5)
3
h3 h
Ig = 2 ⋅b ⋅ , I int = b ⋅ int , I c = κ ⋅ b ⋅ h ⋅ ( h + hint )
2
(6.6)
12 12
h
In the case of Iint << Ig, the moment of inertia of the interlayer can be neglected. The value of
yrig
parameter κ should be between 0 ≤ κ ≤ 1 (App. A, Fig. A.6.9). Parameters κ and Ec were

determined from experiments and can be dependent on time and temperature, κ(T, t).
In the case of κ = 1 the total moment of inertia is equal to the moment inertia of a monolithic
beam with the thickness hbeam = 2×h+ hint. In the case of κ= 0 the glass layers are not bonded
cop
and I =Ig.
The concept of effective thickness was developed by Wölfel (1987). The effective
thickness can be used instead of the actual thickness of the laminated glass. Eqs. (6.7 to 6.9)
are given to describe the shear coupling between two glass layers bonded with interlayer
material. The shear coupling depends mainly on the shear stiffness, G, of the interlayer and
the properties of glass as well as the geometry of the laminate. The shear transfer coefficient,
Γ , by Wölfel is defined by Eq. (6.7):
1
Γ= (6.7)
EJ s hint
1 + 9.6 2
Ghs a 2
J s = h1hs22 + h2 hs21 (6.8)
24
hs h1 hh
hs1 = , hs 2 = s 2 , hs = 0.5(h1 + h 2 ) + hint (6.9)
h1 + h2 h1 + h2
where hint - interlayer thickness, h1 - thickness of 1st glass layer, h2 - thickness of 2nd glass
layer, E - Young’s modulus of glass, a - smallest side length, G - shear modulus of interlayer.
The shear transfer coefficient, Γ, varies between 0 < Γ < 1.
For calculating the deflection of a laminate, the effective thickness, heff, can be used as
follows:
heff = 3 h13 + h23 + 12ΓJ s . (6.10)
The primary interlayer property that influences the strength and deflection is the shear
DT
modulus, G, by Wölfel’s calculations. The greater the shear resistance of the interlayer, the
more effective is the bond between the glass layers. The effective laminate thickness, heff
reaches the “real” (or total) thickness of the laminate in the case of Γ →1.
AR
6.2.2.3 Edge strength of laminate
The edge strength of the laminate is influenced by the edge strength of the glass layers
(Chapter 3) and is also influenced by the interlayer properties (Chapter 4). Water content and
KH
aging of the interlayer material can reduce both the adhesion to the glass surface and bond
strengths, therefore, the stresses can be increased in the edge region of a laminate.
Delamination of glass laminates and their edge stability often creates questions especially in
AN
exposed and butt-joined glazing. By applying laminated glass, the designers and developers
have to take into account these effects and protect the free edges (or holes near point fixings)
for example by water isolating materials which do not react with the interlayer materials.
K.P
6.3 Standardisation
6.3.1 Standards, technical guidelines for glazing that support human
loads
Taking into account the shear transfer between the glass panes can lead to economic design of
t by
glass thickness for some applications e.g. vertical glazing and wind loads. Consideration of
shear transfer between the glass panes is not always allowed (see Chapters 9 and 10).
Glazing, mountings and substructure must be designed and tested to withstand the loads
according to the existing standard. The values for maximum allowable or design stresses
h
(including already a global safety factor of 2.4 against 5%-quantile value of fracture) and
minimal bending strength values given by the product standards are summarised in App. A.,
yrig
Table A.6.1 for different types of structural glasses.

In the case of human loads carried by glazing elements (e.g. stairways, podiums)
construction permission on an individual basis is always required. Impact resistance and
residual load bearing capacities with testing must be submitted (App. A, A.6.3 to 6.6). Glazing
cop
designed to support human loads may be supported at the edges or at individual points and
must be manufactured from laminated safety glass consisting of at least three layers of
tempered or heat-strengthened glass. The external surface of the upper glass layer must be
non-slip. Usually the maximum allowed deflection of the load bearing glass may not exceed
Ls/200 or Ls/100 of the effective span. Load bearing glass structures are sometimes called "a
building part without code of practice", therefore, a special approval procedure must be
carried out. In Germany, it is the ZiE "Zustimmung im Einzelfall" individual test (App. A,
Table A.6.2). This approval must be received from the AbZ “Allgemeine bauaufsichtliche
Zulassung” General Building Supervision Authority.
25
6.3.2 Calculation of bending strength in four-point bending according to
EN 1288-3:2000
A weighted average of the tensile bending stresses can be calculated by applying a factor k to
take into account non-uniformity of the stress field, (see factor k in Eq. (6.11)) and the
calculated bending stress is called the effective bending stress. Equivalent bending strength
can be defined in the case of e.g. patterned glass, for which the irregularities in the thickness
do not allow precise calculation of the bending stress. Usually, glass in bending tests is
subjected to four-point bending, (EN 1288-3:2000, App. A, Fig. A.6.10) or to a coaxial double
ring test (EN 1288-2,5:2000). To determine bending strength in four-point bending the
following formula can be applied:
DT
 3( Ls − Lb ) 
σ bB = k  Fmax + σ bG  (6.11)
 
2
2bh
AR
where, b-width of specimen; h-thickness of specimen; Ls-distance between the centre lines of
the supporting rollers; Lb-distance between the centre lines of the bending rollers; y-central
deflection of the specimen; k=ke-dimensionless factor as function of y/h (to determine the
stress at the middle of the span k=1); σbB-bending strength; σbG-bending stress imposed by the
KH
self-weight of the specimen.
Bending stress imposed by the self-weight of the specimen is:
3ρgL2s
σ bG = (6.12)
AN
4h
Factor k is used when it is required to determine the bending strength of glass where the
effects of the edge conditions are important. For calculating the overall bending strength or
equivalent bending strength of the surface area including the edges, the value k=1 shall be
K.P
used. For calculating the bending strength or equivalent bending strength of the free edges of
the glass pane, k=ke shall be used. The appropriate value of ke for use in Eq. (6.11) shall be
obtained from Fig. 6.5, which gives the value of k=ke as a function of the value of y/h.
h t by
k=ke
yrig
cop
Fig. 6.5 Dimensionless factor k=ke as a function of y/h for 6, 12, 19 mm thickness of glass
(EN 1288-3:2000)
Simple theory (beam model) assumes that there are no stresses across the width of a beam
when it is subjected to bending along its length. Poisson effect (µ=0,23 for float soda lime
silicate glass) generates stresses across the width of wide beams, therefore, the longitudinal
stress cannot be regarded as uniform across the width in the case of wide specimens. The
effect of edgework also influences the strength of the glass in edge regions, which can be
taken into account in factor k. To determine the deflection in four-point bending, the
following formula can be applied (EN 1288-3:2000):
26
3Fmax  LS LS Lb 
3 3 2
Lb
y =  + −  (6.13)
4 Ebh 3  3 6 2 
6.3.3 ASTM E 1300-04:2004
The ASTM E 1300-04 standard contains additional design tables that cover a wider variety of
edge-support conditions, as well as tables (App. A, Figs. A.6.11, A.6.12) for laminated glasses.
The 42 non-factored load charts allow the user to determine load resistance for monolithic,
laminated and insulating glass units of rectangular shape with continuous lateral support along
up to four edges. In 2002, the ASTM E 1300 standard moved from a required 60-second
DT
duration load resistance to a three-second load resistance (Block, 2002). In 2004, an appendix
was added to ASTM E 1300-04. The laminated glass non-factored load resistance charts were
added to the 2002 version of the standard and were based on PVB performance. The new
appendix allows for laminated glass fabricated with other types of interlayer, if the other
AR
material has a Young’s modulus greater than or equal to 1.5 MPa at 50 °C at equivalent three-
second load. The shear modulus of the other interlayer must be greater than or equal to that of
PVB. To determine load resistance, Eq. (6.14) should be applied:
KH
LR= NFL × GTF × LS (6.14)
where, LR is load resistance, NFL is non-factored load (from charts), GTF is glass type factor
(from tables), LSF is load share factor (see App. D).
In ASTM E 1300-04 determined lateral deflections of the supported glass edges may not
AN
exceed LS/175 of the length. ASTM E 1300-04 differ from EN 1288-1-5:2000, where the
allowed deflection may not exceed LS /200 of the effective span. ASTM E 1300-04 gives
71.7 × 103 N/mm2 in Appendix for Young’s Modulus in the appendix, while in
K.P
EN 1288-3:2000, 70 × 103 N/mm2 is given.

6.3.4 Technical guidelines given by glass manufacturers
There are many technical guidelines presented by different manufacturers (AGC, Pilkington,
Saint-Gobain, etc.) based specifically on experiments on their products (App. A, Fig. A.6.13).
t by
It can be helpful in preliminary architectural design or to make pre-calculations on costs.

In the case of special glazing applications, it is preferred to perform engineering calculations
on the glazing element by so-called glass engineers, sometimes also supplemented by
laboratory testing.
h
6.4 Conclusions
yrig
Designers are directed by different standards and guidelines. Where safety can not be
guarantied only with calculations on glazing elements, tests should be done. The structural
behaviour of laminated safety glass lies between two limits, the so-called layered limit, in
which the glass panes react without shear bond and the monolithic limit in which all glass
cop
panes are rigidly connected. For both limits, stresses in the glass pane can be calculated using
the known formulas. In reality, the maximum stress of laminated glass lies between those two
limits.
With increasing temperature, the bond capacity of the laminated glass decreases and creep
becomes more significant. To calculate the bending strength of a single or laminated glass
pane, different influencing factors should be taken into account.
Table A.6.3 indicates that different methods can be applied to calculate the bending
characteristics of laminated glasses. Each method has advantages and also disadvantages.
With some methods we cannot distinguish between float and tempered glass, or the calculated
strength is for the whole glass overall. To perform engineering calculations also supplemented
by laboratory testing, influencing factors e.g. effect of tempering and properties of interlayer
materials as well as temperature dependent behaviour should be studied.
27
Experimental programme
Chapter 7
7. Experimental programme
7.1 Introduction
Influencing factors mentioned in the previous Chapters 3 to 6 affect the strength of glass.
Nowadays glass is often used as a construction material to create load bearing structures
where the bending strength of the material plays an important role in the load bearing
capacity. An experimental programme with numerous single and laminated glass specimens
DT
was carried out. For safety glazing applications, heat–strengthened or heat–tempered glass
should be used. Only the lamination of glass layers do not make it safe. Therefore, the layers
of laminated safety glass should consist of heat–strengthened or heat–tempered glass layers
(Chapter 3). Tempering increases the price of the glass (about 1.5 times), therefore,
AR
sometimes people try to use ordinary float glass (Fig B.7.1) where the use of heat–
strengthened or tempered glass would be necessary. Float glass for point fixed glazing is
especially dangerous, because it is not resistant to high stress concentrations e.g. which
KH
develop around holes (Pankhardt, 2004).
The difference of load bearing capacities between the laminate manufactured from
ordinary float (non safety laminated glass) or tempered glass layers (safety laminated glass)
was also investigated. When glass laminate fractures, the interlayer can keep the pieces in
AN
place. Different types of interlayer materials, both resin and foil (type EVA) were studied in
safety and non safety laminated glass specimens. The main influencing factors on the load
bearing capacity and bending characteristics of the glass specimens were investigated (e.g.
K.P
tempering, temperature).
7.2 Test parameters and test programme
7.2.1 Test parameters
The aims of this Thesis were to study the load bearing capacity of single and laminated glass
t by
specimens in bending tests with different kinds of influencing factors.

Test parameters of single glass specimens were the following:
Constants: test arrangement, width and length of specimens, edgework.
Variables: thickness, type of glass [non heat-treated (float) or tempered], rate of loading,
h
temperature of specimens.
Test parameters of laminated glass specimens were the following:
yrig
Constants: test arrangement, width and length of specimens, thickness of the glass layers
was 6 mm, edgework.
Variables: number of layers (two or three), type of glass (non-safety or safety laminate),
type of interlayer material (resin or EVA foil or without interlayer),
cop
temperature of specimens.
Tests were based on glass panes with a thickness of 6 mm. Therefore, the bending
characteristics in the case of relative large deflections could also be studied. Single glass
specimens with thickness of 12 and 19 mm were also investigated to compare the results of
2×6 mm and of 3×6 mm laminated glass specimens with a possible monolithic upper layered
limit for them. Glass layers laminated only with the use of a spacer (without interlayer
material) were tested to experimentally determine the lower layered limit of 2×6 mm and of
3×6 mm laminated glass specimens.
The glass layers without interlayer material were layered only with an elastic strip with a
width of 5 mm and thickness of 1 mm at the edges as a spacer.
28
7.2.2 Test programme

The glass specimens with constant length of 1100 mm and width of 360 mm were tested in
four-point bending. The size of tested specimens was prepared according to EN 1288-3:2000
standard. Non heat-treated (float) and tempered single layer glass specimens of different
thicknesses (6 mm, 12 mm and 19 mm) and laminated (2×6 mm, 3×6 mm) glass specimens
were tested. Laminated glasses consisting of different types (non heat-treated float or
tempered) and number (two or three) of glass layers with equal thickness (of 6 mm) were
investigated. The applied interlayer materials were resin or EVA foil. The experiments were
carried out at a room temperature of +23°C. Further specimens were cooled to -20 °C or
DT
heated to +60 °C. The schematic diagram of the test programme for single glass specimens is
illustrated in Fig. 7.1.
Test programme of single glass specimens
AR
-20 C° +23 C° +60 C°
KH
Non heat- Tempered Non heat- Tempered Non heat- Tempered
treated (float) treated (float) treated (float)
6 mm
6 mm 6 mm
AN
12 mm
19 mm
K.P
Fig. 7.1 Schematic diagram of test programme for single glass specimens
The schematic diagram of the test programme for laminated glass specimens is illustrated in
Fig. 7.2.
Test programme of laminated specimens with two or three glass layers
t by
with two glass layers with three glass layers

h
-20 C° +23 C° +60 C°

yrig
Non heat- Tempered Non heat- Tempered Non heat- Tempered

treated (float) treated (float) treated (float)
cop
No bond
Resin bonded
Resin bonded Resin bonded
EVA bonded
EVA bonded EVA bonded
Fig. 7.2 Schematic diagram of test programme for laminated specimens

The test programme is described in more detail in Appendix B, Tables B.1 and B.2. The
surfaces of the glass layers in laminated glass panes were bonded with different kinds of
interlayer materials as resin (cast in place resin) or EVA (ethyl-vinyl-acetate) foil. The
influence of temperature on the load bearing capacity of single glass and the laminated glass
panes were studied. Simplified symbols were used to distinguish the studied specimens, these
29
are summarised in Table B.7.1. Tables B.7.2 and B.7.3 of Appendix B summarise the testing
programme.
7.3 Method of testing (four-point bending)
7.3.1 Test specimens
Specimens were manufactured from soda lime silicate float glass with polished edges. Any
intended changes to the condition of the test specimens, like edge working, was completed at
least 24 hours before testing (EN 1288-1:2000). Specimens were stored in the test
environment for minimum one day before being tested (Figs. B.7.2 and B.7.3). If the glass
DT
surface is modified by abrasion, etching, edge working, etc., it is necessary to allow the fresh
damage to heal before the test is done. The continuous surface modification by moisture
affects the damage in a way that can reduce any weakening effect (Wiederhorn, 1967 and
Chapter 5).
AR
The width, length and thickness of specimens were determined as the average of at least
three individual measurements. Thickness was determined on the selected points of the glass
pane with accuracy of 0.01 mm (Fig. B.7.4). The average was taken from all these measured
KH
values for every specimen (Table B.7.4).
For single layer glasses by each testing combination four pieces of specimens, for laminated
glass by each testing combination three pieces of specimens were tested. The required number
of specimens for any combination of parameters was determined according to the standards.
AN
The standard deviation of the test results was at most 10% of the average of measured values.
In total, 160 specimens were tested.
7.3.2 Experimental set up
K.P
7.3.2.1 Force measurement

All glass specimens with a constant span of Ls = 1000 mm and supported at a width of
b = 360 mm were tested in four-point bending (Figs. B.7.5 and B.7.6). The load and deflection
at mid-span of the glass panes were measured in all tests. The test procedure is a semi-
t by
dynamic short-term test. The tests were carried out at a specimen temperature of
+23 °C, +60 °C and -20 °C, respectively (Figs. B.7.7 to B.7.10). The heated or cooled
specimens were insulated to provide an environment with constant temperature and humidity,
Fig. 7.4. The temperature of the specimens and the room temperature were continuously
h
measured during the tests (Fig. B.7.8).

yrig
F
5. 7.
6. 1. 2.
4.
Lb
cop
h
3. Ls
Fig. 7.3 Method of test for four-point bending (EN 1288-3:2000) where, 1.: specimen:
1100×360 mm, 2.: bending roller, 3.: supporting roller, 4.: rubber stripe (3 mm thick,
according to ISO 48:1994), 5.: self-designed transducer, 6.: custom-made insulation (40 mm
thick), Ls: 1000 mm, Lb: 200 mm, h: thickness of the specimen (6 mm, 12 mm, 19 mm or 2×6,
3×6 mm), 7.:self-designed load transfer beam, F: loading force.
The specimens were mounted as shown in Fig. 7.3. Rubber stripe of 3 mm thickness and a
hardness of 40 ± 10 IRHD (in accordance with ISO 48:1994) were placed between the
specimen and the bending and supporting rollers to avoid hard contacts. The bending tests
30
were carried out at 23 ± 5 °C room temperature with relative humidity between 50 % and
65 %. The temperature was kept constant during the test with ± 1°C in order to avoid the
development of thermal stresses. A digital instrument was used during the test to measure the
temperature of the testing room and the relative humidity.
Load was measured with a self-designed force transducer, developed by the author
(Pankhardt, 2004) for an Intron Type 1197 testing instrument and calibrated with a Hottinger
Baldwin Messtechnik (HBM) 200 kN force transducer (No. 76411). A self-designed steel
construction with hinge connections was constructed to transfer the load to the specimens
with four-point bending. Displacement was measured with an HBM type W50 displacement
DT
transducer at the mid of span. Signals of the instruments were transformed with Catman
software to the measured values. Values measured during the tests were simultaneously
recorded by computer. Time was measured by the data-acquisition computer. Modes of
deflection and fracture process as well as the crack pattern of the glass specimens were
AR
recorded with digital optical methods (CMOS SONY Camera, Fig. B.7.11).
7.3.2.2 Rate of loading
The specimen should be bent (EN 1288-3:2000) with a uniformly increasing bending stress at
KH
a rate of (260.4) N/(mm2⋅s) until failure. Therefore, the testing instruments were calibrated to
obtain the displacement increment for single and laminated specimens. The results of
calibrations were a bending stress rate of 260.4 N/(mm2⋅s) in the glass specimens:
AN
- in the case of single glass specimens with thickness of 6 mm: 50 mm/min displacement
(available by Instron Type 1197) should be applied (Table B.7.5).
- in the case of laminated glass specimens: 20 mm/min displacement (available by Instron
Type 1197) should be applied.
K.P
- in the case of single glass specimens with thicknesses of 12 mm or 19 mm: 20 mm/min

displacement should be applied to compare the test results with laminated glass
specimens.
To compare the results at same displacement increment between single glass specimens and
laminated glass specimens, the tests at room temperature were extended as follows:
t by
- 50 mm/min was applied to: single glass specimens with thickness of 6 mm and in the case
of spacer laminated specimens.
- 20 mm/min was applied to: single glass specimens with thicknesses of 6 mm, 12 mm,
19 mm; in the case of spacer laminated specimens and in the case of all laminated
h
specimens.
yrig
Further tests were carried out with specimens thickness of 6 mm, with a decrease of the
loading rate from 20 mm/min to 5 mm/min and 1 mm/min.
The measured actual force and displacement (deflection) data were displayed in real time on
the screen of the data-acquisition computer (Fig. B.7.12). The specimens were tested until
cop
fracture (Fig. B.7.13). Laminated specimens were loaded until all glass layers were fractured.
7.3.2.3 Strain measurement
Strains at selected points on the surface (in Region 1 and Region 2) of the glass panes were
measured with type HBM LY11-10/120 strain gauges (Fig. B.7.13). For the calibration phase
of the tests, type Kyowa KFC-5-D17-11 strain rosettes were also used.
In the case of laminated glass specimens, type HBM LY11-10/120 strain gauges were
placed both at the bottom and at the upper surface of the specimens to measure the strain in
the middle and at the edge regions of the glass pane. For temperature compensation, another
glass specimen had strain gauges applied on its surface and was stored in the same condition
as the tested specimens. Before testing, the compensator specimen was connected with the
tested specimen in half-bridge. The change of resistance in mV/V of the gauges was
31
transferred to the digital channels of a HBM Spider8 instrument. The Catman software which
was installed on the computer is able to transform the measured mV/V data into µm/m after
calibration. Stresses at that same point may be calculated using Hooke’s law for linear elastic
materials, provided its elastic constants are known. If the stress, σz, is normal to the surface of
the body (at the point at which the strains are measured), is zero or no normal loads are
applied to the surface at that point, then Eqs. (7.1) and (7.2) can be used.
σx σy
εx = −ν ⋅ (7.1)
E E
σx σy
ε y = −ν ⋅ +
DT
(7.2)
E E
7.3.3 Scanning electron microscopic analysis (SEM)
AR
To study morphologically the edge region of single glasses, four different types of edges were
prepared:
a) manually arrised edge,
b) machine ground edge,
KH
c) machine ground + acid etched edge,
d) machine polished edge.
A type JEOL JSM-5500LV scanning electron microscope was used. The edge samples were
covered with Au-Pd vapour for electron microscopy. The parameters of electron microscopy
AN
were the following: high vacuum mode, secondary electron (SE) detector, acceleration
voltage 25 kV. Digital photos were taken with magnification of ×50, ×100, ×300, ×1000. In
the photos, the scaling line is also indicated. Analysis of the photos is discussed in Chapter 8.
K.P
7.3.4 Differential scanning calorimetric analysis (DSC) of cured resin

The glass transition temperature and melting temperature ranges were not available for cured
resin, which was used in the laminated glasses. These data were only available for EVA foil.
In order to determine the glass transition and melting temperature ranges, tests were carried
t by
out. Appendix C summarises the properties and test results of interlayer materials. Type of
used resin was: UP-resin based on ortho-phthalic acid, pre-accelerated, light stabilised
(product name: VIAPAL VUP 4808 B/62, manufacturer: CYTEC). Liquid resin and activator
were mixed to create specimens for DSC analysis (Figs. C.2.1 to 2.2). Weight of the liquid
h
resin (styrene content 61.6%) was 50.31 g. Weight of activator (ketoneperoxid) was 0.87 g (it
is equal to 1.73 m% of the resin). Measured gel time was 32 min. at 25 °C room temperature.
yrig
Gel time given by the manufacturer was 35 to 45 min. at 20 °C (in the case of activator 2 m%
of the resin).
DSC tests started at least 24 hours after mixing (end of cure time). The curing process was
exothermal (increase of temperature was about 25 °C after two hours of mixing the
cop
components). The cured resin was highly flexible at room temperature. The dynamic DSC
thermographs of the cured resin were available with the use of Perkin Elmer DSC 7
equipment. A 5 mg sample was tested between temperatures from -60 to 220 °C in a
nitrogenous area with a flow rate of 40 ml/min. The rate of heating was 10 °C/min.
The result of the first heating is indicated in Fig. C.2.3. Heat flow vs. temperature diagrams
(Fig. C.2.4 to 2.5) of the cured resin indicate that the glass transition temperature, Tg, is about
-38 °C, and melting temperature, Tm, is +109 °C (80 °C to 140 °C). Further results are
discussed in Chapter 11.
7.3.5 Physical properties of interlayer used during test programme
The mechanical behaviour of the laminated glass specimens is influenced by the interlayer
material. The studied interlayer materials were cast in place resin and EVA foil. The main
32
properties of the tested products are summarised in Table 7.1. Further informative tensile tests
were carried out with resin and EVA interlayer materials. Appendix C summarises the
properties and informative test results of the tensile tested interlayer materials.
Table 7.1. Physical properties of cured resin and cured EVA foil interlayer (data of Technical
datasheets and the Author’s measurements)
Resin EVA foil
Properties Unit
(UP) (ethylene vinyl acetate)
Specific gravity g/cm3 1.16 0.95
Thickness mm 1 0.4
DT
Tensile strength at 23±2 °C N/mm2 2 14 to 22
Ultimate strain at 23±2 °C % (~160) > 60 > 465
Young’s modulus at 23±2 °C N/mm2 ~ 12 > 1.2
AR
Poisson ratio at 23±2 °C - ~ 0.40 0.32
~ -28 (depending on the
Glass transition temperature °C -42 to -38
content of softeners)
Melting temperature range °C 109 (80 to 140) ~ 76 to 79
KH
Water absorption (°C×24Hours) % - <0.01
Styrene monomer
Other Gel content 95%
content 61.6 m %
Further properties of EVA foil interlayer given by the manufacturer are available in App. C,
AN
Table C.1.1 and Fig. C.1.1.
7.4 Conclusions
K.P
The load bearing capacity of single layer or laminated specimens tested in bending is
influenced by many factors, mentioned in previous Chapters. A testing programme was
determined to investigate the main influencing factors, such as temperature on the bending
characteristics e.g. load bearing capacity and deflection behaviour as well as strains on the
glass surface. In the case of heat-treated glass layers and appropriate interlayer, the laminated
t by
glass can be called laminated safety glass (LSG). Laminated glasses were manufactured with
non heat-treated glass layers or with tempered glass layers. To determine the lower layered
limit and the upper layered limit of tested laminated glasses, specimens without bonded glass
layers (non bonded glass layers) and single layer monolithic specimens with comparable
h
thickness were manufactured.

yrig
cop
33
Effect of tempering
Chapter 8
8. Effect of tempering on load bearing capacity of single
glass pane in bending
8.1 Introduction
An experimental programme was carried out to analyse the effect of tempering on the load
bearing capacity of glasses manufactured with nominal thicknesses of 6, 12 or 19 mm.
Specimens were tested in four-point bending. In the case of tempered glass, the stresses must
DT
first exceed the built-in compression stresses before tension develops. The influence of edge
strength (in Region 2, see Fig B.7.13) also influences the strength of glass products, therefore,
the strains near the edges were also studied. The results of tempered glass specimens were
compared with non heat-treated float glass specimens. This chapter deals with the effect of
AR
glass thickness, of loading rate and of edgework on the bending characteristics of single layer
glasses.
8.2 Experimental results and test observations
KH
Each specimen was tested until fracture, therefore, the maximal force and deformation as well
as maximal strain at selected points of the glass surface can be determined. A significant
difference can be observed between the fracture pattern of non heat-treated and tempered
AN
glass specimens (Fig. B.8.1 and B.8.2). While float glass fractured into shards, the single
tempered specimens fractured dynamically into small pieces with blunt edges (Fig. B.8.3).
8.2.1 Force vs. mid-point deflection relationship
K.P
The load bearing capacity of a single layer tempered glass of 6 mm is about 3.5 times that of
non heat-treated single float glass at room temperature (Fig. 8.1 and App. B, Fig. B.8.4).
16
63.34; 1.62 24.73; 15.52 E_23°C_19mm
1.6 tempered_50mm/min +23 °C 61.63; 1.58
+23 °C
14 E_23°C_12mm
1.4 tempered_20mm/min 50 mm/min 20 mm/min
float_50mm/min 20 mm/min 12 E_23°C_6mm
t by
1.2
Force, F (kN)
float_20mm/min
Force, F (kN)
10 F_23°C_19mm
1 F=0.628y
46.15; 7.94 F_23°C_12mm
0.8 8
11.87; 7.21 F=0.172y F_23°C_6mm
0.6 6
16.20; 0.43 18.76; 0.47 Lineáris
0.4 4 F=0.153y
F=0.026y
h
0.2 17.41; 2.66 61.63; 1.58

2
16.20; 0 18.76; 0 61.63; 0 63.34; 0
0 0; 0 16.20; 0.43
yrig
0 0; 0
0 10 20 30 40 50 60 70
0 5 10 15 20 25 30 35 40 45 50 55 60 65
Deflection, y (mm)
Deflection, y (mm)
Fig. 8.1 Force vs. deflection diagram of Fig. 8.2 Force vs. deflection diagram (avg.)
6 mm thick float glass specimens, average of non heat-treated float (F_) and tempered
value of tested specimens at +23°C. (E_) single layer 6 mm, 12 mm or 19 mm
cop
thick glass specimens tested at +23°C.

In most of the references (B&G glass, 2009; Glassguides, 2009), the load bearing capacity in
the case of tempered single glasses is three to four times higher than in the case of float
glasses. A question arises: is the load bearing capacity of tempered glass always three to four
times higher in the case of different glass thicknesses or in the case of differently applied
loading rates?
Based on these laboratory four-point bending tests, the effectiveness of tempering
decreases with the increase of glass thickness at a loading rate of 20 mm/min. The
relationship between effectiveness of tempering and glass thickness is linear (Fig. 8.3).
Fig. 8.1 and Fig. B.8.4 indicate the influence of loading rate on the ultimate force. Decreasing
the loading rate from 50 to 20 mm/min produced decrease of ultimate force by 8.5 % for float
34
Effect of tempering
glass and 2.5 % for tempered glass. Fig. 8.2 indicates practically the same Young’s modulus
independently of the heat treatment of glass of the same thickness tested in bending.
I have introduced the definition effectiveness of tempering (heat treatment) in the case of
tempered glasses with different thickness to show the effectiveness of tempering on the
increase of load bearing capacity. The effectiveness of tempering shows the proportion
of load bearing properties (e.g. maximal force) of tempered glasses compared to non
heat-treated float glasses with the same thickness. It was experimentally shown that the
effectiveness of tempering depends on the glass thickness and the loading rate
(Pankhardt & Balázs, 2006; Pankhardt & Balázs, 2010a) [1.1 New Scientific Result: The
effectiveness of tempering].
DT
The ratio of ultimate force of tempered glass to that of float glass is 3.67 for 6 mm thickness,
2.98 for 12 mm thickness and 2.15 for 19 mm (Fig. 8.3).
4.0 5.0
3.67 +23 °C 20 mm/min +23 °C
AR
Ratio of ultimate Force, Fu of
3.5 4.5
tempered to float glass (-)

Ratio of ultimate Force, Fu of
2.98 h = 6 mm
tempered to float glass (-)
3.0 4.0 3.67 3.45

2.5 2.15 3.5 3.01
ratio=-0.12h + 4.38
2.0 3.0 50 mm/min
KH
1.5 2.5 2.75 20 mm/min
0.070
1.0 2.0 ratio = 2.757x 5 mm/min
R 2 = 0.821
0.5 1.5 1 mm/min
0.0 1.0
4 6 8 10 12 14 16 18 20 0 5 10 15 20 25 30 35 40 45 50 55 60
AN
thickness, h (mm) Loading rate, (mm/min)
Fig. 8.3 Effectiveness of tempering versus Fig. 8.4 Effectiveness of tempering versus
glass thickness in the case of bending rate of loading with 6 mm glass thickness
K.P
(20 mm/min)
Based on my experimental results I have shown that the effectiveness of tempering decreases
with the increase of glass thickness (Fig. 8.3). Thus, the Bažant type size effect (Bažant,
2004) influences also the effectiveness of tempering. The effectiveness of tempering
decreases 41.1% in the case of 19 mm glasses and 18.8% in the case of 12 mm glass thickness
t by
compared to glasses of 6 mm thickness. It was also shown (Fig. 8.4) that the relationship
between effectiveness of tempering and glass thickness is linear in the nominal thickness
range of 6 and 19 mm at a loading rate of 20 mm/min.
For glasses with nominal thickness of 6 mm it was shown that the effectiveness of tempering
h
decreases with the reduction of the loading rate from 20 mm/min to 1 mm/min (Fig. 8.4), but
yrig
no significant changes with increase of the loading rate from 20 mm/min to 50 mm/min. The
explanation for that is more time for the development of cracks starting from surface scratches
in the case of reduced (5 mm/min, 1 mm/min) loading rates.
Fig. B.8.5 (in App. B) indicates the influence of thickness and tempering on the integral of
cop
force vs. deflection (external work) diagram. The external work increased linearly with
thickness in the case of float glass. However, there was no significant increase of the external
work above 12 mm thickness of tempered glass. The reason for that is the difference of
prestressing during tempering.
With the effectiveness of tempering (heat treatment) there is a possibility to select the
appropriate and economic glass thickness.
8.2.2 Stress-strain relationship
8.2.2.1 Strain in mid-region (Region 1) of pane
Decreasing the loading rate from 50 to 20 mm/min produced decreases of ultimate force by
8.5 % for float glass, while the ultimate strain had no significant decrease. Fig. B.8.6 indicates
the influence of loading rate on strains of single layer glasses with thickness of 6 mm. In the
35
Effect of tempering
case of a loading rate of 50 mm/min, the maximal strains of the edge region (Region 2) are
about 10 % higher than at mid-pane (Region 1) both for float and tempered glasses.
Decreasing the loading rate from 50 to 20 mm/min produced an increase of strains in the edge
region from 10 % up to 16 % for float, and up to 12 % for tempered glasses. The condition of
the glass surface is influenced by the statistical distribution and types of cracks, etc. (see
Chapter 5), which also affects the bending characteristics of glasses. Young’s modulus E =
70000 N/mm2 should be applied to calculate the bending strength of a single glass pane.
8.2.2.2 Strain in edge region (Region 2) of pane
Strains were measured at the bottom (tensioned) surface of the specimens. Fig. 8.5 indicates
DT
the force vs. strain diagram of glass specimens with thicknesses of 6, 12 or 19 mm tested with
a loading rate of 20 mm/min. The tempering influenced also the force vs. strain diagram. In
the case of tempered specimens, the strains are smaller than in the case of float non heat-
treated specimens at the same force level. The stress must first exceed the built-in
AR
compression stresses before tension develops, therefore, the so-called prestressed layers of
tempered specimens help to reduce the strains caused by deflection.
16
tempered_19mm_mid_20mm/min 2075.88; 15.52 2274.99; 15.52
KH
tempered_19mm_edge_20mm/min
14 float_19mm_mid_20mm/min
float_19mm_edge_20mm/min
tempered_12mm_mid_20mm/min +23 °C
12 tempered_12mm_edge_20mm/min
float_12mm_mid_20mm/min 20 mm/min
float_12mm_edge_20mm/min
10
AN
tempered_mid_20mm/min
Force, F (kN)
tempered_edge_20mm/min 2563.46; 7.94 2921.36; 7.94

8 float_mid_20mm/min
float_edge_20mm/min 1125.55; 7.21
1188.82; 7.21
6
K.P
4
1053.20; 2.66
2 1157.63; 2.66 1846.14; 1.58
2063.92; 1.58
590.55; 0.43
508.80; 0.43
0 00;
; ;00
0 250 500 750 1000 1250 1500 1750 2000 2250 2500 2750 3000
Strain, ε ( µ m/m)
t by
Fig. 8.5 Force vs. strain at bottom surface in Region 1 (middle) and Region 2 (edge)
of single tempered and float glasses, thicknesses of 6, 12 or 19 mm tested in four-point
bending with loading rate of 20 mm/min.
h
By illustration of ultimate strain (columns) and ultimate force (dashed lines) as functions of
thickness in one diagram, (Fig. B.8.7) indicates that increase of ultimate force produced no
yrig
further increase of ultimate strain with an increase in the thickness from 12 to 19 mm of

tempered single layer specimens. The ultimate strain is also influenced by the rate of
tempering (tempering process).
1500
cop
-1,82 +23 °C Float_mid

y = 37308.25h
2
1250 R = 0.995 20 mm/min Float_edge
Ratio of maximal strain and ultimate force
-1,89 Tempered_mid
1000 y = 39247.90h
2
R = 0.999 Tempered_edge
750
(µm/m/kN)
-1.88
y = 33977.91h
500 2
R = 1.00
-1,74
250 y = 27757.14h
2
R = 0.995
0
4 6 8 10 12 14 16 18 20
thickness, h (mm)
Fig. 8.6 Ratio of maximal strain and ultimate force at bottom surface in Region 1 (middle)
and Region 2 (edge) of single tempered and float glasses vs. thicknesses of 6, 12 or 19 mm
36
Effect of tempering
Fig. 8.6 indicates that the ratio of maximal strain and ultimate force as function of thickness
decreases with an increase of thickness for both float and tempered specimens calculated in
the Region 1 and Region 2, respectively. This ratio also indicates the effectiveness of
tempering in the case of single glass specimens.
8.2.2.3 Stresses in mid-region (Region 1) of pane
Stresses at selected points on a glass surface can be determined from strain measurements.
Surface stresses on the bottom surface of single non heat-treated float (F) and tempered (E)
glass specimens have been calculated based on Hooke’s law using the theoretical Young’s
modulus of glass E=70000 N/mm2. Bending strength in the mid-pane, σbB, (in Region 1) and
DT
in the edge, σbB,edge, (in Region 2) have been calculated with Eq. (6.11) and are indicated in
Table 8.1.
Table 8.1 Bending strength and surface strength (average for theoretical and for calculated
AR
values) of tested single glass specimens, where letters are: E-tempered glass,
F-non heat-treated glass
Theoretical
Measured values (avg.) surface Calculated bending strength
KH
Specimens
strength, σ
εmax,m
h Middle Edge σbB σbB edge
Fmax, m ymax,m Middle Edge σbG y/h ke
(Region 1) (Region 2) (Region 1) (Region 2)
(Region 1)
AN
(Region 2)
mm kN mm μm/m N/mm2 N/mm2 N/mm2 N/mm2 - - N/mm2
6 1.58 61.63 1846.14 2063.92 129.2 144.5 3.8 157.5 10.5 1.12 176.4
E 12 7.94 46.15 2563.46 2921.36 179.4 204.5 1.9 191.4 3.81 1.14 218.2
K.P
19 15.52 24.73 2075.88 2274.99 145.3 159.2 1.2 145.4 1.29 1.10 159.9
6 0.43 16.20 508.80 590.55 35.62 41.34 3.8 45.6 2.76 1.16 52.9
F 12 2.66 17.41 1053.20 1157.63 73.7 81.0 1.9 65.4 1.42 1.10 71.9
19 7.21 11.87 1125.55 1188.82 78.8 83.2 1.2 68.2 0.62 1.06 72.3
Table 8.1 gives the surface stresses at maximal force which are different from the calculated
t by
bending strengths of single glass specimens with thicknesses of 6, 12 and 19 mm. In the case
of thinner than 10 mm and non heat-treated glasses, the surface stress calculated with the
deformation of glass surface (outer layer) will be considerable. By tempered glasses, with the
increase of the thickness (above 12 mm nominal thickness), the strength values calculated
h
with the basic formulas of the theory of elasticity fit the real values measured on the surface
better than in the case of thin glasses (Fig. 8.7).
yrig
Based on my laboratory bending tests I have shown that the values are different for
bending strength calculated with the basic formula of elasticity given in the standard EN
1288:3-2004 and the maximal value of surface stresses of glass calculated from the result
of strain measurement. With the increase of the glass thickness, the bending strength
cop
decreases compared to the surface strength. The ratio of bending strength to surface
strength determined with strain measurement can be approximated with power
functions both in the case of tempered and non heat-treated float glasses (Pankhardt,
2008a; Pankhardt, 2008b; Pankhardt & Balázs, 2010a) [1.2 New Scientific Result: Bending
strength of single layer glasses].
Fig. 8.7 indicates that the ratio of bending strength to surface strength decreases with increase
of thickness (see also Fig. 8.6). Based on my experimental results, the ratio of surface
strengths determined with strain measurement to the calculated value of bending strength can
be fitted with power functions with the best correlation both in the mid of the pane (R 1
region) and in the edge region (R 2 region).
Fitted curves in Fig. 8.7 indicate the size effect on the ratio of bending strength to maximal
surface stresses of glass panes. The bending strength of a 6 mm thick tempered glass
37
Effect of tempering
specimen is 20% higher than its surface strength. Surface strength results in earlier failure
than bending strength, especially in the case of thin (h < 10 mm) glass panes. In addition to
bending stresses, membrane stresses can develop in thin plates undergoing relatively large
deflections (Hess, 1986; Vallabhan et al., 1987). Simply supported transverse loaded thick
plates are less affected by this geometrical nonlinearity then at all edges supported thin glass
panes. In the case of float non heat-treated and relatively thick (h > 10 mm) specimens, the
strength is considerably influenced by the size effect as mentioned before. The bending
strength of a 19 mm thick float glass specimen is 14% lower than its surface strength.
140%
tempered_Mid
Bending strength / Surface strength
+23 °C
DT
120% tempered_Edge
20 mm/min
float_Mid
100%
float_Edge
σ bB/σ (%)
y = 2.34h -0.35
AR
80% 2
R = 0.88 y = 1.66h -0.17
60% y = 2.08h -0.32
2
R = 0.99 y = 1.62h -0.18
2
2
R = 0.87 R = 0.98
40%
KH
20%
0%
0 5 10 15 20
AN
Thickness, h (mm)
Fig. 8.7 Ratio of calculated bending strength to surface strength in percent versus thickness
of single glass specimens
K.P
The slope of the fitted curves is different for non heat-treated float and tempered glasses (Fig.
8.7, Fig. B.8.8). The slope of curves for tempered glasses (~1:6) is less steep compared to that
of float glasses (~1:3) (Pankhardt, 2008b) as a function of thickness, thus the size effect is
less dominated. In the case of tempered glass, the resulting compressed layer helps to close
cracks initiated on the tensile surface (surface defects e.g. scratches), which can be more
t by
difficult to open and propagate with further loading. The bending strength of tempered glass is
more influenced by the distribution of defects e.g. surface defects, but in the case of float
glass it is influenced by a critical defect. The bending strength of tempered glasses as a
function of thickness is influenced by the Weibull type statistical size effect (Weibull, 1951),
h
but the size effect of float glasses is influenced by linear elastic fracture mechanics and LEFM
provides a good model for these types of glasses (Figs. B.8.9 and B.8.10).
yrig
The force vs. deflection curves indicate linear elastic behaviour until reaching the ultimate
force. Young’s modulus, Eflexural, can be evaluated with the use of Eq. (6.13), which uses the
parameters of ultimate force and ultimate deflection known from the tests. Table B.8.1
indicates the calculated values for specimens with thicknesses of 6 mm,
cop
12 mm and 19 mm. The recommended Young’s modulus is E=70000 N/mm2 for soda lime
silicate glasses. Therefore, the surface strength is calculated with the recommended value, see
Table 8.1 and Table B.8.1. The surface strength is also calculated with E = Eflexural. By
comparing the results of strengths, Eflexural, is higher than the recommended value for both
float and tempered glass with a thickness of 6 mm. Therefore, in the case of specimens with
large deflections (Fig. B.8.11) the calculated strengths (both surface and bending) are
overestimated for thin (6 mm) simply supported specimens with the use of Eq. (6.11) (see
also Pankhardt, Balázs, 2006). In the case of single glass thicker than 12 mm or with
appropriate bonded laminated glass layers (Chapter 10), Eq. (6.13) can be applied.
The surface stresses are influenced more by the surface condition of the glass element
than the bending strength. Impacts on the glass surface by hard, sharp particles e.g. scratching
38
Effect of tempering
are the initial locations of cracks, which can develop and quickly propagate. When the surface
stresses will reach the surface strength of the pane, fracture occurs. The surface strength of
tempered glass is influenced by the tempering process. The maximal compressive stress
produced by tempering in the outer surface layer can be estimated from Table 8.1 and is
indicated by the difference between surface strength of tempered and float glass with the same
thickness. The prestressing of tempered glass with thickness of 6 mm is about 103 N/mm2, of
12 mm, about 123 N/mm2, of 19 mm thickness, about 76 N/mm2. Usually, it should be at least
about 67 N/mm2 (see Table A.3.3).
8.2.2.4 Stresses in edge region (Region 2) of pane
DT
The edge region of glass contains more defects caused by edgework. When glass fractures, it
fails practically at the edges first and the crack propagates in the direction of the middle of the
pane (see Figs. B.8.1, B.8.12). Because of the high (v ≈1480 m/s) crack propagation velocity
in glass (Denoyer et al., 1963, see also Chapter), the fracture process can be observed only
AR
with high-speed cameras. Measurement with strain gauges (Fig. B.8.1) indicated that the
highest stresses were measured in the edge region (Region 2). The theoretical Young’s
modulus of glass E=70000 N/mm2 was applied to determine the surface stress at the selected
KH
points (Region 1 and 2) for lower and upper (loaded) surfaces of single glass specimens. Figs.
in App. B, B.8.13 to B.8.15 illustrate the stress distribution over thickness of glass specimens
at maximal force and ultimate strain and also indicate that edge stresses are in all cases higher
than mid-pane stresses.
AN
120%
y = -0.008h + 1.204 Float_surface
+23 °C 2
118% R = 0.979 Float_bending
20 mm/min
Ratio of edge strength (R2) to mid-pane
2
116% y = -0.001h + 0.017h + 1.042
K.P
Tempered_surface
2
R = 1.000
114% Tempered_bending
112%
strength (R1)
110%
108%
t by
106% y = -0.004h + 1.125

2
R = 0.999 2
y = -0.0005h + 0.010h + 1.049
104% 2
R = 1.000
102%
h
100%
yrig
4 6 8 10 12 14 16 18 20
thickness, h (mm)
Fig. 8.8 Strength ratio for Region 2 (edge) to Region 1 (middle) of float and tempered glass
specimens at bottom surface, calculated with Hooke’s law or with Eq.(6.2), (see Table 8.1)
Fig. 8.8 indicates the strength ratio for Region 2 (edge) to Region 1 (middle) both for float
cop
and tempered glass specimens as a function of thickness. Strengths were calculated with
Hooke’s law (symbol_surface in Fig.8.8) and with Eq. (6.2) (symbol _bending in Fig.8.8).
The fitted curves in the case of float glass specimens are linear and the ratio decreases with an
increase of thickness. In the case of tempered specimens, the fitted curves are polynomial,
where the ratio increases from 6 mm to 12 mm thickness and decreases from 12 mm to
19 mm thickness. In the case of surface stresses, the ratio is higher than that of bending
strength. In the case of float glass specimens, the edge strength is influenced more by
thickness and edge condition than in the case of tempered specimens. Reaching the ultimate
strain in the edge region led to fracture, therefore, the effect of the edge quality is important
on load bearing capacity and durability of glass (Pankhardt, 2010a). Load bearing capacity of
a glass pane with the same thickness decreases with decrease of edge strength. Although the
39
Effect of tempering
tested specimens were manufactured with machine polished edges, further glasses with
different edgework were investigated with electron-microscopic observation (Fig. 8.9 and
Fig. B.8.16).
DT
a) manually arrised edge, left: ×100; right: ×1000
very coarse abrasive size around 420 to 250 micron, (diamond wheel speed in rpm's 1600 to 2400)
AR
KH
b) machine ground edge, left: ×100; right: ×1000
AN
to remove large rough areas of glass, the process begins with abrasive about 105 micron and
requires further processing down to medium size around 53 to 48 micron (speed of feed from 0.5 to
4 m/min)
K.P
t by
c) machine ground + acid etched edge, left: ×100; right: ×1000,

medium abrasive size around 53 to 48 micron, hydrogen-fluoride acid
h
yrig
cop
d) machine polished edge, left: ×100; right: ×1000

very fine cerium dioxide (CeO2) abrasive, size around 37 to 29 micron
Fig. 8.9 Typical edge finishing by various methods
Fig. 8.9 indicates that the edge region of glass was the most damaged by manually arrised
edgework which can be an initiator of cracks (see also Fig. B.8.12). The roughness of the
edge surface decreases with the use of finer abrasives or with acid etching. Edgework reduces
the initiation of cracks caused by cutting the glass pane, see as cut edge in Figs. 8.10 and
B.8.17. The traditional edge work process (Fig. B.8.18) requires a steady stream of water and
an abrasive compound. Abrasives are available in many different sizes (called grits), ranging
from around mesh size 60 (=250 micron, which is a very rough grit used for initial grinding)
40
Effect of tempering
to around mesh size 600 (= 30 micron, an extremely fine grit). Generally, achieving a highly
polished finish involves using a series of finer and finer abrasives (diamond discs or pad, SiC
silica carbide slurry, cerium dioxide, etc.). A finished ground surface (Fig. B.8.19) will appear
whitish and dull, but a polished surface will shine with no visible scratches (Fig. B.8.20).
DT
Fig. 8.10 Typical edge in “as cut” samples a) ×300 surface defects as initiator of cracks;
b) Surface scratch ×50 with” shell-like” fracture, caused by steel razor, F=0.4 kN
AR
Factor k is used to determine the bending strength of glass where the effects of the edge
conditions are important. For the calculation of stresses in the edge region, it is suggested to
increase the values calculated for the middle of the pane with a multiplying factor. For
calculating the bending strength or equivalent bending strength of the free edges of the glass
KH
pane k=ke shall be used. The appropriate value of ke for use in Eq. 6.11 shall be obtained from
Fig. 6.6, which gives the value of k=ke as a function of the value of y/h. I have complemented
the factor (ke) given in the standard EN 1288-3:2000 and I have determined it for glasses with
AN
thickness of 6 mm as function of ratio of deflection to thickness (y/h) in the case of reduced
loading rates of 20 mm/min (Fig. 8.11).
Based on my laboratory results, I have shown that the value of the multiplying factor
(ke) is influenced by thickness and by the loading rate. The difference of the strains on
K.P
the surface of the glass layer in the edge and mid regions increases with the reduction of
the loading rate from 50 mm/min to 20 mm/min (Pankhardt & Balázs, 2006; Pankhardt,
2008a; Pankhardt, 2008b; Pankhardt & Balázs, 2010a) [ 1.3 New Scientific Result: Ultimate
strains measured on surface].
t by
1.18
h = 6 mm 6mm
1.16 12mm
20 mm/min 19mm
1.14
12_meas_20mm/min
1.12 19_meas_20mm/min
6_meas_20mm/min
1.10 6_meas_50mm/min
h = 6 mm
h
1.08
ke (-)
1.06
50 mm/min
yrig
1.04 h = 12 mm
1.02
h = 19 mm 50 mm/min ≈
1.00
2 ± 0.4 N/mm
2
0.98
EN 1288-3: 2000
0.96
0.94
cop
0 5 10 15 20 25
y/h (-)
Fig. 8.11 Multiplying factor ke as a function of y/h for 6, 12, 19 mm thickness of glass,
continuous lines as given in EN 1288-3:2000, dashed lines are measured values with loading
rate of 20 mm/min, rectangles indicate measured values for 6 mm glass with loading rate of
50 mm/min (Pankhardt, 2010a).
The measured strains increase in Region 2 (R 2) with reduction of the loading rate from
50 mm/min to 20 mm/min because of more time of the nucleation of cracks. Slower rates in
increase of stresses, result lower strengths, as there is a longer time for subcritical crack
growth to occur (see also Chapter 5). The value of ke e.g. in the case of y/h=3 should be
increased by 7% when reducing the loading rate from 50 mm/min to 20 mm/min. Fig. 8.11
41
Effect of tempering
indicates that the value of ke can reach 1.17 as a function of deflection/thickness (y/h) for
glass with thickness of 6 mm. This is also confirmed by the constant reducing factor of 0.8
given in the EN 13474-3:2003 standard, which should be used to calculate the design bending
strength of the edge region based on the strength in the middle of the pane. However, this
standard does not consider the loading rate dependency of the reducing factor.
Although, the measured strains in the edges of a “strip of plate” supported on two lines are
usually higher than in the middle of the pane in four point bending, due to the edge finishing
techniques, the edges of a glass pane contain more microscopic damage or defects than the
middle of the pane. This was confirmed also with my scanning electron microscopic
observations (Figs. 8.9 and B.8.16). Thus, the different regions (mid-pane, edge) of the glass
DT
pane should not be considered as having the same strength.
By edge finishing techniques, the glass edge quality is controlled by: the type of edge
finishing; the type of the tool; the rate of grinding machine; the type and amount of coolant,
AR
therefore friction and heat development. Edgework (if friction is significant) may leave
residual stresses in the edge region of the glass pane. The type of grinding coolants can be
water, or water soluble or non-soluble additives (synthetic-oils). The main functions of them
are the reduction of friction during edge finishing and physical and chemical modification of
KH
edge surface (see Fig. B.8.21). The presence of water (as coolant) supports the formation and
enlargement of cracks (see Chapter 5). The handling of glass can also cause flaws on the
surface (Fig. 8.10 b), which can propagate during the lifetime of the glass. It was shown in
AN
Chapter 5 that the nature of flaws is an important factor. A flaw of given size may respond in
different ways, depending on whether there are residual stresses present in glass or cracks
have been formed which is important in the context of lifetime design.
Further investigations on the effect of glass strength in Region 2 with different type of edge
K.P
work and testing speeds should be done (see Chapter 13). According to my experiments, the
refinement of the relationships indicated in Fig. 8.11 of glass thickness of 12 mm and 19 mm
is possible in the future.
8.3 Conclusions
t by
The definition effectiveness of tempering (heat treatment) was introduced. The conclusions
for the effectiveness of tempering can be summarised as follows: I have experimentally
shown that the effectiveness of tempering depends on the glass thickness and the loading rate.
The higher the glass thickness, the lower the effectiveness of tempering. It is influenced also
h
by the Bažant type size effect. The effectiveness of tempering decreases with the reduction of
the loading rate.
yrig
The conclusions for the relationship of bending strength and surface strength can be
summarised as follows: The bending strength of tempered glasses as a function of thickness is
influenced by the Weibull type statistical size effect, but the size effect of float glasses is
influenced by linear elastic fracture mechanics. The ratio of bending strength to surface
cop
strength can be approximated with power functions.

The conclusions for the maximal strain in various regions and the influence of edgework
on the edge strength can be summarised as follows: Although glass is usually defined as
homogen isotropic material, the studied regions (mid-pane and edge) indicate some
inhomogeneity of the surface, which determines the glass strength. Strength of a glass pane
should be investigated at least in two different regions. Based on my laboratory results, I have
shown that the value of the multiplying factor (ke) – used for calculation of edge strength – is
influenced by thickness and by the loading rate. The difference of the strains in the edge and
mid regions increases with the reduction of the loading rate. Sharp crack tips initiated by the
cutting process can be reduced with the use of finer abrasives or acid etching in the edge
finishing process.
42
Effect of lamination on bending characteristics
Chapter 9
9. Effect of lamination on bending characteristics
9.1 Introduction
This chapter summarises the experimental results on the influence of the number of glass
layers and interlayer materials on bending characteristics (including force vs. deflection,
strain, strength). Tests were carried out both for float (non safety) and tempered (safety)
laminated glass specimens at room temperature, 23 °C, in four-point bending. Two types of
interlayer material were studied, EVA foil and resin. The results were compared with single
DT
glass layer of the same thickness (as upper limit). Laminated glass without the use of
interlayer material was also investigated (as lower limit).
AR
The load bearing capacity of laminated glass is influenced by the properties of the glass layers
and the properties of the interlayer material, see Chapters 2 to 6. When laminated glass
fractures, the interlayer material should bond the fractured number of glass layers together
KH
(Fig. 9.1, App. B, Figs. B.9.1 to B.9.4). In the case of laminated glass manufactured without
interlayer material, the fractures are similar to single layer glass specimens (Fig. 9.2).
AN
K.P
Fig. 9.1 Testing of laminated safety glass Fig. 9.2 Test of float glass layers laid
without bond on top of each other
9.2.1 Stages of fracture process of laminated glass
Fig. 9.3 indicates the fracture process in four-point bending of LG specimens consisting of
three glass layers and a plastic interlayer.
t by
No.3
No.2
No.1
(Stage B3)
h
(Stage A)
yrig
(Stage C2) (Stage B2)

cop
(Stage B1)
(Stage C1)
Fig. 9.3 Schematic force vs. deflection diagram with different stages of laminated safety glass
during the fracture process
Three different stages of the fracture can be identified in the laminate. In the first stage
(Stage A), all glass layers help to carry the load to reach the ultimate (maximal) force of the
laminate, e.g. in the case of laminate consisting of three glass layers, the ultimate (maximal)
force is Fu3 with ultimate deflection y3. After Stage A the glass layer in tension fractures and
the maximal force, Fu3 falls to level, Fu23 (force vs. deflection curve of laminate consisting of
43
n-1=2 number of glass layers). This state is called Stage B. During Stage B the deflection
almost remains at y3 (at room temperature, see also Chapter 10). The force starts to increase
from Fu23 to the ultimate force of the non damaged glass layers, Fu2. This state is called
(Stage C). The interlayer also helps in load bearing during Stage C, the deflection increases
from y23 to y2. Stage A can be followed by further B and C stages sequential (with numbering
B3, B2, B1, etc.), which depends on the number of glass layers in the laminate (see App. B
Fig. B.9.5).
9.2.2 Force vs. mid-point deflection relationship
9.2.2.1 Effect of number of glass layers on load bearing capacity of
DT
laminated glass pane
Generally, first the lower limit (so called layered limit) and the upper (so called monolithic
limit) should be determined to study the load bearing capacity of laminated glass. To
AR
determine the lower limit of the load bearing capacity of a laminate, specimens consisting of
two or three glass layers (single layer thickness of 6 mm) without interlayer material were
tested. Fig. 9.4 illustrates the force vs. deflection diagram of float and tempered laminates
KH
without interlayer material, glass layers were laid on each other without bond.
4.5
E_3_D_50mm/min
E_3_D_20mm/min 3.99 4.16
4
E_2_D_50mm/min
AN
3.5 E_2_D_20mm/min
E_1_50mm/min
3.00 3.13
3 E_1_20mm/min 2.87
F_3_D_50mm/min
Force, F (kN)
2.74
2.5 F_3_D_20mm/min
2.37
K.P
F_2_D_50mm/min 2.26
2 F_2_D_20mm/min
1.68 1.73
F_1_50mm/min 1.46 1.55 1.58
1.5 1.40 1.62
F_1_20mm/min 1.27 1.41 1.58 1.55
1.04 1.29 1.37
1 0.97 1.01
0.88
0.81 0.72 0.71
0.68
0.5 0.53 0.68
0.65
t by
0.43 0.47
0 0
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70
Deflection, y (mm)
Fig. 9.4 Force vs. deflection diagram of non heat-treated and tempered
h
laminated glass with 2×6 mm or 3×6 mm glass layers without interlayer material
The load bearing capacity of laminated glass increases with the increase of the number of
yrig
glass layers even in the case of non-bonded layers, but there is no significant decrease in the
ultimate deflections. The fracture process of tested specimens (both for float and tempered)
happened as illustrated in (Fig. 9.3). The load bearing capacity of float glass specimens is
proportional to tempered specimens. The ultimate force of laminate consisting of two or three
cop
tempered glass layers of 6 mm is about three times that of non heat-treated float glass
laminate at room temperature. The upper limit of load bearing capacity of laminated glass can
be characterised with single layer glass (see Chapter 8) with a thickness of 12 mm for
2×6 mm laminated glass, and with a thickness of 19 mm for 3×6 mm laminated glass (note
that glass with a thickness of 18 mm is not manufactured).
The increase of the number of glass layers in laminate without interlayer material (only
laid on each other) produced an increase of ultimate force nearly linear for both float and
tempered glasses (see App. B, Figs. B.9.6 to B.9.10). In the case of float glass specimens
(without interlayer), the deflection at fracture of the first glass layer did not decrease with the
increase of the number of applied glass layers. The ultimate deflection decreased by about
10% for tempered glass specimens (without bonded glass layers) with the increase of the
44
number of glass layers (for n = n+ 1). The reason is the effect of compressed outer layers in
tempered glass.
Based on my laboratory results with laminated glasses (consisting of glass layers with
equal thickness) I have determined that the effectiveness of tempering is influenced by the
number of glass layers and the property of the applied interlayer material. Fig. 9.5 indicates
the effectiveness of tempering as a function of the number of glass layers and as a function of
the thickness of a single glass specimen.
Thickness, h (mm)
0 2 4 6 8 10 12 14 16 18 20 22 24
Ratio of ultimate force, Fu of safety
3.8
DT
to non-safety laminated glass (-)
3.67 EVA_23°C
3.6
resin_23°C
3.4 +23 °C D_1_2_3
3.2 20 mm/min single:6mm_12mm_19mm
3.08
3.0 2.98
AR
2.89 2.84
2.8 2.85
2.6 2.57
single glass layer
2.4
2.2
KH
2.15
2.0
0 1 2 3 4
Number of glass layers, n (pcs)
AN
Fig. 9.5 Effectiveness of tempering versus number of glass layers in the case of laminated and
glasses laminated only with spacer at temperature of +23 °C. Single layer glasses thicknesses
of 6, 12 as well as 19 mm are also indicated (Pankhardt & Balázs, 2010b).
K.P
Experimentally I have demonstrated that the effectiveness of tempering decreases with

the increase of the number of glass layers for symmetrical layered (consisting of glass
layers with equal thickness) laminated safety glasses, when the layers work together. I
have shown for glasses laminated with a spacer (without interlayer material) that the
effectiveness of tempering no decreases with the increase of the number of glass layers
t by
from two to three. Furthermore, I have shown that at room temperature the
effectiveness of tempering is more favourable in the case of resin laminated glasses than
in the case of EVA foil laminated glasses (Pankhardt, 2008b, Pankhardt & Balázs, 2010b),
[2.1 New Scientific Result: Effectiveness of tempering (heat-treatment) in the case of
h
laminated glasses].
With the increase of thickness of single layer or monolithic glasses, the effectiveness of
yrig
tempering is affected by the Bažant-type size effect (Bažant, 2004).

The effectiveness of tempering is influenced by the position of the interlayer and of the
tempered glass layers in laminated glass. Near the neutral axis, the interlayer did not
contribute as much as in the outermost position in the tensioned layer near the surface of
cop
laminated glass in bending. Thus in the case of laminated glass consisting of two glass layers
with equal thickness, the effectiveness of tempering is less influenced by the interlayer
material than in the case of laminated glass consisting of three glass layers.
Statistically the number of defects increases with the increase of the number of layers. In
the case of non-bonded layers (with spacer), tension develops in each glass layer in bending,
therefore, the condition of tensioned surfaces (containing defects) have a similar influence on
the system. Thus, the increase of number of the tempered glass layers has reduced influence
on the effectiveness of tempering.
In the case of the use of interlayer material with appropriate shear resistance, the effectiveness
of tempering decreases for laminated glasses consisting of three tempered glass layers with
equal thickness compared to laminated glass with two glass layers, because the glass layers
45
near the neutral axis contribute less effectively in bending than the outermost tempered glass
layers.
With rigidly bonded (or monolithic) layers, in bending, the surface defects of the
outermost tensioned glass layers most influence the load bearing capacity of laminated glass.
The higher tensile stress develops in the outer glass layer, therefore, the intermediate glass
layer with a resulting compressed layer – initiated on the tempered glass surface over its
thickness –, cannot effectively contribute to the closing of cracks, thus the effectiveness
decreases (see also Fig 8.6). The decrease of the effectiveness of tempering in the case of
rigidly bonded glass layers can reach about 30 % with the increase of thickness.
The effectiveness of tempering has more influence in the case of no appropriate bonded
DT
glass layers (laminated glasses with resin interlayer material at +23 °C), because tensile stress
develops earlier in each glass layer. Therefore, tempering is about 10 % more effective for
laminated glasses consisting of three glass layers with resin interlayer material, than with
AR
EVA foil laminated glasses.
The effectiveness of tempering provides possibilities for the economical determination of
the composition of laminated glass and the selection of the appropriate interlayer material.
The influence of the effectiveness of tempering increases for example with the decrease in the
KH
bond strength of the glass layers (e.g. due to creep of the interlayer material). In the case of
rigidly bonded layers, heat strengthened glass can also be applied for the inner glass layer(s)
which is convenient not only because of the residual load bearing capacity, but the advantage
AN
of the effectiveness of a tempered glass layer in the inner position can not be taken.
9.2.2.2 Effect of interlayer material on load bearing capacity of laminated
glass pane
K.P
A significant increase of ultimate force can only be reached when an appropriate bond is
ensured between the glass layers. Specimens were tested with EVA foil and resin interlayer
materials. Fig. 9.6 indicates the force vs. deflection diagram of laminated glass (non-safety
and safety) consisting of two glass layers (float or tempered) with resin or EVA foil interlayer
materials.
t by
9
+23 °C E_12mm
7.94
8 E_2_F
20 mm/min
E_2_R
6.64
7 E_2_D
h
6.30 E_6mm
6
F_12mm
Force, F (kN)
yrig
5 F_2_F
F_2_D
4 F_2_R
F_6mm
3
2.66 2.87
2.30
2
cop
2.05 1.58 1.88

1.01 1.78
1.58
1 0.51 1.37
0.87
0.41
0.72 0.85 0.92
0.430.68 0.68
0 00
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80
Deflection, y (mm)
Fig. 9.6 Force vs. deflection diagram of non heat-treated float (F_) and tempered (E_)
laminated glass (with 2×6 mm glass layers) specimens, laminated with resin (_R) or EVA foil
(_F) or without interlayer (_D); also, single glass specimens with thicknesses of 6, 12 as well
as 19 mm are illustrated. Dashed lines illustrate the tempered, continuous lines the float
specimens (Pankhardt, 2008b; Pankhardt, 2010e).
The force vs. deflection behaviour of laminated float glass specimens is proportional to that of
tempered laminated specimens at a loading rate of 20 mm/min. The force vs. deflection
46
behaviour of laminated glass with two glass layers is proportional to that of laminated glass
with three glass layers both for EVA and resin interlayer material (App. B, Figs. B.9.11 and
B.9.12). The fracture process of specimens with non bonded glass layers happened quickly
without significant increase of deformation from y3 to y2 or from y2 to y1 (see also Fig. 9.3). In
the case of specimens laminated with interlayer materials, the ultimate force significant
increases compared to non-bonded glass layers. Fig. 9.7 indicates that laminated glass
specimens can exceed the load bearing capacity of single glass specimens (as monolithic
upper limit of load bearing capacity) with comparable thickness, also mentioned in Chapter 6.
The fracture process indicates the post-failure behaviour of laminated glass. The increase of
deformation at Stages B and C characterise the post-failure behaviour of interlayer materials
DT
(discussed in detail in Chapter 12). Fig. 9.7 indicates the force vs. deflection diagram of
laminated glass (non safety and safety) consisting of three glass layers (float or tempered)
with resin or EVA foil interlayer material.
AR
18
E_19mm
+23 °C 15.52 E_3_F
16
20 mm/min E_3_R
15.01 E_12mm
14
E_3_D
12.77
KH
E_6mm
12 F_19mm
Force, F (kN)
F_3_F
10 F_3_R
7.94 F_12mm
8 7.60
7.21 F_3_D
7.36
F_6mm
AN
6 5.85
5.26 4.93
4.49 3.99
4
3.12 3.00
1.402.66 2.37 1.58
2 2.35 2.26 2.13
2.13 1.04 1.27 0.74 0.85 1.93
0.43 1.16 1.55
K.P
0.54 0.650.65 0.71 0.96 1.29

0 00
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85
Deflection, y (mm)
Fig. 9.7 Force vs. deflection diagram of non heat-treated float (F_) and tempered (E_)
laminated glass (with 3×6 mm glass layers) specimens, laminated with resin (_R) or
t by
EVA foil (_F); also, single glass specimens with thicknesses of 6, 12 as well as 19 mm are
illustrated. Dashed lines illustrate the tempered, continuous lines, the float specimens
(Pankhardt, 2008b).
Both Fig. 9.6 and Fig. 9.7 indicate that in the case of EVA interlayer material, the ultimate
h
force, Fu2 and Fu3 is higher than in the case of resin interlayer material at 23 °C. In the case of
EVA laminated specimens, the deformation to maximal force is lower, therefore, EVA
yrig
interlayer material is preferred in those applications, where the load bearing capacity of
laminated glass is important or the deflection should be limited because of the properties of
sealing or water-tightness.
20.0 40
49.89x Tempered_EVA 37.70
cop
18.0 +23 °C y = 1.51e

Float_EVA +23 °C RESIN
2
35
ultimate force of "Stage A", Fu (kN)
33.75
deflection at maximal force Stage A ,
R = 0.98 Tempered_Resin
16.0 20 mm/min Float_Resin 20 mm/min 33.18
Expon. (Tempered_EVA) 30
14.0 15.01 12.77
26.57 EVA
19.78x
y = 1.53e
12.0 25
R 2 = 0.99
10.0 y = 0.41e 57.71x 20
yu (mm)
2 22.15x
8.0 R = 0.98 y = 0.42e
2
6.64 6.30 R = 0.99 15 RESIN
6.0 5.85 11.59 11.00
4.0 10 11.46 10.22 EVA
4.49
2.0 2.30
1.58 2.05 5 Tempered_EVA Tempered_Resin
0.0 0.43
Float_EVA Float_Resin
0.00 0.02 0.04 0.06 0.08 0.10 0.12 0
total thickness of interlayer/total thickness of laminated glass 1 2 3 4
total h int / total h (-) Number of glass layers in laminated glass, n (pcs)
Fig. 9.8 Ultimate force at Stage A of Fig. 9.9 Deflection at maximal force,
laminated glass specimens Stage A of laminated glass specimens
47
By comparing maximal force of laminated glass specimens vs. ratio of total thickness of
interlayer material to total thickness of specimen, Fig. 9.8 also indicates that load bearing
capacity of laminated glass with EVA interlayer increased more than that of resin laminated
specimens. Fig. 9.9 indicates that the deflection decreases more in the case of EVA interlayer
than in the case of resin interlayer with an increase in the number of glass layers (therefore,
the ratio of total thickness of interlayer material to total thickness of specimen).
For quantitative study of the effect of interlayer material on force vs. deflection curve of
laminated glass, forces (Fig. 9.10 and Fig. 9.11) and deflections (Fig. 9.12 and Fig. 9.13) at
different Stages were calculated (see also App. B, Fig. B.9.5).
DT
6 16
Stage B3-C2 Stage B3-C2
14
5 Stage B2-C1 Stage B2-C1
2.73 12
Stage B1 Stage B1 7.41
Force, F (kN)
Force, F (kN)
4
1.49 10 5.41
AR
3 8
2 6
2.38 2.15 5.47
1.62 0.13 4.76 4.52 5.43
1.18 0.29 4
1 0.56 1.62
2 1.29 0.82
0.68 0.87 0.72 0.74 0.85 0.71 1.88 1.78 2.13 1.93
0 1.58 1.55
0
F_2_F F_2_R F_2_D F_3_F F_3_R F_3_D
KH
E_2_F E_2_R E_2_D E_3_F E_3_R E_3_D
Types of laminated float glass Types of laminated tempered glass
Fig. 9.10 Forces at different “Stages” of Fig. 9.11 Forces at different “Stages” of
laminated float glasses laminated tempered glasses
AN
Fig. 9.10 and Fig. 9.11 indicate that in the case of laminate with bonded glass layers, the force
(Fu2 - Fu1) at Stages B2 to C1 significantly increased compared to the force of laminate with
non-bonded layers. Before the fracture of the last glass layer (Stage B1), the force vs.
deflection curve is similar to a single glass specimen strengthened with interlayer material.
K.P
Therefore, the deflection increases from y12 to y1 (Pankhardt, Balázs, 2006). The force of the
remaining glass layer in Stage B1 increases by 19 % in the case of EVA interlayer and by
12.6 % in the case of laminated glass consisting of two tempered glass layers compared to
non-bonded glass layers. The force of the remaining glass layer in Stage B1 increases by
t by
37.4 % (≈2×19 %) in the case of EVA interlayer and by 24.5 % (≈2×12.6 %) in the case of
laminated glass consisting of three tempered glass layers. EVA foil interlayer material is more
effective until reaching the maximal force of the laminate (see also Figs. B.9.9 and B.9.10).
Stage A Stage B2-C1 Stage B3-C2 Stage A Stage B2-C1 Stage B3-C2
h
25 80
Deflection, y (mm)
Deflection, y (mm)
2.75 70 12.97 12.14

yrig
20 0.54 3.05
4.84 2.77 60
3.72 37.76 33.19 6.57 10.82
15 8.90 50 29.93
6.65 5.82 37.84
5.77 40
10 18.68 30 53.84
17.66 48.83
20 33.75 37.70 33.18
5 11.46 11.59 10.22 11.00 26.57
10
0
0 E_2_F E_2_R E_2_D E_3_F E_3_R E_3_D
cop
F_2_F F_2_R F_2_D F_3_F F_3_R F_3_D

Types of laminated tempered glass
Types of laminated float glass
Fig. 9.12 Deflections at different Stages of Fig. 9.13 Deflections at different Stages of
laminated float glasses laminated tempered glasses
The deflection of float or tempered laminated glass consisting of resin bonded two glass
layers at Stage A (until reaching the maximal force) decreases by 30 to 38.5 % compared to
laminate without bonded glass layers. This decrease is 32 to 37.7 % in the case of resin
bonded three glass layers at Stage A.
The deflection of float or tempered laminated glass consisting of EVA bonded two glass
layers at Stage A decreases by 37.3 to 38.6 % compared to laminate without bonded glass
layers. This decrease is 42.1 to 45.6 % in the case of EVA bonded three glass layers at Stage
A.
48
9.2.3 Stress-strain relationship
Fig. 9.14 indicates that the ultimate strain in the Region 1 increases more in the case of EVA
laminated glass specimens, therefore, the appropriate bond of glass layers is ensured at 23 °C
temperature with the use of EVA interlayer, which affects the load bearing capacity of
laminated glass.
3000 2000
+23 °C Tempered_EVA +23 °C Tempered_EVA
y = 1839.67e 4.00x Tempered_Resin 1800 Tempered_Resin
Ratio of ultimate strain in Region 1 to

2500 Float_EVA Float_EVA
20 mm/min R 2 = 0.98 Float_Resin 1600 20 mm/min Float_Resin
ultimate Force of Stage A , ε u / Fu

2213.37 Expon. (Tempered_EVA) Expon. (Tempered_EVA)
ultimate strain in Region 1
1942.37 1400 1373.37 -47.01x
DT
2000 2068.80 y = 1365.49e
1306.28 2 y = 1339.34e
-20.33x
1846.14
at Stage A , εu (µm/m)
1849.51 1200 R = 0.98

y = 1836.53e 0.38x R 2 = 0.99
-48.18x
1500 R 2 = 0.49 1000 y = 1411.50e
(µm/m/kN)
y = 1415.33e -20.93x
y = 503.95e 10.83x 800
2
R = 0.99 2
R = 0.99
1000 831.26 R 2 = 0.98
600
691.92 348.48 311.86
AR
400 302.59 154.87
500 508.80 y = 506.59e 2.51x 605.48 664.19 159.63
200 323.17 151.34
R 2 = 0.98 154.05
0 0
0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.00 0.02 0.04 0.06 0.08 0.10 0.12
total thickness of interlayer / total thickness of laminated glass
total thickness of interlayer / total thickness of laminated glass total h int / total h (-)
KH
total h int / total h (-)
Fig. 9.14 Ultimate strain in the Region 1 at Fig. 9.15 Ratio of ultimate strain and
Stage A of laminated glass ultimate force at Stage A
With the increase of the ratio of total thickness of interlayer to total thickness of laminated
AN
glass in the case of EVA interlayer the ultimate strain increases, therefore, the load bearing
capacity of laminated glass increases. The ultimate strain does not significantly increase in the
case of resin interlayer material with the increase of this ratio at 23 °C. Fig. 9.15 also
K.P
indicates that EVA laminated glass behaves relatively monolithically compared to resin
laminated specimens at 23 °C.
Strain gauges placed on the bottom surface of glass specimens can only measure strains until
the fracture of the lower glass pane. Therefore, further strain gauges were placed on the upper
surface of the upper layer of laminated glass, and were able to measure strains during the
t by
whole fracture process. Fig. 9.16 and Fig. 9.17 indicate the strain of the upper surface of
laminated glass vs. deflection and force vs. deflection curves of tempered laminated glass
consisting of two or three glass layers.
30 -1200
Force_F_19mm
20 mm/min +23 °C
h
Force_F_3_F
Force_F_3_R
yrig
25 Force_F_2_F -1018.60 -1000

Force_F_2_R
Strain of upper glass layer, ε upper,mid
Force_F_3_D -909.79-852.00 -902.80

Strain_F_19mm -791.88
20 Strain_F_3_F -799.22 -800
Strain_F_3_R -715.57 -706.49
Force, F (kN)
Strain_F_2_F -685.63 -666.52 -731.88

Strain_F_2_R
cop
-624.00 -600
(µm/m)
15 Strain_F_3_D -620.85 -628.54

-588.12
-543.84
10 -400
5.85
5 4.49 3.12 -200
3.00
2.35 2.30
2.05 2.13 0.740.870.85
0.65 0.68
0 00 0.41 0.51 0.54 0
0 5 10 15 20 25
Deflection, y (mm)
Fig. 9.16 Strain in middle of upper glass layer of laminated glass specimens consisting of two
(_2) or three (_3) float glass layers, laminated with resin (_R) or EVA foil (_F) or without
interlayer (_D); also, single glass specimens with thickness of 19 mm is illustrated.
49
Fig. 9.16 indicates that strain at the upper surface of the upper glass layer decreases during the
fracture process of laminated float glass in the case of EVA interlayer material. In the case of
resin interlayer laminated float glass, the strain decrease is 3% and it is smaller than that of
EVA laminated specimen (for EVA it is 13 to 24%) after reaching the maximal force (Stage
B). Strain starts to increase with the increase of deflection until the fracture of the next glass
layer (Stage C). Therefore, in the case of small deflections (until y < h) the resin interlayer
transfers forces with relatively small shear deformations. The effective thickness of laminated
glass decreases after the fracture of one glass layer and with the increase of deflection (y > h),
strain also increases. In the case of EVA interlayer, higher strain at the upper surface indicated
that the EVA interlayer material has a higher bond capacity than the resin interlayer material
DT
at 23 °C.
Fig. 9.17 indicates that strain at the upper surface of the upper glass layer increases during the
fracture process when the force reaches the ultimate value from Fu3 to Fu2 and to Fu1 of
AR
laminated glass consisting of two or three tempered glass layers. Therefore, the tempering
plays an important role in the case of safety laminated glass. During the fracture process, the
deflection increases and the maximal load bearing capacity of the individual glass layers can
be reached when the appropriate bond is ensured between the surfaces of glass layers. The
KH
surface strength of individual glass layers influences the post-failure behaviour of the
laminate. In the case of EVA laminated tempered glass specimens, the decrease of strain after
fracture of the first glass layer (Stage B) is higher (by 6 to 8 %) than in the case of resin
AN
laminated float glass specimens. The bond decreases in the case of resin interlayer with the
increase of force and deflection.
60 -3000
+23 °C
K.P
-2581.71
50 20 mm/min -2294.29
-2500
-2399.89 -2291.89

-2131.31 -2190.76 -2226.41
-2066.68
40 -1976.39 -2000
-1874.33 Force_E_19mm
Force, F (kN)
Force_E_3_F
-1743.54 Force_E_3_R
t by
30 -1618.83 Force_E_2_F -1500 (µm/m)

-1279.77 -1322.68 Force_E_2_R
Force_E_3_D
-1164.23
-1138.51 Strain_E_19mm
20 -1000
Strain_E_3_F
15.01 Strain_E_3_R
12.77 Strain_E_2_F
h
10 7.606.30 7.36 Strain_E_2_R -500

Strain_E_3_D
yrig
5.26 4.93 0.96 1.78 1.93 2.13

0 00 0.92 1.16 0
0 10 20 30 40 50 60 70 80 90
Deflection, y (mm)
Fig. 9.17 Strain in middle of upper glass layer of laminated glass specimens consisting of two
cop
(_2) or three (_3) tempered glass layers, laminated with resin (_R) or EVA foil (_F)
Based on my laboratory research carried out with laminated glasses, I have determined that
the interlayer material affects differently the strain of different regions of glass layers. The
higher ultimate strains were measured in the R 2 region in the free edge of the plate.
Therefore, the interlayer material affects the stresses of unsupported edges (or bore holes).
The difference between strains in the middle and edge region is more reduced in the case of
resin laminated glasses than in the case of EVA foil laminated glasses. Fig. 9.18 indicates the
ultimate strains at maximal force of laminated glasses in the middle region of the glass pane
(R 1) and in the edge region (R 2) on the bottom glass surface.
50
2312.39
Region 1 two glass +23 °C
layers 2026.25 1977.44 2213.37
Region 2
2500 three glass 20 mm/min
Region 1 1942.37 2313.32
Ultimate strain in bottom

layers 1738.56
1904.85
2000 Region 2 2068.80
surface, ε u (µm/m)
1906.14
1849.51
934.14 1599.38
1500
658.14 679.81
664.19 831.26
543.30
1000 713.40 638.10 742.49
DT
500 621.61 691.92 Region 2
605.48
Region 1
Region 2 3
0 Region 1
2
F_D_6mm
non F_R EVA F_F
bonded resin E_D_6mm
foil non bonded E_R EVA E_F
resin foil
Number of glass
AR
tempered layers, n
float
(pcs)
Types of laminated glasses
Fig. 9.18 Strain in bottom surface of bottom glass layer of laminated glass specimens
KH
consisting of two or three float (F_) or tempered (E_) glass layers, laminated with resin (_R)
or EVA foil (_F) or without interlayer (_D) at 23°C, loading rate of 20 mm/min, (Pankhardt,
2010b).
Experimentally I have shown that the ultimate deformation (strain) in the region near
AN
the edge is 17 to 19% higher in the case of laminated float and tempered glasses without
interlayer material (only with spacer) and at temperature of +23 °C than that measured
in the middle of the glass surface. Furthermore I have shown that the deformations
K.P
(strains) measured in the edge region of the glass surface exceed by 5 to 12 % the value
measured in the middle region in the case of laminated glasses with EVA foil interlayer
and temperature of +23 °C. This difference is only 2 to 5 % with resin interlayer
material (Pankhardt & Balázs, 2006; Pankhardt, 2008a; Pankhardt, 2008b; Pankhardt &
Balázs, 2010b), [2.2 New Scientific Result: Effect of interlayer material on ultimate strains of
t by
glass].
The chemical bond between the resin and the glass surface affects the ultimate strain of glass.
In the case of resin laminated glasses a chemical link forms between the resin and the silanol
(SiOH) groups on the glass surface (Kadri, 2003). As the resin is liquid, it perfectly fills the
h
space between the glass layers, hence curing it is more ideal for use with imperfectly smooth
glass surfaces (especially for tempered glass) than EVA foil, because the bond behaviour is
yrig
improved in the edge region.

Higher strains measured in region R 2 are confirmed in the research of Pagano (1989) on the
interlaminar shear of composite materials and of Lagunegrand et al. (2006) on laminated
sandwich beams. They have shown that the stress field is complex at a free edge laminate
cop
(Figs. B.9.13 and B.9.14) which implies the occurrence of delamination (especially by thin
plates) and thus the increase of stresses in the region of the free edges. It should be noted that
the thermo-mechanical behaviour of the interlayer materials should be taken into
consideration (see Chapter 11). The chains of the polymer interlayer material arrange
differently in the direction of loading in differently loaded regions (mid-pane, edge of pane),
therefore, the properties of plastics with a hardening effect in tension – such as EVA at
+23 °C (see App. C: Testing of interlayer materials) – affects differently the strain in the
regions of laminated glasses.
9.2.3.3 Stresses in mid-region (Region 1) of pane
To study the bending characteristics of laminated glass specimens, the Young’s modulus
should also be determined. Force vs. deflection curves (Figs. 9.7 and 9.8) indicated linear
51
elastic behaviour until reaching the ultimate force when applying a loading rate of 20
mm/min. The theoretical surface strength of laminated glass at 23 °C is indicated in Table
B.9.1. The calculated Young’s modulus, Efl, represents a global value for laminated glass
including the properties of glass layers and interlayer material.
Table B.9.1 indicates that the increase of strains, therefore surface stresses in Region 2 were
higher than in Region 1. By comparing deflections and surface stresses, Table B.9.1 also
indicates that in the case of higher surface strength and lower deflection, the bond capacity of
the EVA interlayer material is higher at room temperature than in the case of resin interlayer.
The flexural modulus of laminated glass decreases with the increase of the number of glass
layers of laminated glass. The flexural modulus of laminated glass consisting of two or three
DT
glass layers at 23 °C with an EVA foil interlayer is higher than with resin interlayer material.
The bond capacity of interlayer material is influenced by temperature (see Chapters 10 to 11),
rate of loading, etc., which can vary in different regions of laminated glass (e.g. due to
AR
delamination in the edge region).
9.2.3.4 Stresses in edge region (Region 2) of pane
To study the strains, therefore, the ultimate surface stresses of laminated glass without bonded
KH
layers (see Fig. 6.2.a), strain gauges were placed on each surface (upper and lower) of each
individual glass layer of laminated glass both in Region 1 and Region 2. Figs. B.9.15, B.9.16
indicate higher strains, therefore, higher surface stresses in Region 2 of laminated glass,
especially at the bottom surface of specimens. The influence of interlayer material on the edge
AN
stresses has to be taken into account on the load bearing capacity of laminated glass
(Pankhardt, 2008a).
9.3 Conclusions
K.P
Three different stages of the fracture can be identified in the laminate. The load bearing
capacity of float glass specimens is proportional to tempered glass specimens.
The following conclusions can be drawn for the effectiveness of tempering of laminated
glass: I have determined that the effectiveness of tempering is influenced by the number of
t by
glass layers and the property of the applied interlayer material. I have shown that at room
temperature the effectiveness of tempering is more favourable in the case of resin laminated
glasses than in the case of EVA foil laminated glasses. The higher the number of glass layers,
the lower is the effectiveness of tempering in laminate with appropriate bonded glass layers.
h
The effectiveness of tempering provides possibilities for the economical determination of the
composition of laminated glass and the selection of the appropriate interlayer material.
yrig
The following conclusions can be drawn for the interlayer materials of laminated glasses
at room temperature: Laminated glass specimens can exceed the load bearing capacity of
single glass specimens (as monolithic upper limit) with comparable thickness. The load
bearing capacity of laminated glass with EVA interlayer increased more than that of resin
cop
laminated glass. The deflection decreases more in the case of EVA interlayer than in the case
of resin interlayer with an increase in the number of glass layers. EVA interlayer material is
preferred in those applications, where the load bearing capacity of laminated glass is
important or the deflection should be limited because of the properties of sealing or water-
tightness.
I have determined that the interlayer material affects differently the strains of different
regions of glass layers. The influence of interlayer material on the edge stresses has to be
taken into account on the load bearing capacity of laminated glass.
52
Effect of temperature on bending characteristics
Chapter 10
10. Effect of temperature on bending characteristics
10.1 Introduction
When using laminated glass on external surfaces (facades, roofing, etc.), the temperature can
reach about 60 °C in summer or -20 °C in winter (App. B, Fig. B.10.1). The question arises –
especially in the case of load bearing glass constructions exposed to a wide range of
temperatures (such as in certain climates) – how does the temperature affect the load bearing
capacity of single or laminated glasses? This chapter deals with the effect of temperature on
DT
single glasses and laminated glasses. The tested interlayer materials were resin and EVA foil
the properties of which are influenced by temperature, therefore the bending characteristics
of laminated glass are also affected.
AR
Test temperatures of single and laminated glass specimens were -20 °C, +23 °C and +60 °C.
Testing of the interlayer material itself is not enough, because the bond between the glass
KH
surfaces can not be taken into account. Photos were taken during and after testing (App. B,
Fig. B.10.2). Figs. B.10.3 and B.10.4 illustrate the fractured resin laminated tempered glass
specimens, where different fracture patterns can be observed. In the case of glass with a
temperature of +60 °C, the fragments (Figs. 10.1 a, b) were smaller than in the case of glass
AN
with a temperature of -20 °C.
K.P
Fig. 10.1 Schematic representation of the fracture pattern of laminated glass

a) at -20 °C, b) at +60 °C
To better study the fragmentation pattern of tempered laminated glass, the surfaces of
t by
fractured specimens were painted and then cleaned to better observe the contour lines. The
painted fragmentation pattern indicated that it differs with temperature, see Figs. B.10.13 a, b.
I have determined that the cracks are denser at +60 °C than at -20 °C in pane regions which
have higher stresses. At -20 °C, resin laminated glass behaves more rigidly than at room
h
temperature, which is also indicated by the larger regions of fractured specimens. Therefore,
yrig
the temperature sensitivity of the interlayer material influences the fracture pattern of
laminated glass. The change of the fracture pattern and the creation of the regions influence
the load transfer capacity of the interlayer materials in safety glasses at low temperature
(Pankhardt, 2010b).
cop
According to my experimental results, the secondary cohesion effect of the resin

interlayer validates during the development of cracks on the tensioned side of laminated
glasses at low (-20 °C) temperature. During the fracture process of safety glasses, glass
fragments can build partially connected regions. In this region the bond is favourable
compared to single fragments. As a result of this effect, the deflection of the glass will be
reduced under further loads (Pankhardt, 2008b; Pankhardt & Balázs, 2010b; Pankhardt,
2010c), [5.1 New Scientific Result: Effect of temperature on fragmentation pattern of
laminated glasses].
The deflection of safety laminated glasses consisting of two glass layers with a temperature of
-20 °C was 16.1 % lower after the fracture of the first glass layer, at 1.0 kN force level,
compared to deflection at +23 °C, with resin interlayer, and it was 7.5 % lower with EVA foil
interlayer material. The deflection of non safety laminated glasses consisting of two glass
53
layers with a temperature of -20 °C was 23.6 % lower after the fracture of the first glass layer,
at 0.6 kN force level, compared to deflection at +23 °C, with resin interlayer, and it was 5.6 %
lower with EVA foil interlayer material. Figs. B.10.5 to B.10.8 indicate fractured float
laminated glasses consisting of two or three glass layers laminated with resin interlayer
material and Figs. B.10.9 to B.10.12 indicate EVA laminated float glasses.
The importance of this observation is on the post-critical behaviour of laminated glasses,
because the remaining connected regions also influence the residual load bearing capacity and
the flexural stiffness. Better bond is ensured (the elongation of the interlayer material is
reduced on the tensioned side) in larger connected glass surfaces (regions) than in the case of
single fragmented fracture pattern. This observation is confirmed by the so-called tension-
DT
stiffening effect, known in reinforced concrete structures (see Fig. B.10.14, CEB-FIP Model
Code, 1990). In the future further investigations should be done on the results of the
fragmentation pattern, which is not presented in this Thesis.
AR
10.2.1 Effect of temperature on load bearing capacity of single glass
layer
10.2.1.1 Load – deflection relationship
KH
Single float and tempered glasses with different temperatures (-20 °C, +23 °C, +60 °C) were
tested to determine the influence of temperature on the load bearing capacity of glass panes.
Force vs. deflection of tempered glass was not affected by temperature. Non heat-treated float
AN
glass specimens were affected by temperature. Especially at -20 °C, behaviour was more rigid
than at room temperatures, see Fig. 10.2 (Pankhardt 2008b).
2.50
1.6 F_20°C 6 mm 1.62 50 mm/min
1.56
1.58 E_tempered F_float
F_23°C 50 mm/min
K.P
1.4 2.00
F_+60°C 1.62
1.2 1.56 1.58
Force, F (kN)
E_-20°C 1.50
Force, F (kN)
1
E_23°C
0.8 0.79
E_+60°C 1.00
0.79
0.6 0.54
Lineáris 0.54 0.47
0.47 0.50
0.4
t by
0.2
0.00
0 0 -30 -20 -10 0 10 20 30 40 50 60 70
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 Temperature, T (°C)
Deflection, y (mm)
Fig. 10.2 a) Effect of temperature on force vs. deflection of 6 mm thick single glass
h
specimens, b) Effect of temperature on ultimate force of single glass, where the symbols
yrig
mean: E- tempered glass, F- non heat-treated float glass

The load vs. deformation behaviour of tempered glass is easier to predict at different
temperatures than that of float glass. Therefore, in the case of outdoor load bearing
applications, tempered glass should be used. It must also be mentioned here that tempered
cop
glass should be heat soak tested (to avoid spontaneous failure, App. A, Fig. A.2.11) in the case
of outdoor applications (see App. D, Definitions).
10.2.1.2 Ultimate strain in mid-region (Region 1) and edge Region
(Region 2) of pane
Fig. 10.3 indicates that the ultimate strain of Region 2 is higher for both float and tempered
glasses. The ultimate strain in Region 2 is 10 % higher than that of in Region 1 in the case of
float glass at 23 °C. This difference increases up to 23 % with decrease of temperature from
23 °C to -20 °C in the case of float glass, but remains 10 % in the case of tempered glass. The
ultimate strain increases with increase of temperature for both float and tempered specimens.
The ultimate strain in Region 2 is only 4 % higher than that of in Region 1 in the case of
tempered glass at 60 °C, but remains 10 % in the case of float glass.
54
3000
50 mm/min F_1_6mm_mid
F_1_6mm_edge
2500 E_1_6mm_mid
E_1_6mm_edge
Polinom. (E_1_6mm_mid)
2137.75
ultimate strain, ε u (µm/m)

2041.28
2000
1687.75 2056.66
1841.26 Region 2
1500 Region 1
1531.63
Region 2
1000 Region 1
763.22 727.46
565.17
500 658.72
621.20 514.17
DT
-30 -20 -10 0 10 20 30 40 50 60 70
Temperature, T (°C)
Fig. 10.3 Ultimate strain at bottom surface of Region 1 and Region 2 of float and tempered
glasses with thickness of 6 mm
AR
Table B.10.1 indicates ultimate strains of Region 1 and Region 2 and the calculated surface
stresses. In the case of tempered specimens, the ultimate strain is affected more by the
compression in the outer layers than by the thermal expansion of cooled or heated specimens.
Therefore, non heat-treated glass is more temperature sensitive than tempered glass.
KH
10.2.2 Effect of temperature on load bearing capacity of laminated glass
10.2.2.1 Load – deflection relationship
AN
Figs. 10.4 to 10.7 indicate the effect of temperature on force vs. deflection diagrams of non
safety (float) and safety (tempered) laminated glass consisting of two or three glass layers,
laminated with resin or EVA foil interlayer material (App. B, Tables B.10.2 and B.10.3). The
lower limit of the load bearing capacity of laminated glass is indicated by specimens without
K.P
interlayer material.
9
-20°C ;+23 °C;+60 °C E_12mm_23°C
8 7.81 7.94 E_2_R_-20°C
glass: 2×6 mm E_2_R_23°C
7 E_2_R_+60°C
t by
Resin 6.30 F_2_R_-20°C

6
F_2_R_23°C
Force, F (kN)
20 mm/min 5.35
5 F_2_R_+60°C
E_2_D_23°C
4
2.75 2.87
3
h
2.05 2.09
2 1.90
1.66
yrig
1 0.87 0.80 1.37 1.58 1.78

0.51 0.91 0.92 1.18
0 0 0.43 0.500.61
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80
Deflection, y (mm)
Fig. 10.4 Force vs. deflection diagrams of non heat-treated float (F_) and tempered (E_)
cop
glasses (with 2×6 mm glass layers) laminated with resin (_R) interlayer at temperatures of
-20 °C, +23 °C and +60 °C; tempered single glass layer with thickness of 12 mm (E_12_)
as well as specimens without an interlayer (E_2_D_) are indicated. Dashed lines indicate
results for the tempered, continuous lines, the float glasses.
Results showed that the behaviour of laminated glass is influenced by the temperature for
both non heat-treated laminated glass and for tempered laminated glass. By comparing Figs.
10.4 to Fig. 10.5 the resin interlayer material is more temperature sensitive than EVA foil
interlayer material, which affects the behaviour of laminated glasses. With the increase of
temperature, the deflection increases while the ultimate force decreases. This behaviour is
more pronounced in the case of resin interlayer material than in the case of EVA foil
interlayer material. The stiffness of the resin interlayer material significantly increases with
55
the decrease of temperature from 23 °C to -20 °C, therefore, laminated glass with resin
interlayer behaves more rigidly than EVA laminated glass or monolithic single layer glass.
9
-20°C ;+23 °C;+60 °C E_12mm_23°C
8 7.94 E_2_F_-20°C
glass: 2×6 mm 7.16 E_2_F_23°C
7 6.64 E_2_F_+60°C
EVA foil 6.58
F_2_F_-20°C
6 F_2_F_23°C
20 mm/min
Force, F (kN)
5 F_2_F_+60°C
E_2_D_23°C
4
3 2.87
2.30 2.30 2.38
DT
2.25
2 1.88
1.88
1 0.680.80 0.81 1.37 1.58
0.99
0.43 0.51 0.86 0.85
0 0 0.41
AR
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80
Deflection, y (mm)
glasses (with 2×6 mm glass layers) laminated with EVA foil (_F) interlayer at
KH
temperatures of -20 °C, +23 °C and +60 °C.
Based on my laboratory results I have introduced the definition temperature sensitivity of
laminated glasses to show the change of load bearing properties of laminated glasses
with the change of temperature. The definition of temperature sensitivity indicates the
AN
magnitude of the change of a property (e.g. load bearing capacity) of laminated glass
with the change of temperature. I have shown that laminated glasses with EVA foil are
less temperature sensitive than glasses laminated with resin (Pankhardt & Balázs, 2006;
K.P
Pankhardt, 2008b; Pankhardt & Balázs, 2010b; Pankhardt, 2010e), [3.1 New Scientific
Result: Temperature sensitivity of laminated glasses. Change of force versus deflection
relationship with the effect of temperature].
The maximal resisting force increases by 0 to 8 % at -20 °C compared to +23 °C in the case
of laminated glasses consisting of two glass layers and EVA interlayer, but it decreases by 1
t by
to 2 % with increase of the deflection by 28 % at a temperature of +60 °C. The maximal

resisting force increases by 4 to 24 % at -20 °C compared to +23 °C in the case of laminated
glasses consisting of two glass layers and resin interlayer, but it decreases by 15 to 19 % with
increase of the deflection by 35 % at a temperature of +60 °C. App. B, Tables B.10.4 and
h
B.10.5 give measured force values for given deflections (Ls/200 = 5 mm) for better
comparison (see also Figs. B.10.15 to B.10.16). At high temperatures (+60 C°) both interlayer
yrig
materials will soften. At lower temperatures (-20 C°) the resin interlayer behaves more
rigidly. It has an effect on the load bearing capacity of the laminate, (Figs.B.10.17 and
Fig.B.10.18).
Figs. B.10.19.a to B.19.d illustrate the quantitative change of ultimate force and deflection
cop
with change of temperature of laminated glasses with EVA interlayer material, compared to
+23 °C in percent. Figs. B.10.20.a to B.10.20.d illustrate the quantitative change of ultimate
force and deflection with change of temperature of laminated glasses with resin interlayer
material, compared to +23 °C in percent. Values for laminated glass with non bonded layers
at +23 °C were also illustrated. The ratio of standard deviation and the average value (of
ultimate force or deflection) of tested laminated glasses are indicated on the top of the
columns.
In the case of exterior use of laminated glasses, the EVA interlayer material is recommended
compared to the tested resin, thus with EVA the laminated glass is less temperature sensitive
(Pankhardt & Balázs, 2006). Knowing the temperature sensitivity of laminated glasses is
important especially for load bearing glasses, thus it influences the selection of the interlayer
materials and calculations of load bearing laminated glasses.
56
20
18.85 glass: 3×6 mm E_19mm_23°C
18 E_3_R_-20°C
Resin E_3_R_23°C
16 15.52 E_3_R_+60°C
F_3_R_-20°C
14 F_3_R_23°C
12.77
F_3_R_+60°C
Force, F (kN)
12 E_3_D_23°C
10 9.21 10.40 -20°C ;+23 °C;+60 °C
8 7.36
20 mm/min
6 5.46 6.31 5.26
4.49
4.93 4.64
4 2.97 3.19 3.99
DT
3.002.13 2.26 2.37 1.861.92 1.93
2 2.22 1.77 1.30
2.13 0.83 0.85 0.98 1.55
0.46 0.65 0.66 0.68 1.16 1.29
0 0
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80
Deflection, y (mm)
AR
glasses (with 3×6 mm glass layers) laminated with resin (_R) interlayer, at temperatures
of -20 °C, +23 °C and +60 °C.
KH
By comparing Figs. 10.4 to Fig. 10.6 and Figs. 10.5 to Fig. 10.7 it can be seen that the
temperature sensitivity of laminated glass increases with the increase of the number of glass
layers, or with the increase of the total thickness of interlayer material in laminated glass.
18
AN
glass: 3×6 mm E_19mm_23°C
16 15.52 E_3_F_-20°C
15.01 EVA foil E_3_F_23°C
14.49 E_3_F_+60°C
14
13.01 F_3_F_-20°C
12 F_3_F_23°C
F_3_F_+60°C
Force, F (kN)
K.P
10 E_3_D_23°C
8.01 -20°C ;+23 °C;+60 °C
8 7.60 8.08
5.85 20 mm/min
6 5.12 5.82
4.07 5.26
3.99
4 3.33 4.85
3.12 2.13 2.40
2.91 2.26 2.37 2.46
t by
2 2.35 1.08
2.14 2.02 0.80 1.00
0.55 0.54 0.740.95 0.96 1.29 1.43 1.55
0 0
0 10 20 30 40 50 60 70 80 90 100 110
Deflection, y (mm)
h
glasses (with 3×6 mm glass layers) laminated with EVA foil (_F) interlayer, at
yrig

Fig. 10.8 indicates the force versus relative thickness relationships of glass panes with
different temperature, and the divergence of curves increases with increase of the number of
glass layers (with increase of relative thickness of resin), thus indicating the temperature
cop
sensitivity. Experimentally I have shown that the temperature sensitivity of laminated

glasses with resin increases, with the increase of the number of glass layers. According to
my laboratory results, laminated glasses with EVA foil are less temperature sensitive
than laminated glasses with resin foil, with the increase of the number of glass layers
(Pankhardt & Balázs, 2006; Pankhardt, 2008b; Pankhardt & Balázs, 2010b; Pankhardt,
2010c), [3.2 New Scientific Result: Effect of number of glass layers with different temperature
on load bearing properties].
In the case of EVA foil, the maximal force of laminated glass can be increased with the
increase of the relative thickness of EVA foil (until the appropriate bond is ensured at the
given temperature). This is confirmed by the observations of Behr et al. (1984), Vallabhan et
al. (1987), Norville (1997), on laminated glasses with PVB foil and consisting of two glass
layers. According to those, in the case of tests carried out at room temperature, the load
57
bearing capacity increased with the increase of the foil thickness. The load bearing capacity
does not significantly improve (the slope of the curve decreases) at higher temperatures with
the increase of the relative thickness of resin.
Tempered_EVA_-20°C Tempered_EVA_23°C Tempered_EVA_+60°C
Float_EVA_-20°C Float_EVA_23°C Float_EVA_+60°C
Tempered_Resin_-20°C Tempered_Resin_23°C Tempered_Resin_+60°C
Float_Resin_-20°C Float_Resin_23°C Float_Resin_+60°C
Expon. (Float_Resin_23°C) Expon. (Tempered_EVA_23°C) Expon. (Tempered_Resin_23°C)
20
with EVA -20 °C with Resin -20 °C y = 1.49e23.44x
18 y = 1.52e50.08x y = 0.41e57.71x 2
+23 °C 2 18.85 R = 0.98
R 2 = 0.99 R = 0.98
16 15.01 19.78x
y = 1.51e49.89x +60 °C y = 0.75e 40.42x +23 °C y = 1.53e
maximal force, Fmax (kN)
2
14 R 2 = 0.98 2
R = 0.96 R = 0.99
14.49
DT
y = 1.54e47.33x 13.01 y = 0.54e45.54x 12.77 y = 1.53e17.76x
12 +60 °C
2
R = 0.99 +23 °C 2 R 2 = 0.98
R = 1.00
10 -20 °C
7.81 10.40 y = 0.77e18.15x
8 20 mm/min 7.16 +60 °C 2
R = 0.98
AR
6.64 6.30 -20 °C y = 0.42e22.15x
6 6.58 5.85 5.46
5.12 5.35 2
R = 0.99
4.49 +23 °C
4 4.07 2.75 16.53x
2.30 +60 °C y = 0.53e
2 2.30 2.05 3.19 2
R = 0.98
1.58
1.56 2.25 1.66
KH
0.79
0.54
0.43
0
0.00 0.02 0.04 0.06 0.08 0.10 0.12
total thickness of interlayer / total thickness of laminated glass
h int,tot / h tot (-)
AN
Fig. 10.8 Maximal force versus ratio of total thickness of interlayer material to total thickness
of laminated glass at temperatures of -20 °C, +23 °C and +60 °C. Dashed lines indicate the
tempered, continuous lines the non heat-treated, float glasses (Pankhardt, 2010b).
K.P
At high temperatures, resin interlayer material will soften with a decrease of bond strength,
therefore, the ultimate force of laminated glass significantly decreases. The decrease of bond
strength is also indicated by the increase of deflections with an increase of temperature from
-20 °C to +60 °C. Fig. 10.9 indicates the effect of temperature on ultimate deflections at Stage
t by
A. The effect of temperature on load bearing capacity and on deflections, especially in the
case of load bearing laminated glasses, should not be neglected.
80
+60 °C E_R_+60°C E_F_+60°C E_6_12_19mm
E_R_23°C E_F_23°C E_R_-20°C
70 +23 °C E_F_-20°C F_6_12_19mm F_R_+60°C
h
61.63 63.34 F_F_+60°C F_R_23°C F_F_23°C

-20 °C F_R_-20°C F_F_-20°C
60 60.01
ultimate deflection, yu (mm)
yrig
59.18 20 mm/min
51.02
50 45.77
Tempered 46.57
43.23
40 37.70
Float 33.75
33.55 32.36 33.18
32.47
30
cop
28.28 26.57 24.57 24.93

22.16 22.41
20 18.76 17.03
15.18
16.2 16.03 12.19 11.92 11.72 15.88
10 11.46 h>yu 11.00
11.59 10.22
10.96 8.35 8.10
0
4 8 12 16 20
total thickness, h total (mm)
Fig. 10.9 Change of ultimate deflection at Stage A of laminated glass specimens laminated
with resin or EVA foil interlayer materials at temperatures of -20 °C, +23 °C and +60 °C.
The importance of this conclusion is that the number of layers and thickness of the interlayer
material can be optimalised if the load bearing capacity should be taken into account when the
58
application temperature and the exposure class of the laminated glass are already specified
(see Chapter 11).
Based on laboratory experiments, Whitney (1987) has shown on fibre-reinforced polymers in
three point bending that with the increase of the ratio of span length to thickness (Ls/h) the
proportion of shear deflection to total deflection decreases (see Fig. B.10.21). With the
increase of number of glass layers from two to three, the ratio Ls/h decreases from 78.49 to
50.99 in the case of resin laminated glasses and from 82.37 to 54.32 in the case of EVA foil
laminated glasses. Therefore, the load bearing property of laminated glasses consisting of two
glass layers are less influenced by the temperature than those of laminated glasses consisting
of three glass layers. In the case of laminated float glasses consisting of two or three glass
DT
layers (with thickness of single glass layers of 6 mm) deflections remain smaller than the total
thickness of laminated glass with no significant shear deformation of the interlayer material.
The strength factor (SF) of laminated glass can be calculated by Eq. (6.1) and can also be
AR
indicated with a ratio of maximal force of laminated glass to single glass layer with equivalent
thickness (Pankhardt, 2010c). Fig. 10.10 indicates that this ratio significantly decreases with
an increase of temperature in the case of laminated glass with resin interlayer material.
Therefore, it is not preferred to increase this ratio or the strength factor – which is a consensus
KH
nowadays, as mentioned in Chapter 6 – without specifying the exposure class on different
types of laminated glasses.
1.40 1.40
20 mm/min F_2_R F_3_R 20 mm/min F_2_F F_3_F
AN
Strength factor at Stage A , SF (− )
Strength factor at Stage A , SF (− )
1.21 1.20
1.20 E_2_R E_3_R E_2_F E_3_F
1.03 Resin 0.93 0.97 EVA
1.00 1.00
0.98 0.82 0.90 0.86 0.85
0.79 0.86 0.84 0.84
0.80 0.80
0.67 0.81 0.83
0.76 0.77 0.71
0.67
K.P
0.60 0.62 0.60

0.62 0.57
0.40 0.44 0.40
0.20 0.20
0.00 0.00
-30 -10 10 30 50 70 -30 -10 10 30 50 70
t by
Fig. 10.10 Ratio of maximal force of laminated glass to single glass layer with equivalent
thickness at Stage A a) laminated with resin (_R) or b) with EVA (_F) at temperatures of
-20 °C, +23 °C and +60 °C (Pankhardt, 2010c).
h
At low temperatures (-20 °C), the ratio significantly exceeds one, indicating that the interlayer
material acts in flexure. At high temperatures, the ratio remains significantly larger than zero,
yrig
indicating that the interlayer material maintains the ability to transfer horizontal shear force.
Generally, this ratio lies between zero and one.
cop
Fig. 10.11 indicates both strain in the middle of the upper surface of the upper glass layer vs.
deflection and force vs. deflection diagrams. The strain vs. deflection diagram is also affected
by the temperature (see App. B, Tables B.10.6 to B.10.7). At -20 °C and at +23 °C, EVA
behaves relatively monolithically, see also Fig. 10.12. The best bond with use of EVA
interlayer is ensured at +23 °C. Higher force can be reached at -20 °C with a decrease of
strains, as compared to +23 °C. The reason is that the stiffness of EVA interlayer material
increases while the bond between the glass surface and the interlayer material decreases at
low temperatures. With the increase of temperature, the decrease of force and strains with
increase of deflections indicates the softening of the interlayer material. Therefore, the shear
modulus of the interlayer also decreases which decreases the bond between the glass layers.
The load bearing thickness of laminated glass decreases after fracture of a glass layer and with
increase of deflection the strain also increases. The decrease of slope of the curve sections at
59
Stage C2 (between Fu23 and Fu2) and Stage C1 (between Fu12 and Fu1) compared to Stage A
(until Fu3) indicates that the flexural stiffness of laminated glass decreases, see Chapter 9, Fig.
9.3.
50 -1500
Force_F_19mm F_3_F
45 Force_F_3_F_-20°C 20 mm/min -1350
Force_F_3_F_23°C
Force_F_3_F_+60°C
40 -1200
Strain of upper glass layer, εupper,mid

Force_F_3_D_+23°C -1192.08
Strain_F_19mm
35 Strain_F_3_F_-20°C -1050
Strain_F_3_F_23°C -909.79 -904.54
Force, F (kN)
30 Strain_F_3_F_+60°C -900
Strain_F_3_D_+23°C -852.00 -760.08 -731.88
-750
(µm/m)
25
DT
-791.88
-638.91
20 -644.27 -650.33 -676.89 -600
-640.08 -588.12
15 -534.76 -531.03 -450
AR
10 -300
5.12 5.85 4.073.33 -150
5 3.12 2.91 1.00
2.14 2.35 2.020.54 0.55 0.80 0.74 0.95
0 00 0
0 5 10 15 20 25
KH
Deflection, y (mm)
Fig. 10.11 Strain vs. deflection and force vs. deflection diagrams of laminated float glass
consisting of three glass layers with EVA foil interlayer material (F_3_F) at temperatures of
AN
-20 °C, +23 °C and +60 °C; single glass with thickness of 19 mm (F_19mm) as well as
specimens without interlayer (F_3_D) are illustrated.
50 -3000
20 mm/min -2958.76 E_3_F
K.P
45 -2690.80 -2575.22 -2700

-2452.89 -2581.71
-2399.89
40 -2400
Strain of upper glass layer, εupper,mid

-2294.29 -2196.12
35 -2100
-1889.24 -1870.60
Force, F (kN)
30 -1800
Force_E_19mm
-1743.54 Force_E_3_F_-20°C
t by
25 -1611.26 -1500 (µm/m)

-1377.55 -1381.51 Force_E_3_F_23°C
-1279.77 Force_E_3_F_+60°C
20 Force_E_3_D_+23°C
-1200
Strain_E_19mm
15 14.49 15.01 Strain_E_3_F_23°C -900
13.01 Strain_E_3_F_-20°C
10 -600
8.01 7.60 Strain_E_3_F_+60°C
h
8.08 Strain_E_3_D_+23°C
5 4.85 5.26 5.82 -300
2.40
yrig
1.08 0.96 1.43 2.132.46

0 00 0
0 10 20 30 40 50 60 70 80 90 100 110
Deflection, y (mm)
Fig. 10.12 Strain vs. deflection and force vs. deflection diagrams of laminated tempered glass
cop
consisting of three glass layers with EVA foil interlayer material (E_3_F ) at temperatures of
-20 °C, +23 °C and +60 °C; tempered single glass with thickness of 19 mm (E_19mm) as well
as specimens without interlayer (E_3_D) are illustrated.
The bond capacity and strains of the interlayer material play an important role in the post-
failure behaviour of laminated glass during the fracture process (see in detail in Chapter 12).
Stages B1 to B3 indicate how quickly and how effectively the interlayer is able to transfer the
horizontal shear forces to the other remaining glass layers. The decrease of Stage B3 with
increase of temperatures indicate the decrease of effectiveness of the interlayer material to
transfer the forces. The effectiveness of the bond capacity of the interlayer material is
influenced by the simultaneous effect of force and temperature and the fracture pattern during
the fracture process (see Chapter 10.2). Figs. 10.13 and 10.14 indicate that laminated glass
with resin behaves monolithically at -20 °C, but it is temperature sensitive and the bond
60
capacity significant decreases with the increase of temperature. In the case of small
deflections (until y < htot), the resin interlayer transfers forces with relatively small shear
deformations (Fig. 10.13). In the case of resin interlayer laminated glass, the Stage B3 (Fu3 to
Fu23) significant increases with decrease of temperature from +23 °C to -20 °C, and decreases
with the increase of temperature from +23 °C to +60 °C (Fig. 10.14).
50 -1500
Force_F_19mm 20 mm/min F_3_R
Force_F_3_R_-20°C
45 Force_F_3_R_23°C 50 mm/min -1350
Force_F_3_R_+60°C
40 Force_F_3_D_+23°C -1200

Strain_F_19mm
35 Strain_F_3_R_-20°C -1050
Strain_F_3_R_23°C -1031.30
DT
Strain_F_3_R_+60°C -960.00 -902.80
30 Strain_E_3_D_+23°C -900
Force, F (kN)
-730.02 -774.29 -818.56

25 -726.99 -757.28 -799.22
-750
(µm/m)
-685.63
-687.61 -672.93 -600
20 -666.52
AR
-624.00
-578.80
15 -450
10 -300
5.46 4.49 3.19
5 -150
KH
2.97 2.13 3.00 1.77 2.13 0.85
2.22 0.46 0.65 0.83 0.660.68
0 00 0
0 5 10 15 20 25
Deflection, y (mm)
AN
Fig. 10.13 Strain vs. deflection and force vs. deflection diagrams of laminated float glass
consisting of three glass layers with resin interlayer material (F_3_F ) at temperatures of
-20 °C, +23 °C and +60 °C; float single glass with thickness of 19 mm (F_19mm) as well as
specimens without interlayer (F_3_D) are illustrated.
K.P
50 -3000
20 mm/min E_3_R
45 -2563.57 -2700
-2486.45
40 -2400 Strain of upper glass layer, ε upper,mid
-2190.76 -2242.49 -2219.88
-2226.41
35 -1976.39 -2044.43 -2100
-2160.47
-1937.24
t by
Force, F (kN)
30 -1800
-1765.98 -1618.83 -1696.54
25 -1500
(µm/m)
Force_E_3_R_-20°C
-1322.68 Force_19mm
20 -1232.16 Force_E_3_R_23°C -1200
18.85 Force_E_3_R_+60°C
Force_E_3_D_+23°C
15 Strain_E_3_R_-20°C -900
12.77 Strain_E_19mm
h
Strain_E_3_R_23°C
10 9.21 10.40 Strain_E_3_R_+60°C
-600
7.36 Strain_E_3_D_+23°C
6.31 5.26
yrig
5 4.93 -300
4.64 1.30 1.93
0.98 1.92 1.86
0 00 1.16 0
0 10 20 30 40 50 60 70 80 90 100 110
Deflection, y (mm)
Fig. 10.14 Strain vs. deflection and force vs. deflection diagrams of laminated tempered
cop
glass consisting of three glass layers with resin interlayer material (E_3_F ) at temperatures
of -20 °C, +23 °C and +60 °C; tempered single glass with thickness of 19 mm (E_19mm) as
well as specimens without interlayer (E_3_D) are illustrated.
Hussein, Iqbal and Al-Abdul-Wahhab (2005) studied the vinyl acetate content of EVA on the
rheology of polymer modified asphalt and suggested that the tested type of EVA1 of low VA
content would show higher modulus (G’) at high temperature, which is preferred for hot
climates. The reduction of the flow activation energy (see Eq. 4.7 and Chapter 4.4) reduces
the degree of temperature sensitivity, hence, reduces the change of viscosity due to
temperature changes. Therefore, in the case of applying EVA foil in laminated glass as an
interlayer material, it is preferred to investigate EVA foils with low VA content. The exposure
class for laminated glasses with different types of foils should be determined (see also
Chapter 11).
61
Glass strength in Region 2 (edge region) is mainly influenced by the edgework (Chapters 5
and 8). The strength of laminated glass is further influenced by the condition of the interlayer
material near the edges. The edge region of laminated glass is the area most affected by
humidity, temperature or contact with other materials, such as silicon in sealing. If
delamination (Pankhardt & Balázs, 2010b) would occur at the edges, the bond strength in
Region 2 would decrease, therefore the stresses of Region 2 increase, which affects the
strength of the overall laminate. The increase of strains of glass layers in Region 2 highlighted
that mainly this area of laminated glasses are temperature sensitive (Fig. 10.15 and App. B,
Table B.10.7).
DT
Region 1 two glass
layers 20 mm/min
AR
Region 2 3000
Region 1 three glass
Ultimate strain at bottom

2500
glass surface, ε u (µm/m)

Region 2 layers
2000
KH
1500
1000
500
Region 2
Region 1
0
AN
Region 2
-20 23 60 -20 23 60 -20 23 60 -20 23 60 °C

Region 1
3 float_EVA float_resin tempered_EVA tempered_resin

2 Types of laminated glasses at temperatures of
number of glass
-20 °C, +23 °C and +60 °C
K.P
layers, n , (pcs)
Fig. 10.15 Ultimate strain on bottom glass surface at temperatures of -20 °C, +23 °C and
+60 °C of EVA foil or resin interlayer material laminated float or tempered glasses, in mid
pane region (R 1) and edge region (R 2). Loading rate of 20 mm/min (Pankhardt & Balázs,
2010b).
t by
Experimentally I have shown that the reducing effect of resin interlayer material on the
deformation difference in the various regions of glass pane (R 1: mid-pane and R 2: edges)
decreases with the increase of the temperature. It was also mentioned (Chapter 9.2.3.2) that
the complex stress field condition based on the research of Pagano (1989) and Lagunegrand
h
et al. (2006) is influenced by the temperature and thermo-mechanical behaviour of the

yrig
interlayer material. Based on the informal laboratory tensile testing of interlayer materials
(App. C), I have shown that hardening occurs in EVA foil at +23 °C during loading (0.2 and
200 mm/min), at a temperature of -20 °C and +60 °C there is no hardening behaviour. The
force versus elongation curve is nearly linear both for temperature of specimen -20 °C and
+23 °C as well as for +60 °C (with loading rates of 0.2; 2; 20; 200 mm/min), respectively.
cop
The difference of deformations (strains) increases with the decrease of bond strength. In
laminated glasses with a higher temperature, the polymer chains arrange quicker (also under
reduced loading) due to the creep of the interlayer material. Therefore, the difference of
deformations between the mid-pane and edge regions decreases. The resin interlayer creates a
chemical bond with the glass surface (Chapter 4.2.1), which is more effective below room
temperature. By increasing the temperature from 23 °C to 60 °C for float laminated glass
consisting of two glass layers with resin interlayer material, the surface strains at the bottom
surface in Region 2 increase up to the strain value of laminated glass without interlayer
material in Region 2 (strains in Region 2 are about 13 % higher than in Region 1). Therefore,
there is no appropriate bond between the glass surfaces at 60 °C in resin laminated glasses,
delamination can occur. If delamination at the edge region of laminated glass would happen
62
(due to decrease of bond, aging of the interlayer material, etc.), the stresses in Region 2 can
increase up to 19 % and can reach the edge stresses of laminated glass without interlayer
material (see Chapter 9).
In the case of EVA foil at higher temperature (60 °C) the softening is less significant – the
adhesion increases at a small rate – therefore, its compensation effect emerges better (see App.
B, Figs. B.10.22 and B.10.23). The stresses in Region 2 at 60 °C remain about 8 % higher
than in Region 1, according to values calculated at 23 °C.
In the case of calculation of load bearing capacity of laminated glasses with free edges,
containing boreholes and in the case of taking into account durability aspects, a reducing
factor should be applied. Reducing factors should be applied to the calculation of glass
DT
strength for different regions (see Chapter 11). Therefore, it is not preferred to use an overall
design strength, especially for load bearing glass applications in outdoor conditions or for
lifetime predictions.
AR
10.3 Conclusions
The temperature dependent behaviour of the interlayer material affects the load bearing
capacity of the laminate. Based on my laboratory research the following conclusions and new
KH
results can be drawn for the effect of temperature on bending characteristics of laminated
glasses:
I have determined that the fracture pattern is influenced by the temperature of laminated
AN
glasses at -20 °C and +23 °C as well as +60 °C, both for float and tempered glass layers. I
have determined that the cracks are denser at +60 °C than at -20 °C in pane regions which
have higher stresses. During the fracture process of safety glasses, glass fragments can build
partially connected regions. Therefore, the secondary cohesion effect of the interlayer during
K.P
the fracture process should be taken into account.

I have shown, the thermo-mechanical behaviour of applied interlayer material
considerably affects the load versus deflection relationship both in the case of laminated
glasses consisting of (two or three) non heat-treated, float and tempered glass layers with
thickness of 6 mm. I have introduced the definition temperature sensitivity of laminated
t by
glasses. I have shown that laminated glasses with EVA foil are less temperature sensitive than
glasses laminated with resin. Therefore, EVA interlayer material is more effective in
changing temperature conditions in the case of load bearing safety glass applications, where
high ultimate forces have to be resisted (Pankhardt, 2010b).
h
Experimentally I have shown that the temperature sensitivity of laminated glasses with
yrig
resin increases, with the increase of the number of glass layers and with EVA foil laminated
glasses are less temperature sensitive with the increase of the number of glass layers.
The increase of strains in Region 2 highlighted that this area of laminated glasses are
mainly temperature sensitive. It is especially important to study the bond capacity of
cop
interlayer material in those regions, where due to cutting or drilling (for point fixing) the
edges of interlayer are exposed to outdoor conditions. Protection of the highlighted regions is
needed. Reducing factors should be applied to the calculation of glass strength for different
regions. Therefore, it is not preferred to use an overall design strength, especially for load
bearing glass applications in outdoor conditions or for lifetime predictions.
The exposure class for laminated glasses with different types of interlayer materials
should be determined. The recommended testing temperature for laminated glass specimens
should be determined depending on the exposure class (Pankhardt, 2008b).
63
Flexural stiffness
Chapter 11
11. Flexural stiffness of laminated glasses
11.1 Introduction
The flexural stiffness of laminated glass is influenced by temperature. With increasing
temperature the shear stiffness rapidly decreases and creep becomes more significant (Krüger,
1998). According to Wölfel’s (1987) calculations (see Chapter 6), the shear modulus, G, of
the interlayer material has the most influence on the strength and deflection. In the case of
thin or large size glasses (high Ls/htot ratio), where the deformations (deflections) are
DT
considerable, temperature dependent flexural stiffness (Dfl) of the overall laminate is more
significant. This chapter deals with the influence of temperature on flexural stiffness of
laminated glasses. The Author modified the definition of the coupling parameter, κ, which
AR
represents the bond of the glass layers by the interlayer and was originally defined by Krüger
(1998).
KH
11.2.1 Flexural stiffness
The problem for laminated glass calculations is usually the unknown shear stiffness and the
shear modulus, G, of the interlayer material, because of the time dependency of the loads and
AN
because temperature can modify the shear modulus. Deflections can increase with increase of
temperature when the interlayer material softens. Therefore, an appropriate interlayer material
should be selected which is able to transfer forces at the service temperature of laminated
glasses. By determining the exposure class of laminated glasses (see Chapter 11.3), its load
K.P
bearing resistance and its temperature sensitivity should be taken into account (by creating
different classes for them).
The tensile compliance is defined in Chapter 4, Eq. (4.5), where the tensile Young’s modulus
is usually applied for polymer materials. This Thesis proposes to use the flexural stiffness
t by
(Dfl) of laminated glasses to characterise the bending characteristics, which is dependent on

the visco-elastic properties of interlayer materials:
D fl (t , T ) = EI fl (t, T ) (11.1)
where EIfl is the flexural stiffness calculated with Eq. (6.13).
h
5.0E+9
E_2_R F_2_R 2 × 6 mm
4.60E+09
yrig
E_2_F F_2_F
4.5E+9 single 12 mm 2 ply no bond
4.20E+09 2_EVA 2_Resin 20 mm/min
4.0E+9 3.90E+09
2
D fl,2,EVA = -296.36E+3×T - 4.82E+6×T + 4.21E+9
Flexural stiffness, D fl,(Nmm )
2
3.5E+9
3.35E+09
3.0E+9 3.0E+9 3.0E+9 single 12 mm
3.0E+9
cop
2.90E+09
2.5E+9 RESIN EVA
2.25E+09
2.0E+9 D fl,2,resin = -29.97E+6×T + 4.02E+9
2
R = 0.999
1.5E+9
1.0E+9 1.0E+9 1.0E+9 1.00E+09 1.00E+09
1.0E+9
Without bond
Td, 2,EVA ≈ + 96 °C
500.0E+6
Td, 2,resin ≈ + 100 °C
000.0E+0
-30 -20 -10 0 10 20 30 40 50 60 70 80 90 100 110
Temperature, T (°C) T d , Delamination temperature
Fig. 11.1 Flexural stiffness vs. temperature of laminated glasses consisting of 2×6 mm float
(F_) or tempered (E_) glass layers with resin (_R) and EVA foil (_F) interlayer materials
(Pankhardt, 2009; Pankhardt, 2010c)
64
Flexural stiffness
The flexural stiffness of laminated glass types was calculated by using measured average
values of maximal forces (at Stage A) at temperatures of -20 °C, +23 °C and +60 °C. The
flexural stiffness of the overall laminate is influenced by temperature. With increasing
temperature, the flexural stiffness decreases. The tendency of decrease of flexural stiffness at
different temperatures is influenced by the type of interlayer material. The flexural stiffness
affected by temperature is indicated for resin and EVA interlayer materials in Figs. 11.1 and
Fig. 11.2. In Fig. 11, laminated glass consisting of two glass layers without interlayer
material, in Fig. 11.2, laminated glass consisting of three glass layers also without interlayer
material can be seen.
16.0E+9
DT
E_3_R F_3_R 3 × 6 mm
E_3_F F_3_F
13.50E+9 Rigidly bonded 3 ply no bond
14.0E+9 3_EVA 3_Resin 20 mm/min
12.70E+9 11.9E+9 single 19 mm 11.9E+9
12.0E+9
Flexural stiffness, Dfl, (Nmm2)
11.9E+9 11.10E+9
AR
2
D fl,3,EVA = -931.83E+3×T - 31.35E+6×T + 12.42E+9
10.0E+9
EVA
8.10E+9
8.0E+9
RESIN 7.30E+9
6.0E+9
KH
D fl,3,resin = -113.38E+6×T + 11.02E+9 4.10E+9
4.0E+9 2
R = 0.997
1.6E+9 1.6E+9 1.6E+9 1.60E+9
2.0E+9
Without bond 1.60E+9
AN
000.0E+0
Td,resin ≈ +85 °C Td,EVA ≈ +92 °C
-30 -20 -10 0 10 20 30 40 50 60 70 80 90 100 110
Temperature, T (°C) T d , Delamination temperature
Fig. 11.2 Flexural stiffness vs. temperature of laminated glasses consisting of 3×6 mm float
K.P
(F_) or tempered (E_) glass layers with resin (_R) and EVA foil (_F) interlayer materials
(Pankhardt, 2009; Pankhardt, 2010c)
Based on my laboratory results, the flexural stiffness of safety or non-safety laminated
glasses with resin, consisting of two and three glass layers, decreases linearly with the
increase of the temperature in a range between -20 °C and +60 °C. The flexural stiffness
t by
of safety or non safety laminated glass with EVA foil, consisting of two and three glass
layers, can be fitted with a second-degree polynomial function with the best correlation
in a temperature range between -20 °C and +60 °C (Pankhardt & Balázs, 2006;
Pankhardt, 2008b; Pankhardt & Balázs, 2010b; Pankhardt, 2010c), [4.2 New Scientific
h
Result: Change of bending stiffness of laminated glasses as function of temperature].

yrig
The shape of the fitted curves depends on the testing or application temperature range of
laminated glasses (Fig. B.11.1). After determination of the fitted functions, the determination
of delamination temperature is safer than by taking into account the asymptotes of the curves.
The figures indicate that the curve of EVA foil laminated glasses is above the curve of resin
cop
laminated glasses between -10 °C and the delamination temperature, thus the behaviour of
EVA foil is more favourable in this range.
Results without interlayer material indicate the lower limit of flexural stiffness. Therefore, the
temperature can be predicted when the interlayer is not able to transfer shear forces between
the glass layers. The relationships are indicated with the delamination temperatures, Td in
Figs. 11.1 and Fig. 11.2.
Based on my experimental results, I have introduced the definition of delamination
temperature (Td). The delamination temperature gives the temperature (temperature
range) at which the interlayer material cannot appropriately connect the glass layers
and cannot ensure the bond strength. Based on my experimental results, I have
determined the delamination temperature for glasses laminated with EVA foil or resin
and consisting of two or three glass layers, in bending. Based on my experimental
65
Flexural stiffness
results, I have shown that the delamination temperature decreases with the increase of
the number of glass layers (Pankhardt & Balázs, 2010b; Pankhardt, 2010c), [4.1 New
Scientific Result: Introduction of definition of delamination temperature].
The delamination temperature can be determined at the intersection of a fitted curve of
flexural stiffness of laminated glasses with a spacer (lower limit of load bearing capacity) and
a fitted curve of flexural stiffness of laminated glasses with temperatures of -20 °C, +23 °C
and +60 °C (Figs. 11.1 and 11.2). Table 11.1 summarises the glass transition temperature and
delamination temperature ranges.
Table 11.1 Glass transition and delamination temperatures of laminated glasses consisting of
two or three glass layers with 6 mm thickness and with EVA or resin interlayer material
DT
Type of Number of glass Glass transition Delamination
interlayer layers, n temperature, Tg temperature, Td
material pcs °C °C
AR
2 -28 +96
EVA
3 -28 +92
2 -35 +100
Resin
KH
3 -35 +85
When reaching the delamination temperature range, the physical delamination of glass layers
is not certain to happen (see Pankhardt & Balázs, 2010b, and Figs. A.4.10 and A.4.11), but
they are already not bonded or cannot be considered to be bonded. By determination of the
AN
delamination temperature, I have observed that it is near the melting temperature of the
interlayer material (see Table 7.1 and App. C). The delamination temperature is influenced by
the rate and type of loading (e.g. static or cyclic), therefore, further investigations are needed,
K.P
especially in the case of load bearing glasses.

The upper limit of the flexural stiffness of laminated glass can be indicated by the equivalent
monolithic single layer glasses. In this case, the upper limit of flexural stiffness of single glass
layers with thicknesses of 12 mm (Fig. 11.1) and 19 mm (Fig. 11.2) was exceeded by
laminated glasses at those service temperatures where the appropriate bond is ensured and the
t by
stiffness of laminated glasses increased due to the relatively stiff interlayer material (see also
Fig. 10.12).
16.0E+9
20 mm/min with EVA with RESIN
-20 °C
h
14.0E+9
Flexural stiffness, Dfl (Nmm )
single -20 °C
2
yrig
12.0E+9 19 mm
+23 °C
10.0E+9
8.0E+9 +60 °C +23 °C

with EVA with RESIN
cop
6.0E+9
-20 °C -20 °C
4.0E+9 single +60 °C
12 mm +23 °C +23 °C
2.0E+9 +60 °C +60 °C

no bond no bond
no bond no bond
000.0E+0
0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.10 0.11 0.12
Ratio of total thickness of interlayer to total thickness, h int.tot /h tot (-)
Fig. 11.3 Flexural stiffness vs. ratio of thickness of interlayer material to total thickness.
Squares indicate laminated glasses without bonded layers. Triangles indicate laminated glasses with EVA
interlayer and diamonds indicate laminated glasses with resin interlayer, respectively. Empty squares, triangles
or diamonds indicate laminated glasses consisting of tempered glass layers. Solid squares, triangles or
diamonds indicate laminated glasses consisting of float glass layers (Pankhardt, 2010c).
66
Flexural stiffness
Fig. 11.3 indicates the flexural stiffness vs. ratio of total interlayer thickness to total thickness
of laminated glass. The higher values of Dfl in Fig. 11.3 indicate higher flexural stiffness and
stiffer bond. The shift of the values (points) compared to a reference temperature
(e.g. +23 °C) indicates the temperature sensitivity of the interlayer material. The bottom
values (squares) indicate laminated glasses without interlayer material. With increasing
temperature, the shifting of the points tends in the direction of these values. Therefore, resin
laminated glasses (both tempered and float) are more temperature sensitive than EVA
laminated glasses. Fig. 11.3 can properly indicate the temperature or time dependent
behaviour of laminated glasses. With the observation of the shift tendency of the values
(points), the durability of laminated glasses can also be predicted (see Chapter 4, Eq. (4.4)).
DT
11.2.2 Calculations of temperature dependent flexural stiffness of
laminated glasses
The primary interlayer property that influences the strength and deflection is the shear
AR
modulus, G (see Chapter 6). With the increase of the ratio of span length to thickness (Ls/htot)
the proportion of shear deflection to total deflection decreases, therefore, the influence of the
flexural stiffness increases (see Chapter 10.2.2).
KH
Based on my experimental results, I have modified the relationship of temperature
dependent flexural stiffness determined by Krüger (1998). I have determined parameter
κ for laminated glasses with resin or EVA foil, consisting of two and three glass layers,
at temperatures between the glass transition temperature and the delamination
AN
temperature range (Pankhardt, 2008b; Pankhardt & Balázs, 2010b; Pankhardt, 2010c),
[4.3 New Scientific Result: Approaching calculation of temperature dependent bending
stiffness of laminated glasses]. In this Thesis the Author has modified Eq. (6.5) given earlier
K.P
by Krüger (1998):
( E ⋅ I )T ,service = ( Eg ⋅ I g )Td + κ ⋅ ( Ec I c ) (11.2)
- E c I c = ( E ⋅ I )Tg − E g ⋅ I g (11.3)
b ⋅ heff ,T
3
t by
- E g ⋅ I g = 70000 ⋅
d
(11.4)
12
n
- heff ,Td = 3 ∑h
3
i
(11.5)
h
i =1
where
yrig
- (EI)T,service is the flexural stiffness of laminated glass at application service (environment)

temperature;
- (EgIg)Td is the flexural stiffness at delamination temperature, Td;
- (EI)Tg is the flexural stiffness of laminated glass at the glass transition temperature, Tg, of
cop
the interlayer material where the sandwich model (Fig. B.11.2) can be applied when
the interlayer is in a glassy state;
- κ is the coupling term of modified flexural stiffness (EcIc). The value of parameter κ
should be between 0 and 1, based on my laboratory results (Fig. 11.4). In the case of
κ = 1, the stiffness of laminated glass can be calculated with the sandwich model
(Hegedűs, 1983) at the glass transition temperature, Tg. In the case of κ = 0, the glass
layers are not bonded (at delamination temperature). In the case of non bonded glass
layers h = heff,Td should be applied according to Eq. (11.5), where n is the number of
glass layers and hi is the thickness of the ith single glass layer. Parameter κ(T, t) can
be experimentally determined and is dependent on temperature or (and) time (Time-
Temperature Superposition) (Fig. 11.4).
67
Flexural stiffness
The parameter κ versus temperature was determined in Fig. 11.4 in the case of laminated
glasses consisting of two and three glass layers and with resin as well as with EVA foil
interlayer materials. With the increase of temperature, the interlayer softens and thus the bond
strength decreases, therefore the value of parameter κ decreases. The determination of a lower
limit of load bearing capacity is suggested without taking into account the bond between the
glass layers, especially in the case of durability aspects (e.g. creep of polymer, long term
loads, etc.). The temperature dependent behaviour of laminated glass should be studied on the
overall laminate. It is not enough to study the temperature dependent behaviour of the
interlayer material. Therefore, in this Thesis the coupling parameter, κ, was determined
directly for laminated glasses.
DT
1.0
κ3,EVA = 1 κ2,EVA = 1
κ2,EVA = -0.0001T + 0.0001T + 1.0387
2
0.9 2 glass layers

hint,tot/htot=7.85% κ3,EVA= -0.0001T2 - 0.0018T + 0.9752
AR
0.8 2 glass layers
20 mm/min
coupling parameter, κ (-)
hint,tot/htot=3.29%
0.7 3 glass layers
3 glass layers h int,tot/htot=4.35%
0.6 hint,tot/htot=10.20%
EVA
KH
0.5
κ2,resin= -0.0074T + 0.7407
0.4
κ3,resin = -0.0083T + 0.7083
0.3
κ3,EVA = 1 κ2,EVA = 1
AN
0.2 RESIN
T g,EVA = - 28 °C
0.1 T g,resin = - 35 °C Td,2,EVA
Td,2,resin
0.0 Td,3,resinTd, 3,EVA
K.P
-50 -40 -30 -20 -10 0 10 20 30 40 50 60 70 80 90 100 110

Fig. 11.4 Coupling parameter, κ, in the case of resin and EVA interlayer materials in
temperature range from glass transition temperatures, Tg, to delamination temperatures, Td,
of laminated glasses consisting of two or three glass layers (Pankhardt, 2010c).
t by
Table 11.2 summarises the equations for calculating the coupling parameter, κ, between the
testing temperature ranges. The values for the glass transition temperature and delamination
temperature ranges are indicated in Table 11.1.
Table 11.2 Calculation of coupling parameter, κ, at service temperature, Tservice, of laminated
h
glasses consisting of two or three glass layers with 6 mm thickness and with EVA or resin
yrig
interlayer material
Type of Number of Service temperature Calculation of
interlayer glass layers, n ranges, Tservice coupling parameter, κ Eq.
material pcs °C −
cop
Tg ≤ T ≤ + 20 °C 1
2
+ 20 °C < T ≤ Td 2
-0.0001⋅T + 0.0001⋅T +1.0387 (11.6)
EVA
Tg ≤ T ≤ - 10 °C 1
3
- 10 °C < T ≤ Td 2
-0.0001⋅T + 0.0018⋅T + 0.9752 (11.7)
2 Tg ≤ T ≤ Td -0.0074⋅T + 0.7407 (11.8)
Resin
3 Tg ≤ T ≤ Td -0.0083⋅T + 0.7083 (11.9)
The modified Krüger (1998) type relationship, mentioned above, is recommended to use for
approximative calculations of flexural stiffness of plate shaped (htot << B) laminated glasses
supported at two edges, with a relative high ratio of Ls/htot ( > 50). Usually it is acceptable that
analysis based on classical beam theory (with four-point bending) can neglect the effect of
shear deformation by deflection (ASTM D790-99) for high ratio of Ls/htot. This is confirmed
68
Flexural stiffness
also by the researches of Whitney (1987) (see Fig. B.10.21). With the temperature dependent
parameter, κ, the possibility of determination of the flexural resistance at a given deflection is
provided in the application temperature range. In the case of known EIT,service, the maximal
force is predictable for given deflections e.g. yallowed = Ls/200 (see an example in App. B.,
Chapter 11). In the future, further research is required for determination of the parameter, κ,
for other types of interlayer materials.
The value of the glass transition temperature, Tg, and the value of HDT (Heat Distortion
Temperature, see App. D) of the interlayer material should be taken into account to load
bearing glass calculations. The delamination temperature, Td, should be determined
experimentally, where further influencing factors can be taken into account, e.g. TTS
DT
(Chapter 4, Eq. (4.5)). Therefore, delamination temperature, Td, is not equal to the melting
temperature, Tm, of the interlayer material, but Tm also influences Td.
Experimentally results have shown (Chapter 10) that the edge regions of laminated
AR
glasses are the most temperature sensitive. The deformation of laminated glasses in the
middle (R 1) and edge (R 2) regions cannot be considered equal.
Experimentally I have shown that the edge region of laminated glasses with EVA foil is
less temperature sensitive than that of laminated glasses with resin. The effect of
KH
temperature on the edge region of laminated glass panes in approaching calculations of
bending strength should be taken into account with a reducing factor (k1). Based on my
laboratory results, the factor k1 versus temperature was given for laminated glass with
AN
EVA foil or resin (Pankhardt & Balázs, 2006; Pankhardt, 2008a; Pankhardt, 2008b;
Pankhardt & Balázs, 2010b), [3.3 New Scientific Result: Effect of temperature on ultimate
strain of laminated glasses].
1.00
K.P
Resin
0.98 0.97 0.97
EVA
0.96 with EVA
0.94
0.94 0.93 0.93 0.93 20 mm/min
reducing factor, k1 (-)
0.92 0.92
0.90
t by
with RESIN
0.88 0.87
0.86
0.84
Tg, resin Tg, EVA Td, EVA
0.82 Td,resin
h
0.81 0.81
0.80
yrig
-40 -30 -20 -10 0 10 20 30 40 50 60 70 80 90 100

Fig. 11.5 Reducing factor k1, versus temperature of laminated glass with EVA or resin
interlayer material (Pankhardt, 2010b)
The temperature dependent value of factor k1 is recommended to take into account the
cop
thermo-mechanical behaviour of the interlayer material. The upper limit of k1 should be

considered when the interlayer material ensures the appropriate bond, which is influenced by
the glass transition temperature, Tg, of the interlayer material. The glass transition
temperature, Tg, of EVA is -28 °C and for resin it is about -38 °C. Based on the results of
laminated glasses with temperature of -20 °C, the upper limit of k1 can be estimated with the
relationship in Fig. 11.5.
The lower limit of k1 is considered based on the results of laminated glass with a spacer
(without bond). The lower limit shows the difference of stresses between the middle region
(R 1) and edge region (R 2) of a glass pane when the interlayer material cannot ensure
appropriate bond between the glass layers (due to softening of the interlayer or durability
reasons e.g. reduced adhesion). The lower limit is also influenced by the melting temperature,
69
Flexural stiffness
Tm, of the interlayer material. Based on my laboratory results I determined the value of 0.81
for the lower limit of k1, (see also Chapter 9).
A ke factor for single glasses in EN 1288-3:2000 was already determined (Chapter 6.3.2).
In the case of laminated glass, the strength of the edge region (Region 2) is influenced by the
bond behaviour at service temperature of the interlayer material (Chapter 10.2.2.4).
Therefore, the Author defined factor kservice, to calculate the modified bending stress (σb,service)
of Region 2 of laminated glass at service temperature, as:
σb,service,Re gion2 ⋅ kservice = σb,service,Re gion1 (11.10)
- kservice=k1×k2×k3
DT
k1: reducing factor of edge strength (Region 2) at service temperature (Fig. 11.5),
k2: influence of different types of edgework (Future work); k3: influence of loading
rate, etc. (Future work).
If ultimate force and flexural stiffness at service temperature are known (Figs. 11.1 and 11.2)
AR
and the stresses in the outer glass layer were measured (Chapter 9), the effective thickness
(Table B.11.1) and, therefore, the effective Young’s modulus (Table B.11.2) of laminated
glass at service temperature can be determined (see also Figs. B.11.3 and B.11.4). This
KH
temperature dependent effective thickness and Young’s modulus of laminated glass can also
be used in FEM calculations (see example in App. B, Chapter 11). Figs. B.11.3 and B.11.4
indicate the change of effective Young’s modulus of laminated glasses in the temperature
range of -20 °C to +60 °C. With increasing temperature, the shear stiffness rapidly decreases
AN
and creep becomes more significant (Krüger, 1998). Young’s modulus (Et0) depending on
time (see also Fig. A.4.19) or load, e.g. for constant loads, is (Palotás & Balázs, 1980):
E
Et0 = 0 (11.13)
1+ ϕ
K.P
where, E0 is Young’s Modulus for short term loads and ϕ is the creep parameter.
Relaxation is also a well-known phenomenon of plastic materials (Eq. 4.2), which influences
the bending characteristics of laminated glass. The aging of the interlayer material acts on its
bond capacity, etc., therefore, these effects can be defined in the coupling parameter, κ. In the
t by
case of increasing temperature or lifetime predictions on the strength of laminated glass, the
temperature dependent flexural stiffness (Dfl,Tservice) should be studied. Thereby, the influence
of temperature on bond capacity can be taken into account and also the membrane effect
(Vallabhan et al., 1987) from increasing deflections depending on support conditions (Figs.
h
B.11.7 and B.11.8). Further investigations should be done on the change of Poisson’s ratio of
laminated glass with the effective thickness.
yrig
11.3 Exposure classes of laminated glasses

Strength tests of load bearing laminated glasses are suggested to be carried out (besides at
room temperature) at the temperature range according their exposure class (Fig. 11.8). For the
cop
temperature dependent exposure classes of laminated glasses the interlayer properties at least
with Tg and Tm, HDT (see App. D) should be investigated as well.
Based on my laboratory results, I suggest the classification of laminated glasses into
exposure classes. Based on my laboratory results, I have defined the temperature
dependent exposure class (XT). For determination of the temperature dependent
exposure class (XT1 to XT 5), I suggest taking into account the thermo-mechanical
behaviour of the interlayer materials (Pankhardt & Balázs, 2006; Pankhardt, 2008b;
Pankhardt & Balázs, 2010b; Pankhardt, 2010c; Pankhardt, 2010e), [3.4 New Scientific
Result: Exposure classes of laminated glasses]. The exposure classes of laminated glasses can
be determined from the temperature and time (TTS) dependent behaviour of interlayer
material as indicated in Fig. 11.8.
70
Flexural stiffness
DT
AR
Fig. 11.8 Suggested exposure classes at service temperature (from XT1 to XT5) of laminated
glasses dependent on interlayer characteristics (see also Figs. A.4.17 and A.4.19.)
KH
To take into account the service temperature and loading condition, e.g. loading rate, the
exposure class should be determined. The exposure classes are dependent on the type of
polymer if it is e.g. crystalline or amorphous (Chapter 4.3.2). Note that there is no melting
AN
behaviour in the amorphous polymers, therefore, no class XT5 exists for it. The testing
temperature of laminated glasses should be determined according to the service temperature
range (Pankhardt, 2008b).
The development of such an interlayer material is suggested for laminated glasses,
K.P
especially for load bearing glasses, which can be safely applied in the given exposure class.
The new scientific results are confirmed by the development of interlayer materials which
have started to take into account the glass transition temperature e.g. laminated glass
windshields (Okamoto et al., 2009).
By giving the exposure classes (similar to concrete), it is easier for designers to determine
t by
the requirements on load bearing glasses knowing their application conditions (interior,
exterior) and the properties of acting forces (static, dynamic). For constructors it is easier to
select the most appropriate products. In the case of certain exposed exterior load bearing glass
slabs where the creep of the interlayer is significant, the exposure class of the glass pane can
h
be XT4 or XT5, and safety glasses impacted by dynamic loads applied in interior places can
yrig
be XT1 class. If the exposure class of the glazing is known – thereby the requirements are
better defined – the estimation of time of durability and costs can be more precise.
Comprehensive experimental investigation is required for the determination of further
exposure classes.
cop
11.4 Conclusions
Based on the experimental results the definition of delamination temperature, Td, was
introduced. The flexural stiffness of the overall laminate is influenced by temperature. With
increasing temperature, the flexural stiffness decreases. The tendency of decrease of flexural
stiffness at different temperatures is influenced by the type of interlayer material. The
relationship of temperature dependent flexural stiffness determined by Krüger (1998) was
modified. The parameter κ versus temperature was determined directly for EVA and resin
laminated glasses. The effect of temperature on the edge region of laminated glass panes in
approaching calculations of bending strength should be taken into account with a reducing
factor (k1). I suggested the classification of laminated glasses into exposure classes. With the
experimental results, the methods of calculation of laminated glasses can be more precise.
71
Residual load bearing capacity
Chapter 12
12. Residual load bearing capacity of laminated glass
panes
12.1 Introduction
A glazing element can fail during its service life due to various impacts. An interesting aspect
is the residual load bearing capacity after failure of one or more glass layers of multi-layered
laminated glass (post-failure behaviour), especially when it is used in overhead areas e.g.
roofing or in canopies as well as in slabs, where safety demands are high (Pankhardt, 2008b).
DT
The remaining load bearing capacity of glass is influenced by the interlayer material. This
chapter deals with the temperature dependent residual load bearing capacity of laminated
glasses.
AR
12.2 Fracture process of laminated glass in four-point bending
12.2.1 Force during fracture of laminated glass
Three different stages can be identified in the fracture of laminated glass. Fig. B.12.1
KH
indicates the fracture process (Stages A, B and C) in four-point bending of laminated glass
specimens consisting of three glass layers and plastic interlayer (see also Fig. 9.3). Figs.
B.12.2 and B.12.3 indicate the change of load bearing capacity of laminated glasses during the
fracture process. Laminated glasses with non-bonded layers and the ultimate force of single
AN
glass layers with thicknesses 12 mm and 19 mm for comparison are also indicated. It is
indicated that the ultimate force in glass layers of laminated glass is influenced by the
temperature. By comparing the columns of the same type of laminated glass at temperatures
K.P
of -20 °C, +23 °C and +60 °C it can be seen that the ultimate force of the remaining layers
decreases with the increase of temperature. EVA foil is less temperature sensitive than resin
interlayer material in both float and tempered laminated glasses, which is indicated by the
change of force transfer capacity (Stage C1 and C2), Fig. B.12.4 (Pankhardt, 2010b).
12.2.1.1 Effect of number of glass layers
t by
To study the force transfer capacity, the columns of Figs. B.12.2 and B.12.3 are further
divided into parts from No. 1 to No. 5 (Fig. B.12.1).
7.50
Change of force during fracture process,
7.00 5 Fu3-Fu2
h
6.50 20 mm/min
6.00 4 Fu2-Fu23
yrig
5.50
5.00 3 Fu23-Fu1 (Fu2-Fu1)
4.50 5
5 5
∆F (kN)
4.00 2 Fu1-Fu12
1 5
3.50 5
3.00 1 Fu12 (Fu1)
4 4
cop
4 4 5
2.50 4
2.00 4
3 3 3 3 3 3
1.50 1 3 3 3 3 5
3 3 4
1.00 2
2 2 2 2 2 23 2 2 2 2 2 3
2
0.50 1 2
1 1 1 1 1
1 1 1 1 1 1 1 1 1 1 1
0.00
-20°C
23°C
60°C
23°C
-20°C
23°C
60°C
-20°C
23°C
60°C
23°C
23°C
-20°C
23°C
60°C
-20°C
23°C
60°C
23°C
6mm 12mm F_2_R F_2_EVA F_2_D 19mm F_3_R F_3_EVA F_3_D

Fig. 12.1 Change of force during fracture process of laminated float glasses at temperatures
of -20 °C, +23 °C and +60 °C, (Pankhardt, 2010b).
Figs. 12.1 and B.12.5 indicate that the ultimate force of laminated glass consisting of two
glass layers is at least two times that of a single glass layer (when the appropriate bond is
ensured, at temperature of -20 °C and can reach the value of monolithic single glass with a
72
thickness of 12 mm. In the case of laminated glass consisting of three glass layers, the flexural
stiffness decreases with the increase of the total thickness of interlayer material. Therefore,
the ultimate force of a single glass layer with a thickness of 19 mm can only be reached when
all glass layers are tempered and an appropriate bond is ensured. Parts of the columns No. 2
and No. 4 indicate the increase of force with the help of the interlayer material during the
fracture process by Stages C1 and C2 (see Fig. B.12.4). In the case of laminated glass
consisting of three glass layers, part No. 4 of the overall column is about two times that of part
No. 2. Therefore, before the fracture of one glass layer, both interlayer are able to transfer
forces.
12.2.1.2 Effect of tempering on residual load bearing capacity of
DT
laminated glass pane
In safety laminated glasses each layer contains a compression layer near its surface, which has
a favourable effect by closing cracks developed from surface scratches or edgework. As
AR
mentioned in Chapter 8, the effectiveness of tempering decreases with the increase of glass
thickness. Figs. B.12.3 and B.12.5 indicate that the ultimate force of laminated glass
consisting of three glass layers can reach or exceed the ultimate force of tempered monolithic
KH
single glass with a thickness of 19 mm. When appropriate bond is ensured at service
temperature, in the case of reaching the ultimate force of laminated glass consisting of three
glass layers, during the fracture process the force drops to a level higher than the ultimate
force of laminated glass consisting of two glass layers.
AN
12.2.1.3 Effect of temperature sensitivity of interlayer material on
residual load bearing capacity
The residual load bearing capacity of laminated glass during the fracture process is influenced
K.P
by the temperature sensitivity of the interlayer material. Figs. 12.1 and B.12.5 indicate the
decrease of bond capacity of the interlayer material (parts No. 4 and No. 2 of columns).
If one glass layer of three-layer laminated glass fractures at 23 °C, the residual load bearing
capacity is higher than that of two-layer laminated glass, because the interlayer on the top of
t by
the fractured glass layer works as a kind of strengthening (see Fig. B.12.5, dotted line and
columns E_2_R and E_3_R at +60 °C). The drop in load bearing of laminated glasses with
resin interlayer is significant with the increase of the temperature from -20 °C to +60 °C.
However, the temperature sensitivity of laminated glasses with EVA interlayer is less
h
pronounced, there are small drops in load bearing both with reduce or increase of the
temperature compared to +23 °C. The effect of temperature on the residual load bearing
yrig
capacity decreases during the fracture process. The fracture pattern of laminated glass is
influenced by temperature (Chapter 10). The cracks are denser with the increase of
temperature, thus influence the bond capacity of the interlayer material, therefore, the residual
load bearing capacity (App. B., Figs. B.10.3 to B.10.13). The interlayer can transfer forces of
cop
larger non-fractured regions between the glass layer surfaces. In the case of EVA interlayer
material, the fracture pattern is not significantly influenced by temperature, therefore, parts
No. 2 and No. 4 of the columns in Figs. 12.1 and B.12.3 are not significantly influenced by
temperature. The results were also quantitatively analysed. The ratio of residual force, Fb, to
maximal force, Fa (see Fig. B.12.4) by fracture of a glass layer of laminated glass can be
expressed with Eq. B.12.1. Table B.12.1 and Fig. B.12.5 summarise the ratio of residual force,
Fb, to maximal force, Fa.
12.2.2 External work of laminated glasses
Based on the experimental results, it was determined that the residual load bearing capacity of
laminated glasses can be favourably influenced with the increase of the number of glass
layers, when the glass layers are appropriately bonded with interlayer materials. The areas of
73
the force vs. deflection curve indicated in Fig. 12.2 were calculated (Fig. B.12.8) to study the
effect of the number of glass layers on the change of external work during the fracture process
of laminated glasses. The change of the external work can be expressed relative to the results
with room temperature.
450 No.3 glass layer
No.2 glass layer
400
116.33
No. 3.
No.1 glass layer No. 2.
No. 1.
350
156.31
External work, Nm
108.02
20 mm/min
129.54
300
130.03
134.21
250
181.61
DT
200
128.03
131.27
156.04
105.68
140.49
112.02
150
106.38
120.83
23°C 48.59 28.02 42.24

101.63
101.41
97.67
191.90
183.22
AR
100
131.36
23°C 47.7240.38
102.09
90.55
82.41
73.87
72.62
71.74
67.22
66.73
50
64.58
63.09
63.28
23°C 51.31
60°C 47.41
-20°C 46.16
0
23°C
-20°C
23°C
60°C
-20°C
23°C
60°C
23°C
-20°C
23°C
60°C
-20°C
23°C
60°C
KH
E_6mm E_12mm E_2_R E_2_EVA E_ 2_D E_19mm E_3_R E_3_EVA E_3_D
Fig. 12.2 External work during the fracture process of laminated tempered glasses at -20 °C,
+23 °C and +60 °C. Single layer glasses are indicated with their thickness: 6 mm, 12 mm and
AN
19 mm.
I have determined that the relative external work of safety laminated glasses in bending
decreases in the case of resin interlayer material with the increase of the number of glass
layers (from two to three) and increase of the temperature (from +23 °C to +60 °C), but
K.P
it increases in the case of EVA interlayer material (Pankhardt & Balázs, 2006; Pankhardt,
2008b; Pankhardt & Balázs, 2010b), [5.2 New Scientific Result: Residual load bearing
capacity of laminated glasses as function of temperature]. The relative external work of
laminated glasses with three tempered glass layers and resin interlayer decreases by 12.4 %,
but it increases by 25.1 % with the use of EVA foil interlayer with the increase of temperature
t by
from +23 °C to +60 °C. With the decrease of the temperature, the interlayer material behaves
rigidly, the Young’s modulus of resin increases more than of EVA (App. C, Figs. C.2.9 and
C.2.13) and cracks can form in the interlayer material (App. C, Figs. C.3.8 and C.2.17).
Because of the favourable effect of tempering by closing cracks, and thus on the external
h
work during the fracture process with appropriately bonded layers, the external work of safety
yrig
laminated glasses can be higher in bending than that of single glasses with equal thickness.
Load bearing capacity of float glass is influenced by a single critical defect (Chapter 8).
Therefore, with the increase of the number of glass layers, the external work of non-safety
laminated glasses (Fig. B.12.8) in bending is not higher than that of single glasses with equal
thickness. The external work of spacer laminated glass is considerably lower (for two glass
cop
layers: by 51.9 %; for three glass layers: by 38.1 %) than that of single glasses with equal
thickness, and proportionally increases with the increase of the number of glass layers. The
external work is also influenced by the structure of the glass layers in the laminated glass. In
the case of bending, the position from the neutral axis influences the effectiveness both of the
interlayer material and the glass layers (Chapter 9.2.2.1).
12.3 Conclusions
The residual load bearing capacity during the fracture process of laminated glass is influenced
by the interlayer material, which is influenced by temperature. The effect of temperature on
the residual load bearing capacity of laminated glass decreases during the fracture process.
The residual load bearing capacity of laminated glass is influenced by the fracture pattern and
the secondary cohesion effect of the interlayer.
74
Future work
Chapter 13
13. Future work
The effect of temperature on load bearing capacity of laminated glass is an important factor
and until now pure investigations were done on it. Therefore, the durability of the building
materials used in laminated glass will play an important role. In the future, the exposure class
of glazing elements should be determined.
New fields of research based on the laboratory results are: experiments with different loading
rates (e.g. short impact, long term loads); durability questions: study of further reducing
DT
factors (k) on edge strength of laminated glasses, cyclic loading; strain measurement e.g.
regions of bore holes; heat strengthened glasses, further interlayer materials (determination of
further parameter κ) and non symmetric layered laminated glasses; further study of the
influence of the tension stiffening effect.
AR
The development of digital, photoelastic testing under loaded conditions on a reflexive basis
is planned. The procedure can be applied as a non-destructive testing method for load bearing
glasses. The results of my measurements are partially analysed with my own developed
KH
software called StressOptic (which should be developed further) and some results are already
published, App. B. Figs. B.13.1 to B.13.3 (Pankhardt, 2008a).
Architectural tendency is to use larger size glass panes or to expose it to carry high loads, e.g.
glass slabs to carry loads of cars, therefore, strengthening of the glass pane is needed to avoid
AN
too large deflections. For external strengthening of glass panes, relatively lightweight
materials with high strength (e.g. fibre reinforced polymers) are needed (internal
strengthening is prestressing e.g. tempering). Therefore, the Author started to continue testing
K.P
by applying carbon stripes and self-designed carbon T-profiles for strengthening single and
laminated glasses (App. B. Figs. B.13.4. and B.13.5).
h t by
yrig
cop
75
K. Pankhardt Load bearing glasses Summary
Summary of New Scientific Results
1. New Scientific Results: Effect of tempering (heat-treatment) on load

bearing capacity of single layer glasses in bending
1.1 New Scientific Result: The effectiveness of tempering (Pankhardt et al., 2006;
Pankhardt et al., 2010a)
I have introduced the definition effectiveness of tempering (heat treatment) in the case of
DT
tempered glasses with different thickness to show the effectiveness of tempering on the
increase of load bearing capacity. The effectiveness of tempering shows the proportion of
load bearing properties (e.g. maximal force) of tempered glasses compared to non heat-treated
float glasses with the same thickness. It was experimentally shown that the effectiveness of
AR
tempering depends on the glass thickness and the loading rate.
1.2 New Scientific Result: Bending strength of single layer glasses (Pankhardt, 2008a;
Pankhardt, 2008b; Pankhardt et al., 2010a)
KH
Based on my laboratory bending tests I have shown that the values are different for bending
strength calculated with the basic formula of elasticity given in the standard EN 1288:3-2004
and the maximal value of surface stresses of glass calculated from the result of strain
AN
measurement. With the increase of the glass thickness, the bending strength decreases
compared to the surface strength. The ratio of bending strength to surface strength determined
with strain measurement can be approximated with power functions both in the case of
tempered and non heat-treated float glasses.
K.P
1.3 New Scientific Result: Ultimate strains measured on surface (Pankhardt et al., 2006;
Pankhardt, 2008a; Pankhardt, 2008b; Pankhardt et al., 2010a)
Based on my laboratory results, I have shown that the value of the multiplying factor (ke) is
influenced by thickness and by the loading rate. The difference of the strains on the surface of
t by
the glass layer in the edge and mid regions increases with the reduction of the loading rate
from 50 mm/min to 20 mm/min.
h
2. New Scientific Results: Effect of lamination on bending behaviour of

yrig
symmetrical layered laminated glasses consisting of glass layers with equal

thickness
2.1 New Scientific Result: Effectiveness of tempering (heat-treatment) in the case of
laminated glasses (Pankhardt, 2008b; Pankhardt et al., 2010b)
cop
Experimentally I have demonstrated that the effectiveness of tempering decreases with the
increase of the number of glass layers for symmetrical layered (consisting of glass layers with
equal thickness) laminated safety glasses, when the layers work together. I have shown for
glasses laminated with a spacer (without interlayer material) that the effectiveness of
tempering no decreases with the increase of the number of glass layers from two to three.
Furthermore, I have shown that at room temperature the effectiveness of tempering is more
favourable in the case of resin laminated glasses than in the case of EVA foil laminated
glasses.
S 1/3
2.2 New Scientific Result: Effect of interlayer material on ultimate strains of glass
(Pankhardt et al., 2006; Pankhardt, 2008a; Pankhardt, 2008b; Pankhardt et al., 2010b)
Experimentally I have shown that the ultimate deformation (strain) in the region near the edge
is 17 to 19% higher in the case of laminated float and tempered glasses without interlayer
material (only with spacer) and at temperature of +23 °C than that measured in the middle of
the glass surface. Furthermore I have shown that the deformations (strains) measured in the
edge region of the glass surface exceed by 5 to 12 % the value measured in the middle region
in the case of laminated glasses with EVA foil interlayer and temperature of +23 °C. This
difference is only 2 to 5 % with resin interlayer material.
DT
3. New Scientific Results: Effect of temperature on bending behaviour of
symmetrical layered laminated glasses consisting of glass layers with equal
AR
thickness
3.1 New Scientific Result: Temperature sensitivity of laminated glasses. Change of force
versus deflection relationship with the effect of temperature (Pankhardt et al., 2006;
KH
Pankhardt, 2008b; Pankhardt et al., 2010b; Pankhardt, 2010e)
Based on my laboratory results I have introduced the definition temperature sensitivity of
laminated glasses to show the change of load bearing properties of laminated glasses with the
AN
change of temperature. The definition of temperature sensitivity indicates the magnitude of
the change of a property (e.g. load bearing capacity) of laminated glass with the change of
temperature. I have shown that laminated glasses with EVA foil are less temperature sensitive
than glasses laminated with resin.
K.P
3.2 New Scientific Result: Effect of number of glass layers with different temperature on
load bearing properties (Pankhardt et al., 2006; Pankhardt, 2008b; Pankhardt et al.,
2010b; Pankhardt, 2010c)
Experimentally I have shown that the temperature sensitivity of laminated glasses with resin
t by
increases, with the increase of the number of glass layers. According to my laboratory results,
laminated glasses with EVA foil are less temperature sensitive than laminated glasses with
resin foil, with the increase of the number of glass layers.
3.3 New Scientific Result: Effect of temperature on ultimate strain of laminated glasses
h
(Pankhardt et al., 2006; Pankhardt, 2008a; Pankhardt, 2008b; Pankhardt et al., 2010b)
yrig
Experimentally I have shown that the edge region of laminated glasses with EVA foil is less
temperature sensitive than that of laminated glasses with resin. The effect of temperature on
the edge region of laminated glass panes in approaching calculations of bending strength
should be taken into account with a reducing factor (k1). Based on my laboratory results, the
cop
factor k1 versus temperature was given for laminated glass with EVA foil or resin.
3.4 New Scientific Result: Exposure classes of laminated glasses (Pankhardt et al., 2006;
Pankhardt, 2008b; Pankhardt et al., 2010b; Pankhardt, 2010c; Pankhardt, 2010e)
Based on my laboratory results, I suggest the classification of laminated glasses into exposure
classes. Based on my laboratory results, I have defined the temperature dependent exposure
class (XT). For determination of the temperature dependent exposure class (XT1 to XT5), I
suggest taking into account the thermo-mechanical behaviour of the interlayer materials.
S 2/3
4. New Scientific Results: Change of bending stiffness of laminated glasses as

function of temperature
4.1 New Scientific Result: Introduction of definition of delamination temperature
(Pankhardt et al., 2010b; Pankhardt, 2010c)
Based on my experimental results, I have introduced the definition of delamination
temperature (Td). The delamination temperature gives the temperature (temperature range) at
which the interlayer material cannot appropriately connect the glass layers and cannot ensure
the bond strength. Based on my experimental results, I have determined the delamination
temperature for glasses laminated with EVA foil or resin and consisting of two or three glass
DT
layers, in bending. Based on my experimental results, I have shown that the delamination
temperature decreases with the increase of the number of glass layers.
4.2 New Scientific Result: Change of bending stiffness of laminated glasses as function of
AR
temperature (Pankhardt et al., 2006; Pankhardt, 2008b; Pankhardt et al., 2010b;
Pankhardt, 2010c)
Based on my laboratory results, the flexural stiffness of safety or non-safety laminated glasses
with resin, consisting of two and three glass layers, decreases linearly with the increase of the
KH
temperature in a range between -20 °C and +60 °C. The flexural stiffness of safety or non
safety laminated glass with EVA foil, consisting of two and three glass layers, can be fitted
with a second-degree polynomial function with the best correlation in a temperature range
AN
between -20 °C and +60 °C.
4.3 New Scientific Result: Approaching calculation of temperature dependent bending
stiffness of laminated glasses (Pankhardt, 2008b; Pankhardt et al., 2010b; Pankhardt,
K.P
2010c)
Based on my experimental results, I have modified the relationship of temperature dependent
flexural stiffness determined by Krüger (1998). I have determined parameter κ for laminated
glasses with resin or EVA foil, consisting of two and three glass layers, at temperatures
between the glass transition temperature and the delamination temperature range.
t by
5. New Scientific Results: Residual load bearing capacity of laminated glasses

as function of temperature
h
yrig
5.1 New Scientific Result: Effect of temperature on fragmentation pattern of laminated

glasses (Pankhardt, 2008b; Pankhardt et al., 2010b; Pankhardt, 2010c)
According to my experimental results, the secondary cohesion effect of the resin interlayer
validates during the development of cracks on the tensioned side of laminated glasses at low
(-20 °C) temperature. During the fracture process of safety glasses, glass fragments can build
cop
partially connected regions. In this region the bond is favourable compared to single
fragments. As a result of this effect, the deflection of the glass will be reduced under further
loads.
5.2 New Scientific Result: Residual load bearing capacity of laminated glasses as function
of temperature (Pankhardt et al., 2006; Pankhardt, 2008b; Pankhardt et al., 2010b)
I have determined that the relative external work of safety laminated glasses in bending
decreases in the case of resin interlayer material with the increase of the number of glass
layers (from two to three) and increase of the temperature (from +23 °C to +60 °C), but it
increases in the case of EVA interlayer material.
S 3/3
K. Pankhardt Load bearing glasses References
References
References
AGC (2007), All about glass, Flat Glass Europe, AGC Communications Dept., Brussels, p. 25.
Aizenberg, J., Weaver, J.C., Thanawala, M.S., Sundar, V.C., Morse, D.E. and Fratzl P. (2005),
“Skeleton of Euplectella sp.: Structural Hierarchy from the Nanoscale to the Macroscale”
Science, Vol.309, pp. 275-278.
Andrade E.N. (1914), “The flow in metals under large constant stresses,” Proceedings of the Royal
Society, Section A, London, Vol.90, pp. 329-342. doi:10.1098/rspa.1914.0056
DT
Arora, A., Marshall, D.B., Lawn, B.R., Swain, M.V (1979), „Indentation deformation/fracture of
normal and anomalous glasses”, Journal of Non-Crystalline Solids, Vol.31., North-Holland
publishing company, pp. 415-428.
Balázs, Gy. (1984), Építőanyagok és kémia, - Building materials and chemistry, Tankönyvkiadó,
AR
Budapest, ISBN 963 17 79564
Barenblatt, G.I. (1962), “The mathematical theory of equilibrium cracks in brittle fracture”, Adv. Appl.
Mech., Vol.7, pp.55-129.
de la Bastie, F.B.A.R. (1874), Tempering glass, British Patent 2783
KH
Bažant Z. P., (2004), „Scaling theory for quasibrittle structural failure” PNAS, Sept. 14, 2004, Vol.
101, No. 37, pp.13400–13407, doi_10.1073_pnas.0404096101
Behr, R.A., Minor, J.E., Linden, M.P. and Vallabhan, C.V.G. (1985), „Laminated glass units under
uniform lateral pressure”, Journal Struct. Eng., Vol.111, No.5, pp. 1037-1050.
AN
Benedictus, E. (1910), British Patent 1.790
Bene, Zs., Pankhardt, K. (2010d), „Építési üvegek tisztasági kérdései”, („Questions about cleanness of
building glasses”), Műszaki Tudomány az Észak-Alföldi Régióban 2010, MTEAR 2010
Conference presentations, editor: L. Pokorádi, Nyíregyháza, 19. May 2010, DAB Debrecen
K.P
2010, Hungary, ISBN 978-963-7064-24-1, pp. 301-306.

Block, V., Davies, P.S. (2005), “Enhanced edge stability with structural glass laminates” Laminated
glass process and trends, Glass processing days 2005, pp. 1-2.
Brandrup J., Immergut E. H., Grulke E. A. (1989), Polymer Handbook, Vols. 1-2, 4th ed., Wiley-
Interscience, (Ed. Abe A., Bloch, D.R.) New York, U.S.A., ISBN 0-471-47936-5
t by
Brostow, W., D’Souza, N.A., Kubát, J. and Maksimov, R. (1999), “Creep and stress relaxation in a
longitudinal polymer liquid crystal: prediction of the temperature shift factor” Journal of
Chemical Physics, Vol.110, No.19, American Institute of Physics, pp. 9706-9712.
Cannon, D.D., Musso, C.S., Williams, J.C. and Eagar, T.W. (2004), “Analysis of brittle fracture of
h
soda glass bottles under hydrostatic pressure”, ASM International 1547-7029, Journal of Failure
Analysis and Prevention, Vol.4, No.5, pp. 72-77., DOI: 10.1361/15477020420800
yrig
Czarnecki, S., Kursa, M., Lewinski, T. (2008), „Sandwich plates of minimal compliance” Computer
methods in applied mechanics and engineering, Vol. 197, pp. 4866–4881,
doi:10.1016/j.cma.2008.07.005
Chuang, T.J., Fuller Jr., E.R. (1992), “Extended Charles–Hillig theory for stress corrosion cracking of
cop
glass”, Journal of the American Ceramic Society, Vol.75, No.3, pp. 540-545.
Dabbs, T.P., Lawn, B.R. and Kelly P.L. (1982), “A dynamic fatigue study of soda-lime and
borosilicate glasses using small-scale indentation flaws” Phys. Chem. Glasses, Vol.23, p. 58
Dabbs, T.P., Fairbanks, C.J. and Lawn, B.R. (1984), „Subthreshold indentation flaws in the study of
fatigue properties of ultrahigh-strength glass”, Methods for assessing the structural reliability of
brittle materials, ASTM STP 844, (Eds.) Freiman S.W., Hudson C.M., ASTM Philadelphia, pp.
142-153.
Dharani, L.R., Wei, J., Yu, J., Minor, J.E., Behr, R.A. and Kremer P.A. (2005), “Laminated
architectural glass subjected to blast, impact loading”, The Glass Researcher, American Ceramic
Society Bulletin, Vol.84, No.1, pp. 42-44.
Denoyer, J., Pollack, H. (1963), “Measurements of the velocity of crack propagation in glass plates”,
Bulletin of the Seismological Society of America, Vol.53, pp. 87-93.
xi
Emonds et al., (1999) “How the Performance of Diamond Tools in Glass Grinding can be Influenced
by Optimising the Coolant System“, GPD - 1999 Conference, Tampere, Finland
Feingold, J.M., Redner, A.S. (2003), “Stress relief– new Pc-based scanners improve quality and
productivity for glass fabricators” International Glass Review, Vol.1, pp. 63-66.
Ferry, J.D., Child Jr., W.C., Zand, R., Stern, D.M., Williams, M.L. and Landel, R.F. (1957), “Dynamic
mechanical properties of polyethyl methacrylate” Journal of Colloid Science, Vol.12, No.1, pp.
53-67.
Fuller Jr., E.R., Lawn, B.R. and Thomson, R.M. (1980), “Atomic modelling of chemical interactions at
crack tips” Acta Metallurgica, Pergamon Press Ltd., Vol.28. pp. 1407-1414.
Gehrke, E., Ullner, Ch., and M. Hähnert (1991), “Fatigue limit and crack arrest in alkali-containing
silicate glasses”, J. Mater. Sci., Vol. 26, pp. 5445–5455.
DT
Glaesemann, G.S. (1991), „Optical fiber failure probability predictions from long-length strength
distributions”, Proc. of the 40th International wire and cable symposium, St. Louis, Mo. pp. 819-
825.
Glaesemann, G.S., Walter D.J. (1991), „Method for obtaining long-length strength distributions for
AR
reliability prediction”, Optical Engineering, Vol.30, No.6, Corning, New York, pp.746-748.
Glathart, J.L., Preston, F.W. (1968), „The behaviour of glass under impact: theoretical considerations”
Glass Technology, Vol. 9, pp. 89-100.
KH
Glick, S.H., Pern, F.J., Watson, G.L., Tomek, D. and Raaff J. (2001), „Performance degradation of
encapsulated monocrystalline-si solar cells upon accelerated weathering exposures” Proceedings
of NCPV Program Review Meeting, Lakewood CO, pp. 307-308.
Glick, S.H., Pern, F.J., (2003), “Adhesion strength study of EVA encapsulants on glass substrates”,
AN
NCPV and Solar Program Review Meeting 2003, NREL/CD-520-33586, p. 942.
Greaves, N. (2005), “Glass in class”, 40.IOP Publishing Ltd, Specialfeature: Materials, Physics
Education, pp. 422-429.
Gulati, S.T. (1966), “Delayed cracking in automotive windshields”, Materials Science Forum, Vols.
K.P
210-213, Transtec Publications, Switzerland, pp. 415-424.

Gulati, S., Roe, T., Vitkala, J. (2001), “Importance of edge finish on thermal tempering”, Glass
Processing Days 2001 pp. 72-78, Glass Processing Days 2005, pp. 422-429.
Häuser, K., Kramer, B., Schmidt, RW., Walk, R. (2002), „Design with glass“, Interpane, Vol.6.,
Interpane, Lauenförde, p. 224.
Hegedűs, I. (1983), „Szendvicsszerkezetek szilárdsági és alakváltozási jellemzői“ – “Strength and
t by
deformation characteristics of sandwich structures”, Építőanyag praktikum, Ch. 14, (Ed.)

Balázs, Gy., Műszaki könyvkiadó, Budapest, p. 690.
Hertz Jr., D.L., (1979) “Mechanics of elastomers at high temperatures” presented at the High
temperature electronics and instrumentation Seminar, Houston, Texas, Dec. 3-4, 1979. pp. 2-14.
h
Hess, R. (1986), “Glasdickenbemessung: Bemessung von Einfach- und Isolierverglasungen unter

Anwendung der Membranwirkung bei Rechteckplatten grosser Durchbiegung” – “Calculation of
yrig
glass thickness: Calculation of single and of thermal-insulation glass unit with application of
membrane effect by large deflection of rectangular shaped plates” Institut für Hochbautechnik,
ETH Zürich
Hillig, W.B., Charles R.J. (1965), High-Strength Materials, (Ed.) Zackey, V.F., Wiley, New York, pp.
cop
682-705.
Hooper, J.A. (1973),”On the bending of architectural laminated glass” International Journal of
Mechanical Sciences, Vol.15, No.4, pp. 309-323.
Horváth, I., Szabó, I. (2003), "Feszültségmérési módszerek alkalmazása az üveggyártásban” –
“Application of strength measurement methods in glass manufacture”, Építőanyag, Vol.55,
No.4, pp. 141-150.
Hrma, P. (2006), “High-temperature viscosity of commercial glasses”, Ceramics-Silikáty Vol.50, No.2,
pp. 57-66.
Hussein, A.I, Iqbal, M.H. and Al-Abdul-Wahhab, H.I. (2005) „Influence of Mw of LDPE and vinyl
acetate content of EVA on the rheology of polymer modified asphalt” Rheol. Acta Vol.45,
Springler, pp. 92–104.
Ito, S., Tomozawa, M. (1982), “Crack blunting of high silica glass” Journal of the American Ceramic
Society, Vol.65, No.8, pp. 368-371.
xii
Jodidio, P. (2001), Architecture Now!, Vol.1. Taschen, Köln, p.75., pp. 134-135., p.403., 545. ISBN
3-8228-6065-4
Jodidio, P. (2002), Architecture Now!, Vol.2., Taschen, Köln, pp. 68-71., pp.182-185., p.499. ISBN 3-
8228-1594-2
Jofeh, C., Perry, A. (2005), „Sea of glass ‘Refloating’ Brunel’s ss Great Britain”, The Arup Journal,
No.3, pp.57-58.
Kadri, I., Lambrechts, P. and Cornelis, A. (2003), “Adding freedom to contemporary architecture with
resin glass laminates”, Glass processing days 2003, Poster 6, pp. 444-448.
Klindt, L.B., Klein, W. (1981), Az üveg mint építőanyag, - Glass as building material, Műszaki
Könyvkiadó, Budapest
Krüger, G. (1998), “Temperature effects on the structural behavior of laminated safety glass” Otto-
DT
Graf-Journal, Vol. 9, pp. 153-163.
Kollár, L.P. (2006), A mérnöki stabilitáselmélet különleges problémái, - Special problems of theory of
engineering stability, Akadémiai kiadó, Budapest, pp. 692-739.
Lakatos, T., Johansson, L.G. and Simmingsköld, B. (1972), “The effect of some glass components on
AR
the viscosity of glass”, Glasteknisk Tidskrift, Vol.27, No.2, pp. 25-28.
Lakatos, T. (1976), “Viscosity-temperature relations in glasses composed of SiO2-Al2O3-Na2O-K2O-
Li2O-CaO-MgO-BaO-ZnO-PbO-B2O3”, Glasteknisk Tidskrift, Vol.31, No.3, pp. 51-54.
KH
Lawn, B.R, Evans A.G. (1977), “A model for crack initiation in elastic/plastic indentation fields”,
Journal of Materials Science, Vol.12, pp. 2195-2199.
Lawn, B.R. (1985.10), „Interfacial forces and the fundamental nature of brittle cracks”, Appl. Phys.
Lett., Vol.47, No.8, pp. 809-811.
AN
Lawn, B.R. (1985), „Indentation: Deformation and fracture process”, Kurkjian, C.R. (Ed.), Strength of
inorganic glass, Plenum Publishing Corporation, New York, pp. 67.
Lawn, B.R., Marshall, D.B. and Dabbs, T.P. (1985), „Fatigue strength of glass: a controlled flaw
study”, Kurkjian, C.R. (Ed.), Strength of inorganic glass, Plenum Publishing Corporation, New
K.P
York, pp. 249-259.

Littleton, J.T., Preston, F.W. (1929), „The velocity of crystallisation in soda-lime-silica glasses. The
strength of glass containing cracks”, Journal of the Society of Glass Technology, Vol.13,
pp.350- 351.
Loughran, P. (2003), Falling glass problems and solutions in contemporary architecture, Birkhäuser-
Publishers for Architecture, Basel-Berlin-Boston, ISBN 3-7643-6712-1
t by
Lagunegrand, L., Lorriot, Th., Harry, R., Wargnier, H. (2006) „Design of an improved four point
bending test on a sandwich beam for free edge delamination studies”, Composites: Part B, 37,
pp. 127–136, doi:10.1016/j.compositesb.2005.07.002
Marshall, D.B., Lawn B.R. (1977), “An indentation technique for measuring stresses in tempered glass
h
surfaces” Journal of the American Ceramic Society, Vol. 60 No.1-2, pp.86-87.

Marshall, D.B., Lawn, B.R. (1985) “Surface flaws in glass” Kurkjian, C.R. (Ed.), Strength of inorganic
yrig
glass, Plenum Publishing Corporation, New York, pp. 171-180.

Mayer, G., Trejo, R., Lara-Curzio, E., Rodriguez, M., Tran, K., Song, H., and Ma, W.H. (2005),
„Lessons for New Classes of Inorganic/Organic Composites from the Spicules and Skeleton of
the Sea Sponge Euplectella aspergillum”, Mater. Res. Soc. Symp. Proc. Vol., Materials Research
cop
Society, pp. Y.2.4.1-8.

Michalske, T.A. (1977), “The stress corrosion limit: Its measurement and implications”, Fracture
Mechanics of Ceramics, Vol. 5, Surface Flaws, Statistics and Microcracking, (Ed.) Bradt, R.C.,
Evans, A.G., Hasselman, D.P.H., and Lange, F.F., Plenum Press, New York, pp. 277–89.
Miyasaka, Y., Kikuta, M. (2003), „Study on the fracture pattern of heat strengthened glass”, Glass
processing days 2003, Session 31, pp.677-679.
Mould R. E., Southwick, R. D. (1959) “Strength and static fatigue of abraded glass under controlled
ambient conditions : II ” J. Amer. Ceram. Soc., Vol. 42, No.12, pp. 582-592.
Nemes R. (2005), „Habüveg adalékanyagos könnyűbetonok” - “Lightweight concretes from glass
bubbles” Thesis at BME Dept. of Construction Materials and Engineering Geology
Nijsse, R. (2003), Glass in Structures, Birkäuser Publishers, Berlin
Norville, H.S. (1997), „The effect of interlayer thickness on laminated glass strength” Glass
processing days, 13–15 Sept. ’97, pp. 138-142., ISBN 952-90-8959-7
xiii
Okamoto et al. (2009), “Laminated glass and interlayer for use in such a laminated glass” Patent
application publication, Pub. No.: 2009/0092841 A1, Date: 04/09/2009 Assignee: Asahi Glass
Company, Honda Motor Co., Ltd., Tokyo US, pp.1-7.
Pagano N.J. (1989), Interlaminar response of composite materials. Composite material series, Vol. 5.
Amsterdam, Elsevier
Palotás, L., Balázs, Gy. (1980), Mérnöki szerkezetek anyagtana 3. Beton-Habarcs-Kerámia-Műanyag,
Tenet of materials of engineering structures 3. Concrete-Mortar-Ceramic-Plastic, Akadámiai
Kiadó, Budapest, p. 311., pp. 775-788.
Pankhardt, K. (1998), Diploma: Part 1.: Design, plans of a shopping-centre. Part 2.:Calculations of
glass and steel structures of overhead and façade elements. Part 3.:Calculations of heat-transport
through different types of glazing and optimalization of façade elements. Part 4.:Future of glass.,
DT
in German
Pankhardt, K. (2000), „Az üveg tartórendszerek fejlődése” – Development of load bearing glass
structures, Alaprajz, Vol.7, No.5, pp. 14–18., in Hungarian
Pankhardt, K. (2003), „Különleges üveg tartószerkezetek a magyar építőiparban”, - Special load
AR
bearing glass structures in Hungarian building industry, Építőanyag Vol. 55. No.3, pp. 106-111.
in Hungarian
Pankhardt, K., (2004), „Load-Bearing glass structures”, Periodica Polytechnica - Civil Engineering,
KH
Vol.48, No.1-2., pp. 157-172.
Pankhardt, K., Balázs, L.Gy. (2006), „New opportunities of structural glazing, loadbearing glass
structures.” Proceedings (Report) "Responding to tomorrow’s challenges in structural
engineering” Report, IABSE Symposium Budapest 2006; Vol.92, No.34-35, the whole paper on
AN
CD with 11 pages.
Pankhardt, K., Nehme, S.G. (2007), „Experimental studies on carbon fibre reinforced white cement
lightweight mortar and concrete elements” Proceedings of the 3rd Central European congress on
Concrete Engineering, Visegrád, Hungary, pp. 269-274.
K.P
Pankhardt, K. (2008a), „Investigation on load bearing safety glass” Eligehausen R., Gehlen C. (Ed.),
7th International PhD Symposium in Civil Engineering, Stuttgart, pp. 53-62.
Pankhardt, K. (2008b), „Investigation on load bearing capacity of glass panes”, Periodica
Polytechnica - Civil Engineering, Vol.52, No.2, pp. 73-82., doi:10.3311/pp.ci.2008-2.03
Pankhardt, K., Balázs, L. Gy. (2010a), „Study of edge strength of load bearing glasses”, Építőanyag,
Vol. 62, No.1, pp. 15-22.
t by
Pankhardt, K., Balázs, L. Gy. (2010b), „Temperature dependent load bearing capacity of laminated
glass panes”, Periodica Polytechnica - Civil Engineering, Vol. 54., No.1, pp.11–22 doi:
10.3311/pp.ci.2010-1.02
Pankhardt, K. (2010c), „Temperature dependent flexural stiffness of laminated glass panes”, Periodica
h
Polytechnica - Civil Engineering, under print in Vol. 54. No. 2.

Pankhardt K. (2010e), „Üvegtartók” („Glass load bearing structures”), 14. Fémszerkezeti konferencia,
yrig
Göd, 2010.10.14., (14th Metal structure conference, 14.10.2010., Göd, Hungary), printed in:
Könnyűszerkezetes építés, Vol.VI., No.2. 2010. October, pp. 27-36., in Hungarian
Pearman, H. (2002), Contemporary world architecture, Phaidon Press Limited, London, p.116., 196.,
276., 370. ISBN 0 7148 4203 6
cop
Pern, F.J., Glick, S.H. (2003), „Adhesion Strength Study of EVA Encapsulants on Glass Substrates”,
NCPV and Solar Program Review Meeting 2003, NREL/CD-520-33586 p. 942
Redner, A.S. (2000), „Automated Measurement of Edge Stress in Automotive Glass”, Strainoptic
Technologies, Inc., Glass Proceeding days 2000, Poster 24.
Peeterbroeck, S., Breugelmans, L., Alexandre, M., BNagy, J., Viville, P., Lazzaroni, R. and Dubois P.
(2007), „The influence of the matrix polarity on the morphology and properties of ethylene vinyl
acetate copolymers–carbon nanotube nanocomposites” Composites Science and Technology,
Vol.67, No.7-8, pp. 1659-1665.
Rimez, B., Rahier, H., Van Assche, G., Artoos, T., Biesemans, M. and Van Mele, B. (2008), „The
thermal degradation of poly(vinyl acetate) and poly(ethylene-co-vinyl acetate), Part I:
Experimental study of the degradation mechanism” Polymer Degradation and Stability, Vol.93,
pp. 800-810.
xiv
Prinssen, G. (2001), Transparenz und Sinnlichkeit, - Transparency and sensuality, Handbuch Farb- und
Spezialglas, (Ed.) Schott AG., Mainz
Roque, R., Hiltunen, D.R., and Buttlar, W.G. (1995), „Thermal cracking performance and design of
mixtures using Superpave™, Symposium for Superpave™ implementation, Journal of the
Association of Asphalt Paving Technologists, 64, pp. 718–735.
Sarikaya, M., Fong, H., Sunderland, N., Flinn, B.D., Mayer, G., Mescher, A. and Gaino, E. (2001),
„Biomimetic model of a sponge-spicular optical fiber—mechanical properties and structure” J.
Mater. Res., Vol.16, No.5, pp. 1420-1428.
Savineau, G.F. (1999), „Laminated architectural glass test performance criteria and reality” Glass
Processing Days 1999, pp. 388-392., ISBN 952-91-0885-0
Schittich, C., Staib, G., Balkow, D., Schuler, M., Sobek, W. (1999), Glass Construction Manual,
DT
Birkhäuser Publishers, Basel, pp. 65-69, p. 102., ISBN 3-7643-6077-1
Schwarzl, F., Staverman, A.J. (1952) „Time-temperature dependence of linear viscoelastic behavior”,
J. Appl. Phys., Vol. 23, No. 8, pp. 838 – 843.
Sobek, W., Kutterer, M., Messmer, R. (1999), „Shear stiffness of the interlayer in laminated glass”,
AR
Proceedings of the 6th international conference on architectural and automotive glass (Glass
Processing Days 1999), Tampere, Finland, pp. 360-365.
Suetomi, J., Omi, S., Nakaya, K., Aratani, S. (2001), „Stress generation in laminated glass plates”,
KH
Glass Processing Days, 18–21 June 2001, Poster 23, p. 526-529.
Széwald, O. (2001), „Üvegfeszültségek kialakulása és mérése a lámpagyártásban” – “Development of
stresses and measurement in lamp manufacture”, Építőanyag, Vol. 53, No.4., pp. 111-114.
Thompson, T. (2005), „Rheological study of linear and nonlinear viscoelastic behavior for silica-
AN
reinforced polybutadiene and polystyrene” Thesis presented to the Graduate Faculty of the
University of Akron
Török, Á. (2007), Geológia mérnököknek, - Geology for engineers, Műegyetemi kiadó, Budapest, pp.
44-81
K.P
Tucker, R.T. (2002), "Primers and Adhesion", Proc. of first thin-film module reliability national team
meeting at NREL, Golden, CO., Sept. 4-5.
Vallabhan, G., Minor, J., Nagalla, S. (1987), „Stresses in layered glass units and monolithic glass
plates”, Journal of structural Engineering, Vol. 113, 1987, pp. 36-43.
Vallabhan, C.V.G., Das, Y.C., Magdhi, M., Asik, M., Bailey, J.R., (1993) “Analysis of laminated glass
units, Journal of Structural Engineering, Vol. 119, No. 5., pp. 1572–1585.
t by
Van Duser, A., Jagota, A. and Bennison, J. (1999), „Analysis of glass/polyvinyl butyral laminates
subjected to uniform pressure", Journal of Engineering Mechanics, Vol.125, No. 4, pp. 435-442.
Weibull, W.A. (1951), “A statistical distribution function of wide applicability,” Journal of Applied
Mechanics, Vol. 18, No. 3, pp. 293–297.
h
Whitney, J. M. (1987), Structural analysis of laminated anisotropic plates, Technomic Publishing

Company, Inc., Pennsylvania.
yrig
Wiederhorn, S.M. (1967), „Influence of water vapor on crack propagation in soda-lime glass”,
Journal of the American Ceramic Society., Vol.50, No. 8., pp. 407-414.
Wiederhorn, S.M., Bolz, L.H. (1970), „Stress corrosion and static fatigue of glass”, Journal of the
American Ceramic Society, Vol.53, pp. 543-548.
cop
Wiederhorn, S.M., Lawn, B.R. (1979), „Strength degradation of glass impacted with sharp particles: I,
Annealed surfaces”, Journal of The American Ceramic Society, Vol. 62., No.1-2., pp. 66-70.
Wiederhorn, S.M., Lawn, B.R., Hockey, B.J. (1979), „Effect of particle impact angle on strength
degradation of glass” Journal of the American Ceramic Society, Vol.62, No.11-12, p.640.
Wiederhorn, S.M., Dretzke, A. and Rödel, J. (2002), „Crack growth in soda–lime–silicate glass near
the static fatigue limit”, Journal of the American Ceramic Society, Vol.85, No.9, pp. 2287–2292.
Woesz, A., Weaver, J.C, Kazanci, M., Dauphin, Y., Aizenberg, J., Morse, D.E. and Fratz, P.,
(2006),”Micromechanical properties of biological silica in skeletons of deep-sea sponges”, J.
Mater. Res., Vol.21, No.8, Materials Research Society, pp. 2068-2078.
Wölfel, E. (1987) „Nachgiebiger Verbund eine Näherungslösung und deren
Anwendungsmöglichkeiten”, - “Compliant bond as approximate solution and application
possibilities of it”, Stahlbau No.6, pp.173-180.
xv
Zimmermann, H., Merkwitz, M., Starck, A. (2001), Viscontrol-stabilising the glass production process
by controlling the viscosity, International glass journal, Faenza,Vol.115., pp. 45-48.
Standard specifications, Technical regulations
ASTM E 1300-04:2004, “Standard practice for determining load resistance of glass in buildings”
ASTM International, 2004 July (Beuth Verlag GmbH)
ASTM C1048-85:1985, "Standard specification for heat treated flat glass – kind HS, kind FT coated
and uncoated glass" ASTM
ASTM D790-99: “Standard Test Methods for Flexural Properties of Unreinforced and Reinforced
Plastics and Electrical Insulating Materials”, ASTM
ANSI: Z 97.1 -1984- Safety glazing materials used in buildings -Safety performance specifications and
DT
methods of test., American National Standards Institute (ANSI)
BSI 5051: Part 1: 1998, "Bullet-resistant glazing– Part 1: Specification for glazing for interior use"
CEB-FIP Model Code 1990, Comitè Euro-international du Bèton – Federation Internationale De La
Precontrainte, Lousanne, 1990.
AR
DIN 18516-T4:1990, “Außenwandbekleidung, hinterlüftet; Einscheiben-Sicherheitsglas;
Anforderungen, Bemessung, Prüfung“
ISO 48:1994, “Rubber, vulcanized or thermoplastic - Determination of hardness (hardness between 10
IRHD and 100 IRHD)”, TC45/SC2
KH
EN 356:1999 “Security glazing – Testing and classification of resistance against manual attack”
EN 572-1:2004,”Glass in building - Basic soda lime silicate glass products – Part 1: Definition and
general physical and mechanical properties” CEN, Brussels
EN 572-2:2004, “Glass in building - Basic soda lime silicate glass products – Part 2: Float glass”
AN
EN 572-9:2004, "Glass in Building - Basic soda lime silicate glass products – Part 9: Evaluation of
conformity/Product standard”, February 2004, CEN, Brussels
EN 1288-1:2000, “Glass in Building – Determination of the bending strength of glass – Part 1:
Fundamentals of testing glass”, 2000, CEN, Brussels, pp. 7-8.
K.P
EN 1288-2: 2000, "Glass in Building – Determination of the bending strength of glass – Part 2:
Coaxial double ring test on flat specimens with large surface areas"
EN 1288-3: 2000, “Glass in Building – Determination of the bending strength of glass – Part 3: Test
with specimen supported at two points (four-point bending)”, 2000, CEN, Brussels; pp. 8-11.
EN 1288-5: 2000, "Glass in Building – Determination of the bending strength of glass – Part 5:
t by
Coaxial double ring test on flat specimens with small surface areas"
EN 1863-1, “Glass in building – Heat strengthened soda lime silicate glass - Part 1: Definition and
description”
EN 12150-1: 2000, "Glass in Building – Thermally toughened soda lime silicate safety glass – Part 1:
h
Definition and description" CEN, Brussels

EN 12150-2: 2004, "Glass in Building – Thermally toughened soda lime silicate safety glass – Part 2:
yrig
Evaluation of conformity/Product standard ", February 2004, CEN, Brussels

EN ISO 12543:2000, „Glass in building - Laminated glass and laminated safety glass” Part 1:
Definitions and description of component parts, Part 2: Laminated safety glass , Part 3:
Laminated glass, Part 4: Test methods for durability, Part 5: Dimensions and edge finishing ,
Part 6: Appearance”, CEN, Brussels
cop
EN 12600:2003, “Pendulum test – Impact test method and classification for flat glass ”
EN 14178-1:2004, "Glass in Building - Basic soda lime silicate glass products – Part 1: Float glass”,
February 2004, CEN, Brussels
prEN 13464-1-5.:1999-2003, „Glass in building – Design of glass panes – Part 1 : General basis of
design, 1999, Part 2: Design for uniformly distributed loads, 2000, Part 3: Design for line loads,
2003, Part 4: Design for concentrated loads, Part 5: Design for superposition of loads”, CEN,
Brussels
EN 14449:2005, „Glass in building - Laminated glass and laminated safety glass -Evaluation of
conformity/Product standard”, CEN Management Centre: rue de Stassart, 36 B-1050 Brussels
EN 15682-1:2007, “Glass in building - Heat soaked thermally toughened alkaline earth silicate safety
glass - Part 1: Definition and description”, CEN Management Centre: rue de Stassart, 36 B-1050
Brussel
xvi
Glass Technical Document TD-138, Heat Treated Glass for Architectural Glazing. pdf, pp.1.
Information sheet, Supplement to Technical Regulations 4.4.3 „Use of glass and acrylic glass in stand
construction and design inside fair halls”, 12. April 2002
ISO 48:1994, „Rubber, vulcanized or thermoplastic - Determination of hardness (hardness between 10
IRHD and 100 IRHD)”, TC45/SC2
ISO 75-1:2004 Plastics – Determination of temperature of deflection under load - Part 1: General test
method, ISO Online, CH-1211 Geneva 20, equivalent ASTM D648 - Standard Test Method for
Deflection Temperature of Plastics Under Flexural Load
Technical datasheet VIAPAL 4808 VUP B/62 resin (2004), given by the manufacturer CYTEC
Surface Specialties, Surface Specialties Austria LtD., Graz
Technical guide EVASFE (2005), for the application of EVASFE to laminated glass, given by the
DT
manufacturer Bridgestone, Japan
TRLV „Technische Regeln für die Verwendung von linienförmig gelagerten Verglasungen”
„Technical regulations for the application of linearly supported glazing”, DIBt (German Institute
for Building Techniques)
AR
TRAV (E-TRAV):2001 „Technische Regeln für die Verwendung von abstrurzsichernden
Verglasungen” „Technical regulations for using crash proof glazing”, DIBt (German Institute
for Building Techniques)
WWW Links
KH
Anderson, (2001), „Applications and processing of ceramics” Introduction to Materials Science,
AN
Chapter 14-16, Applications and Processing of Ceramics, University of Virginia, Dept. of
Materials Science and Engineering, http://people.virginia.edu/~lz2n/mse209/Chapter14-15.pdf
Banks, R. (2008), “Chemistry tutorials by Richard Banks” Boise State Univ., Dept. of Chemistry and
Biochemistry, Retrieved 2008
K.P
http://chemistry.boisestate.edu/people/richardbanks/glassblowing/glassblowing_history.htm
Barker, D.B., Yang, Y. (2006), „Effect of proof testing on optical fiber fusion splices” C01-34,
CALCE Electronic Products and Systems Center, University of Maryland, Retrieved 2006,
http://nepp.nasa.gov/docuploads/D55C314D-E985-4241
8AEA58F450DC7F69/1_3_Effect_of_Proof_Testing.pdf
t by
Bloch, T. (2009), “The Glass Harmonica”, http://www.finkenbeiner.com/gh.html

Block, V. (2002), “New version of glass strength standard is major departure ASTM E1300, a work in
progress, guides design and compliance”, Guide to glass codes
http://www.glasswebsite.com/techcenter/articles/Glass%20Guide1028.pdf
Chaplin, M. (2008), “Water structure and science” http://www.lsbu.ac.uk/water/ Rheology
h
primer.htm, Retrieved 2008

yrig
http://www.glassforeurope.com/theindustry/FactsAndFigures/floatprocess/Pages/default.aspx, Glass
for Europe, Retrieved 2008
Gibbs P. (1996), “Is glass liquid or solid?” updated PEG January 1997.
http://www.netmug.org/~oscar/pdf/Isglassliquidorsolid.pdf
Goubard, D., Pollacchi, B., Martin, S. and Charrière, B. (2009), “A practical way of predicting low
cop
temperature tack of hot-melt pressure sensitive adhesives” Bostik, Centre de Recherches et

Développements Ribécourt, France http://www.bostik.com/upload/news/file11.pdf,
http://www.afera.com/files_content/papervitiants2006.pdf
Hilmer, M. (2009), ” Historical Development of Glass Music” http://www.glasmusik.com/hist1.htm
Kamath, M.G., Dahiya, A., Hegde, R.R, Kannadaguli, M. and Kotra R. (2004), “Chemical bonding”
http://www.engr.utk.edu/mse/pages/Textiles/Chemical%20Bonding.htm, Retrieved 2008
Mathweb (2009), “Deflection temperature testing of plastics”
http://www.matweb.com/reference/deflection-temperature.aspx, Retrieved 2009
Musical glasses (2009), In Encyclopaedia Britannica. Retrieved January 21, 2009, from Encyclopaedia
Britannica Online: http://www.britannica.com/EBchecked/topic/399168/musical-glasses
Pankhardt, K. (2009), “Effect of temperature on load bearing capacity of laminated
glasses”„Hőmérséklet hatása a laminált üvegek teherbírására” DAB Symposium
xvii
“Day of science” 11th of Nov. 2009, DAB and Technical Faculty Dept. of Civil Engineering "A
Tudomány Napja" 2009. November 11.,
Poster http://www.mk.unideb.hu/images/stories/epito_poszterek/pankhardt_kinga.jpg
B&G glass (2009), „Glass terminology” http://www.bgglass.com/glass%20terms.htm, Retrieved in
2009.
Glassguides (2009), Achitect’s guide to glass & metal, „Thermal stress, Part two”.
http://www.glassguides.com/index.php/archives/610#
WWW reference of pictures

Grogan, M. (2006), Photo taken by Grogan M., University of Virginia, Date 27.01.2006
DT
http://en.wikipedia.org/wiki/Prince_Rupert%27s_Drop
Kengo Kuma, Atami, Japan (1995) The Japan Architect. Summer, 2000, No.38, p. 120.
National Geographic Channel (2002), captured frame “bullion window”
AR
The Bakken Library and Museum (2007), http://www.thebakken.org/exhibits/mesmer/glass-
armonica.htm
KH
AN
K.P
h t by
yrig
cop
xviii
K. Pankhardt Load bearing glasses Appendix A
APPENDIX A
Contents
Appendix to literature review A1
A.1 to Chapter 1 Research significance and aims of Thesis A1
A.2 to Chapter 2 Properties of glass A4
A.3 to Chapter 3 Float glass as building material A8
A.4 to Chapter 4 Plastic interlayer materials of laminated glass A12
A.5 to Chapter 5 Strength and durability of glass A17
A.6 to Chapter 6 Bending strength of float and laminated glasses A22
DT
Appendix to literature review
AR
A.1 to Chapter 1 Research significance and aims of Thesis
KH
AN
K.P
Figs. A.1.1 a) Mezzanine glass roof in Gresham Palace (2002) b) placing of glass pane on
t by
glass beams spanning 9 m, calculations of glass roof and glass beams spanning 9 m and
coordination at finishing site by K. Pankhardt (Glasmetal Ltd.)
h
yrig
cop
Figs. A.1.1 c) Mezzanine glass roof in Gresham Palace (2002), jointed glass beams
spanning 2×4.5=9 m, calculations by K. Pankhardt d) Testing of glass beams (BME)
A1
DT
Figs. A.1.1 e) Load bearing glass slab in Sándor Palace, Budapest, calculations by K.
Pankhardt, in 2001
AR
KH
AN
Figs. A.1.1 f) Load bearing “Kossuth, and Andrássy-bridge” glass slabs in Gresham
K.P
Palace, calculations by K. Pankhardt, in 2002 (Glasmetal Ltd.)

h t by
yrig
Figs. A.1.1 g) Load bearing “Gresham-bridge”cable tesioned bridge, calculations by K.

Pankhardt, in 2002 (Glasmetal Ltd.)
cop
Figs. A.1.1 h) Point fixed glazing of MEO bridges, calculations by K. Pankhardt, in 2001
A2
DT
Figs. A.1.1 i) Jumbo size safety insulating glass units in Gresham Palace, calculations by
K. Pankhardt, in 2002 (Glasmetal Ltd.)
AR
KH
AN
Figs. A.1.1 j) Point fixed glazing, Office building Tél utca 14 th district, Budapest,
K.P
calculations by K. Pankhardt, in 2001 (Glasmetal Ltd.)

h t by
yrig
Figs. A.1.1 k) Point fixed glazing inside Public prosecutor’s building, Budapest,
calculations by K. Pankhardt, in 2003 (Glasmetal Ltd.)
cop
Figs. A.1.1 l) Safety laminated glass balustrade in BME Department of Construction

materials and engineering geology (Pankhardt, 2006)
A3
DT
Fig. A.1.2 Tension tested laminated glass specimens (Pankhardt, Nehme 2004, BME)
AR
A.2 to Chapter 2 Properties of glass
KH
AN
a) b)
Fig. A.2.1 a) Silica minerals (agate), b) Silica minerals with hexagir crystal shape
(left: amethyst, right: quartz), Museum of minerals, Siófok, Hungary
K.P
t by
a) b)
h
Fig. A.2.2 a) Silica minerals (opal), Museum of minerals, Siófok, Hungary b) The (shell-
like) fractured surface of non heat-treated glass (19 mm thick specimen of this Thesis)
yrig
cop
Fig. A.2.3 Shell-like fractured surface of non heat treated glass (19 mm thick specimen of
this Thesis)
A4
a) b) c)
Fig. A.2.4 a) Natural glass of sea-sponge, b) Photo by Max-Planck-Institut für Kolloid-
und Grenzflächenforschung, c) Photo by University of California at Santa Barbara
DT
(Aizenberg, 2005)
AR
KH
AN
K.P
a) b)
Fig. A.2.5 a) Effect of loading rate on the mechanical response of sea sponge Euplectella
aspergillum fibres b) Fracture surface of fibres tested in tension, showing central region
t by
(SEM) with fast fracture (Mayer et al., 2005)

h
yrig
cop
Fig. A.2.6 Stress–strain diagrams of the Antarctic sponge Rosella racovitzea fibre (spicule)
and silica glass rods in three-point bending (glass rods were General Electric -Type-214
high-purity fused silica. Samples 20–25 mm in length were tested in a three-point bending
fixture with a 10 mm span and an articulated centre loading rod with a crosshead rate of
0.5 mm/min. The diameter of the spicules ranged from 0.34 to 0.56 mm, whereas the silica
rods were 1.14 mm in diameter (Sarikaya et al., 2001)
A5
Fig. A.2.7 Early glass pots and vases in Lebanon, Fig. A.2.8 „Prince Rupert's Drop”
Beit Eddine, 2002 (Grogan, 2006)
DT
AR
KH
Fig. A.2.9 Crystal Palace London 1851, Joseph Paxton (1801-1865), with 71540 m2 glass
surface, (Schittich et al., 1999)
AN
K.P
Fig. A.2.10 The principal steps in the manufacture of float glass (Schittich et al., 1999)
t by
Table A.2.1 Composition of glass types by m% of their components

1 2 3 4 5 6
soda lime container fibre- boro- high silicate
h
Components silicate glass, glass silicate lead glass

yrig
m% glass (ancient wool glass crystal

Roman (Pyrex)
glass)
Silica, SiO2 67.5-77.3 62.5-81 48.6-78 63-84 35.0 92-98
Soda, Na2O 12-15 11-15 13-17 4-8 -- --
cop
Lime, CaO 7-9 7-12 5-11 0-2 -- --

Potash, K2O 0-2 0-2 0-2 0-3 7.2 --
Magnesia, MgO 3-4 0-3 1-5 -- -- --
Alumina, Al2O3 0.1-2 1-3 0-6 2-7 -- 0.5
Iron Oxide, Fe2O3 0.1-1.5 0-0.4 0-0.6 -- -- --
Boric Oxide, B2O3 -- -- 3-9 10-15 -- 3.0
Lead Oxide, PbO -- -- (0.01) -- -- 58.0 --
Glass transition
270-550 > 550 < 550
temperature, °C
A6
DT
a) b)
Fig. A.2.11 a) NiS (nickel-sulphide) inclusion in the centre of glass thickness,
AR
b) Diameter of NiS vs. surface residual stress (Miyasaka, Kikuta, 2003)
Table A.2.2 Glass product groups by manufacturer AGC (2008)
KH
Base products Processed products
Product Standard Primary Standard Secondary Standard
processing processing
Float glass EN 572-1,2 Coated glass EN 1096- Thermally EN 12150-1
AN
(magnetron 1-3 toughened EN 14479-1
coated, pirolytic) glass
Patterned EN 572-1,5 Mirror EN 1036-1 Heat- EN 1863-1
K.P
glass strengthened
glass
Wired EN 572-1,6 Surface Enamelled EN 1863-1,
glass treatment and silk- EN 12150-
(acid etching, screen 1,
t by
sandblasting etc.) printed glass EN 14479-1

Polished EN 572-1,3 Laminated glass EN 12543- Chemically EN 12337-1
wired glass 1-6 toughened
glass
h
Profiled EN 572-1,7 Bent glass

glass
yrig
Drawn EN 572-1,4 Insulating EN 1279-1-

sheet glass glass 6
Spandrels
cop
Fig. A.2.12 “Jumbo” size (6.00 m× 3.25 m) glass by Rákosy Glass Ltd.
A7
Fig. A.2.13 Cutting of glass
DT
A.3 to Chapter 3 Float glass as building material
AR
KH
AN
K.P
Fig. A.3.1 Forming of glass at high temperatures (Snapshot of Nat. Geo. Channel, 2004 )
Table A.3.1 Viscosity points of soda lime silicate glasses

Viscosity points Temperature, °C Viscosity, Pa⋅s
t by
Practical melting point (melting temperature) 1550 10

Working point (working temperature) 650 – 1100 103
(Littleton) softening point 675 – 725 106.6
Dilatometric softening point 675 – 725 108.5
1012
h
Annealing point (tempering temperature) 550 – 600

Glass transition temperature 550 1012.6
yrig
Strain point (stress forming temperature) 450 – 500 1013.5
Table A.3.2 Basic physical properties of soda lime and alkaline earth silicate glasses
cop
Properties Symbol Soda lime Alkaline earth Unit

silicate glass silicate glass
Density (at 18°C) ρ 2500 2700 kg/m3
Thermal expansion α 88 × 10-7 80 × 10-7 1/°C
(between 20°C and 300°C)
Specific heat capacity c 0.72×103 0.7×103 J/(kg⋅°K)
Thermal conductivity λ 1 0.8 up to 1.1 W/(m⋅°K)
Mean refractive index nR 1.5 1.5 -
(to visible radiation at 589.3 nm)
A8
liquid
glass supercooled liquid
Stress forming
No stress in glass
Deformation temp.
Softening point
Viscosity
DT
Working temp.
Liquid temp.
AR
Melting range
Temperature (T) of glass
KH
a) b)
Fig. A.3.2 a) Viscosity (η) vs. temperature (T) diagram of glass (Balázs, 1984),
b) Viscosity (η) vs. temperature (T) diagram of different glasses (Anderson, 2001)
AN
K.P
t by
Fig. A.3.3 Front entrance of Gresham Palace with service load of people on the cantilever
overhead glazing, calculations by Pankhardt, 2004
h
yrig
cop
Fig. A.3.4 Fractured cantilever overhead, Humanic store in Vienna, 2007
A9
Fig. A.3.5 Fracture of wired glass pane due to thermal expansion in overhead glazing of
a hotel building, Siófok, 2007
DT
AR
KH
AN
Fig. A.3.6 The principal steps in the manufacture of tempered glass (feed in, heating,
cooling) in Rákosy Glass Ltd.
K.P
h t by
yrig
Fig. A.3.7 “Leopard spots” of tempered glass by polarised light

cop
a) b) c)
Fig. A.3.8 Fracture pattern of a) float glass tested in four-point bending (Pankhardt, 2008b);
b) of heat-strengthened glass (insulating glass unit, National theatre, 2004); c) tempered
tensioned glass specimen (Pankhardt, 2004)
A 10
Fig. A.3.9 Fracture of float glass pane due to thermal expansion, office gallery, Budapest,
DT
2007
AR
KH
Fig. A.3.10 Distribution of cracks in thickness of tempered glass, tensioned specimen with
hole (Pankhardt, Nehme, 2004)
AN
K.P
t by
Fig. A.3.11 Acid etched glass surface by Rákosy Glass, 2009

h
yrig
Fig. A.3.12 Stress distributions over the thickness of a) tempered glass, b) of heat-
cop
strengthened glass and c) chemically strengthened glass
Table A.3.3 Requirements on compressive stress in outer layer for tempered and heat-
strengthened glasses (ASTM C1048-85:1985)
Compressive stress in outer layer of Heat tempered Heat-strengthened
glasses glass glass
Compressive stress of Region 1, N/mm2 min. 69 min. 24
Compressive stress of Region 2, N/mm2 min. 67 no requirement
A 11
DT
Fig. A.3.13 Recycled glass products by Geofil Ltd., Tatabánya, Hungary
A.4 to Chapter 4 Plastic interlayer materials of laminated glass
AR
KH
AN
K.P
Fig. A.4.1 Arnhem Zoo Bridge, Rijkerswoerd, Netherlands (Nijsse, 2003)

The walls with two glass layers are 10.16 mm thick and the roof is 7.62 mm thick. The beams
are spanning 3.81 m between the two buildings with three glass layers of altogether 12.7 mm.
h t by
yrig
a) b)
cop
Fig. A.4.2 Bullet and explosion resistant glass, a) Glascon conference and exhibition
Stuttgart, Germany (2003) b) product of Rákosy Glass
Fig. A.4.3 Fire “resistant” laminated glass product PYRANOVA, product of Shott Glass,
Class: F30
A 12
Fig. A.4.4 The principal steps in the manufacture of laminated safety glass with PVB foil
interlayer (Schittich et al., 1999)
DT
140
130 C°
120
100 100 C°
Temperature (C°)
AR
80
60
room temp.
40
30 -90 min. 30 min.
KH
20
0
0 50 100 150 200 250
Time (min.)
b)
AN
a)
Fig. A.4.5 a) Autoclave process of EVA foil laminate (Technical guide of Evasafe, 2005);
b) Vacuum & autoclave lamination process by Rákosy Glass
K.P
h t by
yrig
Fig. A.4.6 Laminating with resin “cast in place”

cop
Fig. A.4.7 Chemical composition of EVA (Rimez et al., 2008)
A 13
DT
Fig. A.4.8 Change of molecular structure of EVA during laminating process (Technical guide
of Evasafe, 2005)
AR
KH
AN
K.P
Fig. A.4.9 Stress–strain curves of EVA materials at room temperature

with different VA content (Peeterbroeck et al., 2007)
h t by
yrig
Fig. A.4.10 Delamination of laminated glass pane around the edges (Pankhardt, 2008a)
cop
a) b)
Fig. A.4.11 Delamination of resin laminated glass pane a) one month after finishing, b)
three months after finishing
A 14
10.0
(GPa)
0.10
DT
(GPa)
Fig. A.4.12 Shear modulus, G, and temperature, T, for various molecular weights and cross-
AR
linking (Hertz Jr. ,1979)
KH
AN
K.P
Fig. A.4.13 Semi-crystalline polymers (Anderson, 2001)

h t by
3.
G’ 4.
yrig
1.
2.
G”
cop
tanδ : G”/G’
Fig. A.4.14 Rheological state of interlayer as a function of temperature (glassy region -

transition region - rubbery plateau – flow zone) of hot-melt pressure sensitive adhesives
(Goubard et al., 2009)
A 15
G
G”
δ G’
Fig. A.4.15 Vector relationship of G', G" and G
DT
AR
KH
AN
Fig. A.4.16 Amorphous and crystalline polymers (Hertz Jr., 1979)
K.P
h t by
Fig. A.4.17 Temperature dependent stress vs. strain diagram of polymers

yrig
cop
Fig. A 4.18 Master compliance curve (Roque et al. 1995, Thompson 2005)
A 16
DT
AR
KH
Fig. A.4.19 Semi-crystalline polymers (Anderson, 2001). The Young’s modulus of plastics are
dependent on time and temperature. The (viscoelastic) relaxation modulus can be defined as
AN
Er(t) = σ(t)/ε. Stress decreases with time due to molecular relaxation processes.
K.P
A.5 to Chapter 5 Strength and durability of glass

h t by
a) b) c) d)
yrig
cop
e)
Fig. A.5.1 Finishing techniques of glass pane by Rákosy Glass Ltd.
Drilling a hole in glass pane a) drilling machine, b) drilled glass core c) grinding machine
for drilled hole d) finished glass hole in pane e) hand cutting of float glass pane
A 17
DT
a) b)
Fig. A.5.2 a) Curved shaped glass roof with water collection. Hydra-Pier in
Haarlemmermeer, Netherlands. Design: Asymptote Architecture. Photos by Christian
AR
Richters (Jodidio, 2002), b) Leaching of glass, office gallery, Budapest, 2007
KH
AN
K.P
a) b)
Fig. A.5.3 a) „Sharp” Vickers indentation and generated crack types from the deformation
zone (Lawn, 1985), b) Schematic geometry of cracks by Vickers indentation (Dabbs,
Fairbanks and Lawn, 1984)
h t by
yrig
Fig. A.5.4 Optical micrograph of Vickers indentation in soda-lime glass viewed from below
contact a) at full load (P=40 N) and b) at complete unload. Radial cracks: R, lateral: L.
(Arora et al. 1979)
cop
Fig. A.5.5 Scanning electron micrograph of Vickers indentation in a) unetched b) etched

soda lime glass. The indentation load is 4 N, width of field is 40 µm. Radial cracks initiate
from the shear fault structure in the zone (Dabbs, Fairbanks and Lawn, 1984).
A 18
< 0,01 mm 45 MPa

0,01 mm 40 MPa
0,02 mm 35 Mpa
0,05 mm 30 Mpa
0,10 mm 25 MPa
> 0,10 mm 20 Mpa
Fig. A.5.6 Distribution of surface damage a) new glass; b) weathered glass; c) glass with
inherent damage and their influence on bending strength of glass (Schittich et al., 1999)
DT
AR
KH
Fig. A.5.7 Optical micrograph of soda lime glass surface impacted with silica carbide (SiC)
particles, velocity 113 m/s (Wiederhorn and Lawn, 1979)
AN
K.P
Time to critical nucleus, tc (s)
h t by
yrig
Indentation load, P (N)
Fig. A.5.8 Median times for radial crack initiation as function of indentation load for soda-
cop
lime glass in water, air or inert environments. Time, tc=T+ td (where T is hold time for
indentation load, and td is the delay time =0) (Lawn, 1985 data courtesy Fairbanks, C.J)
Table A.5.1 Aging of silicate glass in aqueous Si(OH)4 solution at 90°C

(Ito, Tomozawa, 1982)
Strength (MPa)
Flaw (damage) type
t=0 t = 10 days
Abrasion 63.3 6 2.4 83.2 6 13.0
Machining 50.2 6 1.9 67.2 6 10.0
Indentation radial crack 47.0 6 3.7 58.5 6 7.8
Indentation cone crack 36.0 6 2.2 39.6 6 3.0
A 19
DT
Fig. A.5.9 Strength of Vickers-indented (P=5N) soda-lime glass as a function of aging time
(stored in silicone oil aging environment between indentation and strength tests) (Lawn et al.,
AR
1985).
KH
AN
K.P
t by
Fig. A.5.10 a) Dynamic fatigue of Vickers-indented soda-lime glass in water. Data at single
contact load for as-indented and post-indentation annealed specimens. b) Response of
indentation cracks in soda-lime glass, in water, to applied stress. Data at fixed stressing rate,
for as-indented and post-indentation annealed specimens (Lawn et al., 1985)
h
yrig
cop
Fig. A.5.11 Crack growth in silicate glasses. All specimens were immersed in distilled
water (Wiederhorn and Bolz, 1970)
A 20
Fig. A.5.12 Figures a) and b) show a crack front containing light bands separated by a few
dark bands. The dark bands show no regularity along the arrested crack front and also vary
DT
in size. The angle of crack growth again changed when the stress intensity factor was
increased, and the inclined cracks rejoined each other to form a single crack surface again.
The portion marked by the square was examined by AFM (Wiederhorn, Dretzke, Rödel 2002)
AR
KH
AN
K.P
Fig. A.5.13 Scanning electron micrographs of glass surfaces impacted with SiC particles at
v=94 m/s a) impacting angle 90°, b) impacting angle 15° (Wiederhorn, Lawn, Hockey, 1979)
h t by
yrig
cop
Fig. A.5.14 Strength decrease of glass as a function of impact angle of SiC particles at
v=94 m/s. Data points represent mean and standard deviation. Rupture was effected in a
ring-on-ring arrangement (flexural tension), with the target face lying on the tension side.
(Wiederhorn, Lawn, Hockey, 1979)
A 21
A.6 to Chapter 6 Bending strength of float and laminated glasses
DT
Fig. A.6.1 Main viewing ports at observation deck, height 222 m, Auckland’s Sky Tower,
architect: Craig Craig Moller Ltd., structural engineering: Beca Carter Hollings & Ferner
AR
Ltd. 1994-1997.
G, E, η glass (h1)
KH
ηi Gi, Ei interlayer
material AN
G, E, η glass (h2)
Fig. A.6.2 a) Kelvin-Voigt model of creep Fig. A.6.2 b) Maxwell model of relaxation of
of laminated glass pane a laminated glass pane
K.P
h t by
Fig. A.6.3 Laminated glass subjected to impact with an impactor of 10×5 cm and 4.1 kg
a) hard missile (Kadri I., et. al , 2003), b) large soft missile with spherical impacting end,
yrig
c) large soft missile with flat impacting end (Dharani et. al, 2005) used mainly in hurricane
exposed areas.
cop
Fig. A.6.4 Laminated glass subjected to ball drop test a) hard missile (Dharani et. al, 2005)
b) according to EN 356:1999 with 4.11 kg h=1.0-6.0 m falling ball impact (Kadri I., et. al ,
2003).
A 22
DT
Fig. A.6.5 Laminated glass subjected to pendulum test a) soft missile (Kadri I., et. al , 2003)
b) 45(50) kg bag or 1kg steel ball (D10), h=1.2-1.5m -ANSI Z.97:1984, or 50 kg tyre, h=1.2m
–EN 12600:2003.
AR
KH
AN
Fig. A.6.6 Glass sea structure. Conservation of ship SS Great Britain (Isambard Kingdom Brunel,
1806-1859). Often dubbed the ‘first great ocean liner’. From 1845 to 1886 she served as a North
K.P
Atlantic passenger liner, a Crimean war and Indian Mutiny troopship, and made over 30 round trips
between England and Australia. Impact testing of prototype panels (4,35×1,5 m, thickness of 21 mm
of laminated heat-strengthened glass) was undertaken to validate the design and thus ensure both the
safety of visitors beneath and the integrity of the glass plate to protect the dehumidified space beneath
from flooding. Two potential impacts were tested: a 1.5 kg hammer dropped from 15 m up in the
t by
ship’s rigging and the 6 m fall of a 100 kg person from the ship’s deck. Neither was allowed to cause
significant leakage (Jofeh, C., Perry, A., 2005)
h
yrig
cop
Fig. A.6.7 Rigidly bonded glass layers (theoretic: Eglass=Einterlayer)
Fig. A.6.8 Theory of elastic bonded glass layers,

a) glass remains rigid, b) glass is also elastic
A 23
DT
AR
Fig. A.6.9 Coupling parameter κ(T) for a load duration t = 10 s and t = 1000 s
(Krüger,1998)
KH
Table A.6.1 Design and bending strength values of different types of glazing
Product standard TRLV (Guideline)
Characteristic values Design strength
Bending strength Overhead Vertical
AN
Bending strength
characteristic values glazing glazing
(N/mm²) (N/mm²)
Float glass 45* 12** 18***
K.P
Casted flat glass 25 8 10

Profile glass 45 - -
Tempered safety float glass 120* 50 50
Casted safety glass - 37 37
Tempered float glass, ceramic coated on
75* 30 30
t by
tension side
Heat strengthened glass 70* 29 29
Heat strengthened glass, ceramic coated
45* 18 18
on tension side
h
Laminated glass (float) - 15 22.5

*prEN 13474-1:1999; **only applicable for upper glass in insulating glass units, the lower glass must
yrig
be a laminated glass, single float in overhead glazing forbidden; ***slope up to 10°

Table A.6.2. Standardization of glazing in Germany before and parallel to the existing
DIN EN standards
cop
Vertical glazing
Overhead Walking
Only for separating Prevention from
glazing areas
areas falling
Sloped overhead Glass façade, Glass façade, Slabs,
glazing Separation wall Balustrade Stairs
Supported TRLV,
TRLV TRAV, ZiE ZiE
along edges DIN 18516 T4*
Supported at ZiE,
ZiE TRAV, ZiE ZiE
points DIN 18516 T4*
Glued AbZ
* it is allowed only in special cases
A 24
Lb .
Ls
Fig. A.6.10 Specimen tested in four point bending (EN 1288-3:2000)
DT
AR
KH
AN
Fig. A.6.11 Nonfactored load chart for glass Fig. A.6.12 Deflection chart for glass
simply supported along two parallel edges simply supported along two parallel
K.P
(ASTM E 1300-04) edges. Chart A 1.25 of ASTM E 1300-04

h t by
yrig
cop
Fig. A.6.13 Technical guideline of INTERPANE for four-side supported laminated glass. For
example, glass type ipasafe S30 thickness of 32 mm should be used for a glazing supported at
side length of 95 cm×105 cm (Häuser et al., 2002).
A 25
Table A.6.3 Methods of calculations on laminated glass subjected to bending

Method Reference Contexts Advantages Disadvantages
Concept Wölfel Coupling of glass For laminated glass
heff = h + h + 12ΓJ s
3 3
1
3
2
of (1987) layers can be taken consisting of two glass
effective Γ : shear transfer into account. layers, supported on
thickness, coefficient, varies between: Effect of shorter sides.
heff 0 ≤ Γ ≤ 1. temperature and Primary (and only)
0: non bonded glass layers duration of loading interlayer material
1: rigidly bonded glass can be taken into property is its shear
layers account. modulus, G.
DT
Effective Norville M Coupling of glass Parameter q, was
W eff = =
section (1997) σ layers can be taken determined with
modulus,  h q ⋅ h q ⋅ h int  into account. experiments, only for
Weff = b ⋅h⋅ + +  PVB foil, with two
3 3 2 
AR
sides supported,
q: ratio of horizontal shear uniformly distributed
force transferred by PVB loaded panes.
foil between glass layers,
varies between:
KH
0≤ q≤ 1.
Coupling Krüger ( E ⋅ I )beam = Coupling of glass Parameter κ, was
para- (1998) layers can be taken determined with
= Eg ⋅ I g + Eint ⋅ I int + Ec ⋅ I c ≈
AN
meter, into account. experiments, only for
κ ≈ Eg ⋅ I g + Ec ⋅ I c Effect of PVB foil, with two
- I c = κ ⋅ b ⋅ h ⋅ (h + hint ) 2 temperature and sides supported, three-
κ : represents the coupling duration of loading point bended panes.
K.P
of glass layers , varies can be taken into

between: account.
0≤ κ≤1
0: non bonded glass layers
1: rigidly bonded glass
t by
layers
Standard EN 1288-  3( Ls − Lb )  The strength of Can be applied for
3:2000 σ bB = k  Fmax + σ bG  pane in edge region single glasses. In the
 
2
2bh
k = 1, in the middle of pane can be determined, case of approximative
therefore, regions calculations with two
h
k = ke, in the edge region of

pane, can be determined with different sides supported, four-
yrig
with diagrams as function strength can be point bended panes.

of y/h defined.
ASTM E The load resistance of glass Variability of Standard was based on
1300-04 pane: supports. The effect PVB performance
cop
(2004) LR = NFL × GTF × LS of temperature can under uniformly

(with use of load charts and be taken into distributed loading
diagrams) account. with load duration to
failure in three
seconds.
FEM - Dependent on application. Freedom by design. Laboratory
experiments required.
Experi- - Dependent on application. Surface condition Size effect in the case
ments (or treatment) of of scaled-down
glass can be taken specimens.
into account.
A 26
K. Pankhardt Load bearing glasses Appendix B
APPENDIX B
Contents
Appendix to test parameters and test results B1
B.1 to Chapter 7 Experimental programme B1
B.2 to Chapter 8 Effect of tempering on load bearing capacity of single glass pane in bending B6
B.3 to Chapter 9 Effect of lamination on bending characteristics B11
B.4 to Chapter 10 Effect of temperature on bending characteristics B15
B.5 to Chapter 11 Flexural stiffness of laminated glasses B24
B.6 to Chapter 12 Residual load bearing capacity of laminated glass panes B31
B.7 to Chapter 13 Future work B35
DT
Appendix to test parameters and test results
AR
B.1 to Chapter 7 Experimental programme
KH
AN
K.P
Fig. B.7.1 Glass separation wall (failure started by all panes - about 120 pieces),
Visegrád, Hungary (2008)
t by
Table B.7.1. Simplified symbols of the studied specimens used in this Thesis
Symbol of specimens
Type Number of Type of interlayer
Float Tempered
h
of glass glass layers material

specimens specimens
yrig
F_6mm E_6mm
Tempered (symbol: E)
single glass,
F_12mm E_12mm
Float (symbol: F)
with thickness of 6 mm; 12 mm or 19 mm

F_19mm E_19mm
with spacer: (symbol: D) F_2_D E_2_D
2 with resin: (symbol: R) F_2_R E_2_R
cop
with EVA foil: (symbol: F) F_2_F E_2_F

with spacer: (symbol: D) F_3_D E_3_D
3 with resin: (symbol: R) F_3_R E_3_R
with EVA foil: (symbol: F) F_3_F E_3_F
B1
Table B.7.2 Number of tested non heat-treated float glass specimens with thicknesses of 6,
12 and 19 mm (where symbols are: R - resin interlayer, F - EVA interlayer, D -
non bonded glass layers)
Number Type of Rate of loading (mm/min)
Type Temperature
of glass interlayer
of glass of specimens 50 20 5 1
layers material
- 20 °C 4
1 - + 23 °C 4 4 4 4
thickness of single glass layer: 6 mm
+ 60 °C 4
D + 23 °C 3 3
- 20 °C 3
DT
R + 23 °C 3
Float (symbol: F)
2 + 60 °C 3
- 20 °C 3
F + 23 °C 3
AR
+ 60 °C 3
D + 23 °C 3 3
- 20 °C 3
R + 23 °C 3
KH
3 + 60 °C 3
- 20 °C 3
F + 23 °C 3
+ 60 °C 3
AN
12 mm 1 - + 23 °C 4
19 mm 1 - + 23 °C 4
K.P
Table B.7.3 Number of tested tempered glass specimens with thicknesses of 6, 12 and 19 mm
(where symbols are: R - resin interlayer, F - EVA interlayer, D - non bonded
glass layers)
Number Type of Rate of loading (mm/min)
Type Temperature
of glass interlayer
t by
of glass of specimens 50 20 5 1
layers material
- 20 °C 4
1 - + 23 °C 4 4 4 4
thickness of single glass layer: 6 mm
+ 60 °C 4
h
D + 23 °C 3 3
- 20 °C 3
Tempered (symbol: E)
yrig
R + 23 °C 3
2 + 60 °C 3
- 20 °C 3
F + 23 °C 3
+ 60 °C 3
cop
D + 23 °C 3 3
- 20 °C 3
R + 23 °C 3
3 + 60 °C 3
- 20 °C 3
F + 23 °C 3
+ 60 °C 3
12 mm 1 - + 23 °C 4
19 mm 1 - + 23 °C 4
B2
Fig. B.7.2 Storage of the specimens
DT
AR
KH
Fig. B.7.3 Preparation of specimens before testing
AN
2
3 2 1
K.P 1
2
1
2
3 4
3
3 2 1
t by
1
Fig. B.7.4 Locations of thicknesses measurements
h
Table B.7.4 Measured parameters of single layer glasses

yrig
Average Dimension Nominal

Measured parameter
value value
density of glass ρglass 2.50 g/cm3 2.50 g/cm3
length of specimen L 1099 mm 1100 mm6 5 mm
cop
width of specimen b 358 mm 360 mm6 5 mm

thickness of specimen hnom,6 mm 5.87 mm 6 mm
thickness of resin interlayer hint Resin 1.00 mm 1 mm
thickness of EVA foil interlayer hint EVA 0.40 mm 0.4 mm
B3
DT
AR
Fig. B.7.5 Test arrangement Fig. B.7.6 Testing of heated or cooled glass
specimens
KH
AN
K.P
Fig. B.7.7 Specimens in heating Fig. B.7.8 Continuous temperature measurement

t by
chamber heated up to +60°C during testing

h
yrig
cop
Fig. B.7.9 Specimens in refrigerator Fig. B.7.10 Cooled float glass specimens
cooled down to -20°C after testing
B4
INSTRON testing
Digital instruments instrument TYPE 1197
to measure temperature
and humidity of testing
Self-designed
room
force transducer
HMB Spider 8 +
Catman software
Recording of
DT
data, with data
acquisition Camera
computer
AR
Extending of
channels of
Spider with MGC
KH
Fig. B.7.11 Test set-up
AN
Table B.7.5 Calibration to determine the loading rate
Calibration Time Strain (in mid-pane, Stress Stress at rate of
50 mm/min with rosette gauge) (2 60,4)
s µm/m N/mm2 N/(mm2 .s)
K.P
E_1_1_50 mm 20 388.114 27.168

60 1403.200 98.224 1.776
E_1_2_50 mm 20 438.171 30.672
60 1459.429 102.160 1.787
t by
F_1_1_50 mm 10 107.4286 7.53

20 335.0857 23.456 1.593
F_1_2_50 mm 10 165.2571 11.568
20 394.0571 27.584 1.602
Average 1.690
h
yrig
INSTRON
TYPE 1197
Test specimen
cop
Camera
Catman software
Version 3.5
Recording of test
data
Fig. B.7.12 Test set-up
B5
DT
AR
Fig. B.7.13 Regions 1 and 2 of glass surface strain measurements (Pankhardt, 2008a)
KH
B.2 to Chapter 8 Effect of tempering on load bearing capacity of single
glass pane in bending
AN
R1 R2
K.P
t by
Fig. B.8.1 Fracture Fig. B.8.2 Fracture pattern Fig. B.8.3 Fracture of single
pattern of float glass of painted tempered glass tempered glass
h
2
E_1_1_50 mm
yrig
E_1_2_50 mm
1.8 E_1_3_50 mm
E_1_4_50 mm
E_1_1_20 mm
E_1_2_20 mm
1.6 E_1_3_20 mm
E_1_4_20 mm
F_1_1_50 mm
1.4 F_1_2_50 mm
F_1_3_50 mm
F_1_4_50 mm
F_1_1_20 mm
Force, F (kN)
1.2
cop
F_1_2_ 20 mm
F_1_3_20 mm
F_1_4_20 mm
1
0.8
0.6
0.4
0.2
0
0 10 20 30 40 50 60 70
Deflection, y (mm)
Fig. B.8.4 Force vs. deflection of glass of 6 mm thickness, where symbols: E - tempered
glass; F - non heat-treated float glass; first numbering is the No. of the plies, second numbers
from 1 to 4 – number of individual specimen, 50 mm/min or 20 mm/min are the loading rates.
B6
250
Integral of force, F vs. deflection, y

Integral of F_y = 1.24h + 168.32
191.90
200 183.22
Tempered
diagram, (Nm)
Float
150
Adatsor3 Integral of F_y = 21.99h - 80.60
100
51.31 Integral of F_y = 2.95h - 12.97 42.79

50
23.16
4.36
DT
0
4 6 8 10 12 14 16 18 20
Thickness, h (mm)
Fig. B.8.5 Integral of force vs. deflection diagrams (external work) for tempered and float
AR
glass specimens tested in four-point bending up to failure
1.8
tempered_mid_50mm/min 2041.28; 1.62
KH
1841.26; 1.62
1.6 tempered_edge_50mm/min 1846.14; 1.58
tempered_mid_20mm/min
1.4 tempered_edge_20mm/min
float_mid_50mm/min 2063.92; 1.58
1.2 float_edge_50mm/min
Force, F (kN)
AN
float_mid_20mm/min
1
float_edge_20mm/min
+23 °C
0.8
514.17; 0.47 20 mm/min
0.6
508.80; 0.43 50 mm/min
K.P
565.17; 0.47
0.4 590.55; 0.43
0.2
0 0
0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200
Strain, ε ( µ m/m)
t by
Fig. B.8.6 Force vs. strain at bottom surface in Region 1 (middle) and Region 2 (edge) of
single layer tempered or float glass specimens, thickness of 6 mm tested in four-point
bending with loading rate of 20 mm/min or 50 mm/min.
h
yrig
3500 16
Float_mid 20 mm/min 15.52
Float_edge
Tempered_mid 2921.36 14
3000 +23 °C
Strains at maximal load, ε (µm/m)
Tempered_edge
Float_Max.Force
Tempered_Max.Force 2563.46
- 12
2500 2274.99
Force, Fmax (kN)
cop
2063.92 2075.88 10
2000 1846.14
7.94 7.21 8
1500
1157.63 1188.82 6
1053.2 1125.55
1000
4
590.55
508.8
500 1.58 2.66 2
0.43
0 0
6 mm 12 mm 19 mm
Thickness, h (mm)
Fig. B.8.7 Force vs. strain at bottom surface in Region 1 (middle) and Region 2 (edge)
of single tempered and float glass specimens with thicknesses of 6, 12 or 19 mm tested in
four-point bending with loading rate of 20 mm/min.
B7
10
tempered_Mid
tempered_Edge
float_Edge
Bending strength / Surface strength

float_Mid
Hatvány (tempered_Edge)
float
σ bB/σ
1
0 0
tempered 1 10 100
Thickness, h (mm)
DT
0
Fig. B.8.8 Ratio of bending strength to surface strength vs. thickness of single glasses,
AR
logarithmic scales
KH
AN
Fig. B.8.9 LEFM and Weibull-type size-effect Fig. B.8.10 Example of Malpasset Dam
K.P
laws (Bažant, 2004) and its size effect (Bažant, 2004)
Table B.8.1 Calculated Young’s modulus, bending strength and surface strength (average
for theoretical and for calculated values) of tested single glass specimens, where first symbols
are: E-tempered glass, F-non heat-treated glass, Numbers -No. of glass plies
t by
Theoretical Calculated Surface strength Bending strength,

surface strength Young's calculated with σbB
Specimens
20mm/min
(E=70000 N/mm2) modulus E = Efl with Eq. (6.11)

h Mid Edge Mid Edge Mid Edge
h
Efl
Bottom pane Bottom pane Bottom pane Bottom pane Bottom pane Bottom pane
yrig
mm N/mm2 N/mm2 N/mm2 N/mm2 N/mm2 N/mm2 N/mm2

6 129.2 144.5 83556 154.3 172.5 157.5 171.9
E_1 12 179.4 204.5 68725 176.2 200.8 191.4 210.5
19 145.3 159.2 60805 126.2 138.3 145.4 155.1
cop
6 35.62 41.34 86510 44.0 51.1 45.6 50.1

F_1 12 73.7 81.0 61879 65.2 71.6 65.4 70.0
19 78.8 83.2 59219 66.7 70.4 68.2 71.0
Fig. B.8.11 Large deflections by testing of 6 mm glass specimens in four-point bending
B8
Fig. B.8.12 Typical scratch pattern produced by grinding, with scratches extending over
width of glass edge
Stress distribution over thickness of glass layer, Fmax =0.43 kN Stress distribution over thickness of glass layer, Fmax =1.58 kN
single layer float glass, h = 6 mm, T = 23°C single layer tempered glass, h = 6 mm, T = 23°C
DT
6 6
-41.34 -35.62 Mid_20mm/min -144.47 -129.23 Mid_20mm/min
5 5
thickness (mm)
Edge_20mm/min
thickness (mm)
Edge_20mm/min
4 4
3 3
2 2
AR
1 1
35.62 41.34 129.23 144.47
0 0
-60 -40 -20 0 20 40 60 -200 -150 -100 -50 0 50 100 150 200
2 2
Stress (N/mm ) Stress (N/mm )
KH
Fig. B.8.13 Calculated ultimate surface stresses with Hooke’s law of 6 mm thick single
glass specimens a) float, b) tempered, tested with loading rate of 20 mm/min.
Stress distribution over thickness of glass layer, Fmax =2.66 kN Stress distribution over thickness of glass layer, Fmax =7.94 kN
AN
12 12
-81.03 -73.72 Mid_20mm/min -204.50 -179.44 Mid_20mm/min
10 10
thickness (mm)
thickness (mm)
Edge_20mm/min Edge_20mm/min
8 8
6 6
K.P
4 4
2 2
73.72 81.03 179.44 204.50
0 0
-100 -50 0 50 100 -300 -200 -100 0 100 200 300
2 2
Stre ss (N/mm ) Stre ss (N/mm )
Fig. B.8.14 Calculated ultimate surface stresses with Hooke’s law of 12 mm thick single glass
t by
specimens a) float, b) tempered, tested with loading rate of 20 mm/min.

Stress distribution over thickness of glass layer, Fmax=7.21 kN Stress distribution over thickness of glass layer, Fmax=15.51 kN
18 -145.31 18 Mid_20mm/min
-83.22 -78.79 Mid_20mm/min -159.25
h
16 16
thickness (mm)
thickness (mm)
Edge_20mm/min 14 Edge_20mm/min
14
12 12
yrig
10 10
8 8
6 6
4 4
2 83.22 2 145.31 159.25
78.79 0
0
-100 -50 0 50 100 -200 -150 -100 -50 0 50 100 150 200
2 2
cop
Fig. B.8.15 Calculated ultimate surface stresses with Hooke’s law of 19 mm thick single
glass specimens a) float, b) tempered, tested with loading rate of 20 mm/min.
Fig. B.8.16 a) electron-microscope type JEOL JSM-5500LV b) prepared specimens for

edge microscopy
B9
Fig. B.8.17 Typical edge in “as cut” samples a) surface defects ×50, b) surface of float
glass ×300
DT
AR
KH
Fig. B.8.18 Edgework of a glass pane a) manually arrised b) machine ground
AN
K.P
Fig. B.8.19 Defects, as initiators of cracks on manually arrised edge of a glass pane
h t by
a) b) c)
yrig
Fig. B.8.20 Edge of glass panes a) as cut, b) machine ground, c) polished edge
cop
Fig. B.8.21 Effect of coolant concentration vs. type on mean value of edge strength -
preliminary tests at Tampere University (Emonds et al., 1999)
B 10
B.3 to Chapter 9 Effect of lamination on bending characteristics
DT
Fig. B.9.1 Test of laminated Fig. B.9.2 Test of Fig. B.9.3 Test of non safety
glass specimens laminated safety glass laminated glass specimens
specimens
AR
KH
AN
K.P
Fig. B.9.4 Tested laminated safety glass pane

h t by
yrig
cop
Fig. B.9.5 Schematic force vs. deflection diagram at different stages during the fracture
process of a) single glass layer and of laminated glass b) consisting of two glass layers, c)
consisting of three glass layers
B 11
20.0
Fmax_tempered_6_12_19mm Fmax_float_6_12_19mm
Fmax_Tempered_50mm/min Fmax_Tempered_20mm/min 80
18.0 Fmax_Float_50mm/min Fmax_Float_20mm/min
y_tempered_6_12_19mm y_float_6_12_19mm
y_tempered_50mm/min y_tempered_20mm/min
y_float_50mm/min y_float_20mm/min 70
16.0

63.34 15.52
14.0 61.63 58.15 60
ultimate force, Fu (kN)

12.0 53.84 51.53 50
46.15 48.83
10.0
40
8.0 7.94
7.21 30
6.0 18.68 17.66
17.80 24.73
18.76 17.41 18.11 20
4.0 16.20 3.13 2.87 4.16
DT
1.62 3.99
2.0 1.58 2.66 11.87 10
0.47 1.01 1.46
0.43 0.97 1.40
0.0 0
0 1 2 3 4
AR
Fig. B.9.6 Force or deflection vs. number of glass layers of float and tempered laminated
glasses (with 2×6 mm or 3×6 mm glass layers) without interlayer material. Single glass with
thicknesses of 12 mm and 19 mm are also indicated. Dashed lines illustrate the deflection,
KH
continuous lines the force.
6.0 80
Fmax_Tempered_50mm/min Fmax_Tempered_20mm/min
+23 °C Fmax_Float_50mm/min Fmax_Float_20mm/min
y_tempered_50mm/min y_float_50mm/min 70
AN
y_float_20mm/min y_tempered_20mm/min
5.0
63.34
61.63 58.15 60
4.16
Deflection, y (mm)
4.0 3.99
ultimate force, Fu (kN)
50 mm/min 53.84 51.53 50

48.83
20 mm/min 3.13
K.P
3.0 40
2.87
30
2.0 1.62
18.68
1.58 17.80 18.11
18.76 17.66 20
16.20 1.46
1.0 1.40
1.01 10
0.47 0.97
t by
0.43
0.0 0
0 1 2 3 4
Fig. B.9.7 Force or deflection vs. number of glass layers of float and tempered laminated
h
glass (with 2×6 mm or 3×6 mm glass layers) without interlayer material,

enlargement from Fig. B.9.6
yrig
20.0 85
Fmax_tempered_6_12_19mm
18.0
Fmax_float_6_12_19mm +23 °C
y_tempered_6_12_19mm 75
y_float_6_12_19mm 20 mm/min
16.0 65
61.63
14.0 15.52
cop ultimate force, Fu (kN)
55
12.0 46.15
45
10.0
7.94 35
8.0 7.21
25
6.0 16.20 17.41 24.73
4.0 15
2.66
1.58 11.87 5
2.0
0.43
0.0 -5
4 6 8 10 12 14 16 18 20
thickness, h (mm)
Fig. B.9.8 Force or deflection vs. glass thickness of float and tempered single glass layers
B 12
8.00 18.00
F_1_6_12_19mm +23 °C E_1_6_12_19mm +23 °C
7.21 16.00 15.52
7.00 F_F_foil EVA E_F_foil EVA
15.01 20 mm/min
Ultimate force, Fu (kN) 20 mm/min 14.00 E_R_resin
F_R_resin
Ultimate force, Fu (kN)

6.00 5.85 E_D_only spacer 12.77
F_D_only spacer 12.00
5.00
4.49 10.00
4.00 7.94
8.00
2.66 6.64
3.00 6.00
2.30 6.30
2.00 4.00 3.99
2.05
1.40
1.00 2.00 2.87
1.01 1.62
0.47
0.00 0.00
1 2 3 4
DT
1 2 3 4
Left: Fig. B.9.9 Ultimate force vs. number of float glass layers in laminated glass,
Right: Fig. B.9.10 Ultimate force vs. number of tempered glass layers in laminated glass,
AR
where symbols mean: 6mm_12mm_19mm are the single layer glass specimens
18
+23 °C 15.52
E_19mm
16 E_3_F
20 mm/min
15.01
KH
14 E_2_F
E_12mm
12
E_6mm
Force, F (kN)
10 F_19mm
AN
7.94 F_2_F
8 7.60
7.21 6.64
F_3_F
6 5.85 F_12mm
5.26
F_6mm
K.P
4
3.12
2.30 2.66
2 2.35
0.68 0.74 1.88 2.13
0.43 1.58
0.410.54 0.85 0.96
0 0
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85
Deflection, y (mm)
t by
Fig. B.9.11 Force vs. deflection diagram of non heat-treated float (F_) and tempered (E_)
laminated glass (with 3×6 mm and 2×6 mm glass layers), with EVA foil (_F); also single
glass specimens with thicknesses of 6, 12 as well as 19 mm are illustrated. Dashed lines
illustrate the tempered, continuous lines, the float specimens.
h
18
yrig
+23 °C E_19mm
15.52
16 E_3_R
20 mm/min E_12mm
14 E_2_R
12.77 E_6mm
12
F_19mm
cop
Force, F (kN)
F_3_R
10
F_12mm
7.94
8 F_2_R
7.21 6.30 7.36
F_6mm
6
4.49
4.93
4 3.00
2.13 2.66
2 2.05
0.87 0.85 1.78 1.93
0.65 1.16 1.58
0.51 0.92
0 0
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80
Deflection, y (mm)
Fig. B.9.12 Force vs. deflection diagram of non heat-treated float (F_) and tempered (E_)
laminated glass (with 3×6 mm and 2×6 mm glass layers), with resin interlayer (_R).
B 13
Fig. B.9.13 Loaded laminate where x is Fig. B.9.14 Three-dimensional stress field at
DT
the distance from the edge (Lagunegrand a free edge laminate for an average loading
et al., 2006) of σzz=1 MPa (Lagunegrand et al., 2006)
Table B.9.1 Surface strength and calculated Young’s modulus of laminated glass specimens,
AR
where letters are: E – tempered glass, F – float non heat-treated glass, Numbers –No. of glass
plies, laminated with resin or EVA foil interlayer, loading rate of 20 mm/min.
Measured Theoretical surface strength Calculated Young's
laminated
Type of
Force Deflection (E = 70000 N/mm2) modulus

Inter-
layer
KH
glass
F y Region 1 Region 2 Efl

2 2
kN mm N/mm N/mm N/mm2
EVA 6.64 33.75 144.8 161.9 72466
AN
E_2
Resin 6.30 37.70 129.5 133.3 53275
EVA 2.30 11.46 48.4 52.0 73850
F_2
Resin 2.05 11.59 42.4 44.7 56279
K.P
EVA 15.01 26.57 154.9 161.9 59688

E_3
Resin 12.77 33.18 136.0 138.4 33639
EVA 5.85 10.22 58.2 65.4 60495
F_3
Resin 4.49 11.00 46.5 47.6 35689
t by
Stress distribution over thickness of glass layers, Fmax =1.01 kN Stress distribution over thickness of glass layers, Fmax=2.87 kN
non bonded float glass layers, h = 2×6 mm, T = 23°C non bonded tempered glass layers, h = 2×6 mm, T = 23°C
12 12
-34.61 Mid_20mm/min -112.20 Mid_20mm /min
-36.54 -116.77
10 Edge_20m m/min 10 Edge_20mm /m in
thickness (mm)
thickness (mm)
8 8
-53.60 -47.12
6
-133.43-129.80
6
h
47.12 53.60
4 116.77 129.80
4
2
yrig
2
43.51 49.94 133.43
0 111.96
0
-100 -50 0 2 50 100
Stress (N/mm ) -200 -100 0 2 100 200
Stre ss (N/mm )
Fig. B.9.15 Calculated ultimate surface stresses (Hooke’s law) of 2×6 mm thick laminated
glass without bonded glass layers a) float, b) tempered, loading rate of 20 mm/min.
cop
Stress distribution over thickness of glass layers, Fmax =1.40 kN Stress distribution over thickness of glass layers, Fmax=3.99 kN
non bonded float glass layers, h = 3×6 mm, T = 23°C non bonded tempered glass layers, h = 3×6 mm, T = 23°C
18 18
-47.61 -36.15 Mid_20mm/min -112.15 -108.71 Mid_20mm /min
16 16
Edge_20mm/min Edge_20mm /min
14 14
thickness (mm)
thickness (mm)
-126.98 12
-53.35 -53.35 12 36.15
47.61 -126.98
10 10 108.71 112.15
8
-46.07 8 -128.32
-53.03 6 -141.84 6 126.98
53.35
53.35 4 126.98
4
2 46.07 53.03 2 128.32 141.84
0 0
-200 -100 0 2 100 200 -200 -100 0 2
100 200
Fig. B.9.16 Calculated ultimate surface stresses (Hooke’s law) of 3×6 mm thick laminated
glass without bonded glass layers a) float, b) tempered, loading rate of 20 mm/min.
B 14
B.4 to Chapter 10 Effect of temperature on bending characteristics
1 Temperature at the window

2 Temperature between sunshade and inner pane
3 Ambient temperature
4 Room temperature
DT
Fig. B.10.1 Effect of ambient temperature on the temperature of window glazing (Schittich
et. al, 1999)
AR
KH
AN
Fig. B.10.2 Schematical views of observation locations of fractured area of
a) float and of b) tempered glass panes
K.P
t by
Fig. B.10.3 Fracture pattern of Fig. B.10.4 Fracture patterns of resin laminated glass
resin laminated glass at +60 °C specimens at -20 °C a) fractured specimen, b) regions
c) painted regions
h
yrig
cop
Fig. B.10.5 Fractured region (width: avg. Fig. B.10.6 Fractured region (width: avg.
34,5 cm) of resin laminated float glass 29 cm) of resin laminated float glass
consisting of two glass layers at -20 °C consisting of three glass layers at -20 °C
39 cm) of resin laminated float glass 35 cm) of resin laminated float glass
consisting of two glass layers at +60 °C consisting of three glass layers at +60 °C
B 15
34 cm) of EVA foil laminated float glass 42 cm) of EVA foil laminated float glass
consisting of two glass layers at -20 °C consisting of three glass layers at -20 °C
DT
AR
KH
34 cm) of EVA foil laminated float glass 42 cm) of EVA foil laminated float glass
consisting of two glass layers at +60 °C consisting of three glass layers at +60 °C
AN
K.P
Fig. B.10.13 Fractured regions (area 10×10 cm) of EVA foil laminated tempered glass
t by
consisting of three glass layers a) at -20 °C b) at +60 °C (from 10 to 20 cm at centre line,

arrow indicates the direction of support position)
h
yrig
cop
Fig. B.10.14 Idealised behaviour of a reinforced concrete tie for tension-stiffening effect
(CEB-FIP Model Code, 1990)
B 16
Table B.10.1 Effect of temperature on strains and surface stresses at bottom of

single layer glass with 6 mm thickness, loading rate of 50 mm/min
Mid Edge Young’s Mid Edge
Region 1 Region 2 modulus Region 1 Region 2
Specimens
Fmax y εmax.,avg. E σmax.,avg
kN mm µm/m N/mm2 ×103 N/mm2
-20 °C 0.79 28.28 621.20 763.22 70 43.5 53.4
F 0.47 18.76 514.17 565.17 70 36.0 39.6
23 °C
float 0.54 22.16 658.72 727.46 70 46.1 50.9
60 °C
-20 °C 1.56 59.18 1531.63 1687.75 70 107.2 118.1
DT
E 1.62 63.34 1841.26 2041.28 70 128.9 142.9
23 °C
tempered 1.58 60.01 2056.66 2137.75 70 144.0 149.6
60 °C
AR
Table B.10.2 Average of measured values of maximal load of single layer glasses and
laminated glasses, where the symbols mean: F_- float glass, E_- tempered glass; numbers: -number of
glass layers; _F -EVA foil interlayer material, _R -resin interlayer material, _D - without interlayer material
Float Tempered
KH
Temp. Ultimate force, Fu (kN) Temp. Ultimate force, Fu (kN)
(°C) Number of glass layers (°C) Number of glass layers
1 2 3 1 2 3
AN
-20 2.30 5.12 -20 7.16 14.49
F_F 23 2.30 5.85 E_F 23 6.64 15.01
60 2.25 4.07 60 6.58 13.01
-20 2.75 5.46 -20 7.81 18.85
K.P
F_R 23 2.05 4.49 E_R 23 6.30 12.77

60 1.66 3.19 60 5.35 10.40
-20 0.79 -20 1.56
F_D 23/50mm 0.47 0.97 1.46 E_D 23/50mm 1.62 3.13 4.16
F_6mm 23/20mm 0.43 1.01 1.40 E_6mm 23/20mm 1.58 2.87 3.99
t by
60 0.54 60 1.58
F_12 mm 23 2.66 E_12 mm 23 7.94
F_19 mm 23 7.21 E_19 mm 23 15.52
Table B.10.3 Average of measured values of deflections at maximal force of single layer
h
glasses and laminated glasses

yrig
Float Tempered
Temp. Deflection, y (mm) Temp. Deflection, y (mm)
(°C) Number of glass layers (°C) Number of glass layers
1 2 3 1 2 3
cop
-20 10.96 8.35 -20 33.55 22.41

F_F 23 11.46 10.22 E_F 23 33.75 26.57
60 16.03 11.92 60 43.23 32.36
-20 12.19 8.10 -20 32.47 24.93
F_R 23 11.59 11.00 E_R 23 37.70 33.18
60 15.18 15.88 60 51.02 46.57
-20 28.28 -20 59.18
F_D 23/50mm 18.76 17.80 18.11 E_D 23/50mm 63.34 58.15 51.53
F_6mm 23/20mm 16.20 18.68 17.66 E_6mm 23/20mm 61.63 53.84 48.83
60 22.16 60 60.01
F_12 mm 23 17.41 E_12 mm 23 46.15
F_19 mm 23 11.87 E_19 mm 23 24.73
B 17
Table B.10.4 Measured and calculated values for load and deflection of laminated glasses
consisting of two glass layers, where the symbols mean: E_- tempered glass, F_- float glass;
number: e.g._ 2_- number of glass layers; _F- EVA foil interlayer material, _R- resin
interlayer material, _D – without interlayer material
Measured values Calculated Calculated
Laminated glass Force Deflection Force Deflection
consisting of two (avg.) (avg.) F y
glass layers Fmax ymax at Ls/200 = 5 mm at F = 0.4 kN
kN mm kN mm
-20°C 7.81 32.47 1.20 1.66
DT
E_2_R 23°C 6.30 37.70 0.84 2.39
60°C 5.35 51.02 0.52 3.81
-20°C 2.75 12.19 1.13 1.77
AR
F_2_R 23°C 2.05 11.59 0.88 2.27
60°C 1.66 15.18 0.55 3.66
-20°C 7.16 33.55 1.07 1.87
E_2_F 23°C 6.64 33.75 0.98 2.03
KH
60°C 6.58 43.23 0.76 2.63
-20°C 2.30 10.96 1.05 1.91
F_2_F 23°C 2.30 11.46 1.00 2.00
60°C 2.25 16.03 0.70 2.85
AN
E_2_D 23°C 2.87 53.84 0.27 7.50
F_2_D 23°C 1.01 18.68 0.27 7.40
K.P
Table B.10.5 Measured and calculated values for load and deflection of laminated glasses
consisting of three glass layers, where the symbols mean: E_- tempered glass, F_- float glass;
number: e.g._ 3_- number of glass layers; _F- EVA foil interlayer material, _R- resin
interlayer material, _D - without interlayer material
t by
Measured values Calculated Calculated

Laminated glass Force Deflection Force Deflection
consisting of three (avg.) (avg.) F y
glass layers Fmax ymax at Ls/200 = 5 mm at F = 0.4 kN
h
kN mm kN mm
yrig
-20°C 18.85 24.93 3.78 0.53

E_3_R 23°C 12.77 33.18 1.92 1.04
60°C 10.40 46.57 1.12 1.79
-20°C 5.46 8.10 3.37 0.59
F_3_R 23°C 4.49 11.00 2.04 0.98
cop
60°C 3.19 15.88 1.00 1.99

-20°C 14.49 22.41 3.23 0.62
E_3_F 23°C 15.01 26.57 2.82 0.71
60°C 13.01 32.36 2.01 0.99
-20°C 5.12 8.35 3.07 0.65
F_3_F 23°C 5.85 10.22 2.86 0.70
60°C 4.07 11.92 1.71 1.17
E_3_D 23°C 3.99 48.83 0.41 4.90
F_3_D 23°C 1.40 17.66 0.40 5.05
B 18
250%
E_2_R F_2_R E_2_F F_2_F
at deflection of Ls /200 = 5mm, %

E_3_R F_3_R E_3_F F_3_F
197% 200%
Change of force, ∆F,

165%
143% 150%
128%
115%
109%
107% 100% 100% 78%
105% 71%
70%
63%
50% 62%
60%
58%
49%
0%
DT
-40 -20 0 20 40 60 80
Temperature, T , °C
Fig. B.10.15 Change of force with change of temperature at a deflection of Ls/200 = 5mm
AR
250%
E_2_R F_2_R E_2_F F_2_F
E_3_R F_3_R E_3_F F_3_F
at force level of F=0.4 kN, %
200% 203%
Change of deflection, ∆y,
172%
167%
150% 161%
KH
159%
143%
139%
93% 96% 100% 100% 130%
92%
87%
78%
69%
60% 50%
AN
51%
0%
-40 -20 0 20 40 60 80
Temperature, T , °C
K.P
Fig. B.10.16 Change of deflection with change of temperature at a force level of 0.4 kN
8
20 mm/min F_19mm
7.21 F_3_F
7 F_3_R
F_12mm
ultimate Force at Stage A , Fu (kN)
5.85 EVA F_2_F

5.46 6 F_2_R
F_3_D
Resin F_2_D
t by
5 F_6mm
5.12 4.49 Expon. (F_2_R)
4 4.07
3.19
2.75 3 2.66
2.30 2.25
2.30 2
2.05
1.40
h
1 1.01 1.66
0.79 0.47 0.54
0
yrig
-30 -20 -10 0 10 20 30 40 50 60 70 80

Temperature, T °C
a)
25
E_19mm
20 mm/min E_3_F
E_3_R
E_12mm
ultimate Force at Stage A , Fu (kN)
18.85 20
cop
E_2_F
E_2_R
Resin 15.52 EVA E_3_D
E_2_D
14.49 15 E_6mm
15.01 Polinom. (E_3_F)
13.01
12.77
0 10.40
7.81 10
7.94
6.64 6.58
7.16 6.30
5
3.99 5.35
2.87
1.56 1.62 1.58
0
-30 -20 -10 0 10 20 30 40 50 60 70 80
Temperature, T °C
b)
Fig. B.10.17 Ultimate force of a) laminated float (F_) glasses and of b) laminated
tempered (E_) glasses consisting of two (_2_) or three (_3_) glass layers and with resin (_R)
or EVA foil (_F) interlayer material (Pankhardt, Balázs, 2006)
B 19
40
F_6mm 20 mm/min
ultimate deflection at Stage A , yu (mm)

F_2_D
35 F_3_D
F_19mm
F_12mm
30 F_2_F
28.28 F_2_R
F_3_R
25 F_3_F 22.16
Polinom. (F_3_R)
20 18.76
17.66 18.68 16.03
15.88
15 Resin 17.41 11.87
12.19 15.18
11.59
10.96 10 11.92
8.10 11.46
8.35 11.00 EVA
5 10.22
DT
0
-30 -20 -10 0 10 20 30 40 50 60 70 80
Temperature, T °C
a)
AR
70
63.34
59.18 60.01
ultimate deflection at Stage A , yu (mm)
60
53.84 Resin 51.02
50 48.83 46.57
KH
46.15
37.70 43.23
40
32.47 33.75 32.36
33.18
33.55 30 26.57
24.93 E_6mm
24.73 E_2_D
20 EVA E_3_D
22.41
AN E_12mm
E_2_R
20 mm/min E_3_R
10 E_2_F
E_3_F
E_19mm
0 Polinom. (E_2_F)
-30 -20 -10 0 10 20 30 40 50 60 70 80
K.P
Temperature, T °C
b)
Fig. B.10.18 Ultimate deflection at maximal force of a) laminated float (F_) glasses and of
b) laminated tempered (E_) glasses consisting of two (_2_)or 3 (_3_) glass layers and with
resin (_R) or EVA foil (_F) interlayer material, also of single layer glasses
t by
180% change of force F_2_EVA 180% change of force E_2_EVA

Change of ultimate force, ∆Fu, and deflection, ∆yu
y = 0.749e 0.194x 163% y = 0.787e 0.167x 160%

R 2 = 0.930 change of deflection R 2 = 0.902
160% change of deflection 160%
140%
140% 140% Mozgó átl. 2 sz. 128%
h
120% 120% 108%

100% 100% 100% 100%
100% 98% 99% 99%
100% 96% 100%
yrig
80% 80%
60% 44% 60%

43%
40% 40%
20% 20%
cop
0% 0%
-20 °C +23 °C +60 °C no bond -20 °C +23 °C +60 °C no bond
Fig. B.10.19.a Change of ultimate force and Fig. B.10.19.b Change of ultimate force and
deflection of laminated glass consisting of two deflection of laminated glass consisting of two
float glass layers and EVA interlayer tempered glass layers and EVA interlayer
B 20
200% change of force F_3_EVA 200% change of force E_3_EVA

184%
y = 0.622e 0.240x 173% y = 0.622e 0.253x
180% change of deflection 2
180% change of deflection R 2 = 0.952
R = 0.956
160% 160%
140% 140%
122%
120% 117% 120%
100% 100%
97% 100% 100%
100% 88% 82% 100% 84% 87%
80% 70% 80%
60% 60%
40% 24% 40% 27%
20% 20%
0% 0%
DT
Fig. B.10.19.c Change of ultimate force and Fig. B.10.19.d Change of ultimate force and
deflection of laminated glass consisting of deflection of laminated glass consisting of
three float glass layers and EVA interlayer three tempered glass layers and EVA
AR
interlayer
KH
180% change of force F_2_Resin change of force E_2_Resin
0.155x y = 0.834x 0.389
y = 0.828e 161% 160%

change of deflection R 2 = 0.834 change of deflection R 2 = 0.932 143%
160% 135%
140% 124% y = -0.250x + 1.512
140% 134% y = -0.274x + 1.595 131%
R 2 = 0.989 R 2 = 0.965
Change of ultimate force, ∆Fu,
120%
120% 100% 100%
105%
100% 100%
AN and deflection, ∆yu
100%
100% 86% 85%
81%
80% 80%
60% 49% 60%

46%
40% 40%
K.P
20% 20%
0%
0%
-20 °C +23 °C +60 °C no bond
-20 °C +23 °C +60 °C no bond
Fig. B.10.20.a Change of ultimate force and Fig. B.10.20.b Change of ultimate force and
deflection of laminated glass consisting of two deflection of laminated glass consisting of two
t by
float glass layers and resin interlayer tempered glass layers and resin interlayer
180% change of force F_3_Resin y = 0.718x 0.586 160% change of force E_3_Resin
161% 148% 147%

h
R 2 = 0.976 change of deflection 140%

160% change of deflection 140%
144% y = -0.368x + 1.820 y = 0.741x 0.515
Hatvány (change of
yrig
140% 122% y = -0.300x + 1.560 R 2 = 0.973 R 2 = 0.968

120%
R 2 = 0.982
120% 100% 100%
100% 100% 100%
81%
100% 75%
74% 71% 80%
80%
60%
60%
cop
31%
40% 40% 31%
20% 20%
0% 0%
Fig. B.10.20.c Change of ultimate force and Fig. B.10.20.d Change of ultimate force and
deflection of laminated glass consisting of deflection of laminated glass consisting of
three float glass layers and resin interlayer three tempered glass layers and resin
interlayer
B 21
Table B.10.6 Average of measured strains in Region 1 (mid-pane) at maximal force of single
layer glasses and laminated glasses, where the symbols mean: F_- float glass, E_- tempered
glass; numbers: -number of glass layers; _F- EVA foil interlayer material, _R- resin
interlayer material, _D - without interlayer material
Float Tempered
Temp. Strain, ε (µm/m) Temp. Strain, ε (µm/m)
Number of applied glass Number of applied glass
(°C) (°C)
layers layers
1 2 3 1 2 3
-20 615.96 654.52 -20 1985.24 1896.47
DT
F_F 23 691.92 831.26 E_F 23 2068.80 2213.37
60 785.01 666.17 60 2238.83 2083.57
-20 696.23 570.64 -20 2042.80 2191.11
F_R 23 605.48 664.19 E_R 23 1849.51 1942.37
AR
60 622.60 714.76 60 2087.18 2211.38
-20 621.20 -20 1531.63
F_D 23/50mm 514.17 471.82 878.21 E_D 23/50mm 1841.26 1840.54 1607.53
F_6mm 23/20mm 508.80 621.61 543.30 E_6mm 23/20mm 1846.14 1599.38 1738.56
KH
60 658.72 60 2056.66
F_12 mm 23 1053.20 E_12 mm 23 2563.46
F_19 mm 23 1125.55 E_19 mm 23 2075.88
AN
Table B.10.7 Average of measured strains in Region 2 (edge) at maximal force of single layer
glasses and laminated glasses, where the symbols mean: F_- float glass, E_- tempered glass;
K.P
numbers: -number of glass layers; _F- EVA foil interlayer material, _R- resin interlayer
material, _D - without interlayer material
Float Tempered
Temp. Strain, ε (µm/m) Temp. Strain, ε (µm/m)
Number of applied glass Number of applied glass
t by
(°C) (°C)
layers layers
1 2 3 1 2 3
-20 646.14 678.17 -20 2027.65 2092.19
F_F 23 742.49 934.14 E_F 23 2313.32 2312.39
h
60 836.50 657.55 60 2415.50 2262.52

-20 706.25 562.49 -20 2279.30 2569.40
yrig
F_R 23 638.10 679.81 E_R 23 1904.85 1977.44

60 720.70 701.36 60 2360.85 2136.00
-20 763.22 -20 1687.75
F_D 23/50mm 565.17 575.04 926.56 E_D 23/50mm 2041.28 2148.00 1970.10
cop
F_6mm 23/20mm 590.55 713.40 658.14 E_6mm 23/20mm 2063.92 1906.14 2026.25
60 727.46 60 2137.75
F_12 mm 23 1157.63 E_12 mm 23 2921.36
F_19 mm 23 1188.82 E_19 mm 23 2274.99
B 22
DT
AR
Fig. B.10.21 Ratio of shear deflections to total deflections (yshear/ytotal) as function of ratio of
span length to thickness (Ls/h) in three-point bending with tests of fibre-reinforced polymers
(Whitney, 1987)
20 mm/min
50 mm/min
3500
F_2_F_mid
F_2_F_edge
E_2_F_mid
KH 20 mm/min
50 mm/min
3500
F_3_F_mid
F_3_F_edge
E_3_F_mid
Ultimate strain at Stage A , ε u (µm/m)
3000 3000
AN
Ultimate strain at Stage A , εu (µm/m)
E_2_F_edge E_3_F_edge
Polinom. (E_2_F_mid) 2415.50 Polinom. (E_3_F_mid)
2313.32 2262.52
2500 2500 2312.39
2027.65 2092.19
2000 2238.83 2000 2213.37
2068.80 Region 2 Region 2
1985.24 2083.57
Region 1 1896.47
Region 1
1500 Region 2 1500 Region 2
K.P
Region Region 1
836.50 934.14
1000 1 742.49 1000
646.14 678.17 666.17
785.01 831.26
500 691.92 500
615.96 654.52 657.55
0 0
-30 -20 -10 0 10 20 30 40 50 60 70 -30 -20 -10 0 10 20 30 40 50 60 70
Temperature, T (°C) Temperature, T (°C)
t by
a) b)
Fig. B.10.22 Strain at maximal force (Stage A) in Region 1 and 2 of laminated glass with
EVA foil (_F) interlayer material a) consisting of two glass layers or b) of three glass layers
h
3500 3500
yrig
20 mm/min E_2_R_edge 20 mm/min E_3_R_edge

E_2_R_mid E_3_R_mid
Ultimate strain at "Stage A", ε u (µm/m)
50 mm/min 50 mm/min
Ultimate strain at "Stage A", ε u (µm/m)
3000 F_2_R_edge 3000 F_3_R_edge

F_2_R_mid F_3_R_mid
Polinom. (E_2_R_mid) 2569.40 Polinom. (E_3_R_mid)
2279.30 2500 Tempered
2360.85 2500 Tempered
2191.11 2211.38
1904.85 1977.44
2000 2000 2136.00
2042.80 2087.18 1942.37
1849.51 Region 2 Region 2
cop
1500 1500
Region 1 Region 2 Region 1
Region 2
1000 Region 1 720.70 1000 Region 1 Float
706.25 638.10 Float 570.64 679.81 714.76
696.23 500 500 664.19 701.36
605.48 622.60 562.49
0 0
-30 -20 -10 0 10 20 30 40 50 60 70 -30 -20 -10 0 10 20 30 40 50 60 70
Temperature, T (°C) Temperature, T (°C)
a) b)
Fig. B.10.23 Strain at maximal force (Stage A) in Region 1 and 2 of laminated glass with
resin (_R) interlayer material a) consisting of two glass layers or b) of three glass layers
B 23
B.5 to Chapter 11 Flexural stiffness of laminated glasses
DT
AR
Fig. B.11.1 Interlayer dependent bending behaviour of laminated glasses
KH
AN
n
∑z E h
i =1
i i i
=0 (B.11.1)
K.P
Fig. B.11.2 Sandwich model, where hi is the thickness of the ith layer in sandwich
(Hegedűs, 1983)
t by
The effective thickness of layered sandwich at glass transition temperature, Tg, is:
n
h
∑ E h (h i i i
2
+ 12zi2 )
heff ,T =
yrig
i =1
n
(B.11.2)
∑E h
g
i i
i =1
cop
The effective Young’s modulus of layered sandwich at glass transition temperature, Tg, is:
n
∑E h i i
Eeff ,T = i =1 (B.11.3)
g
heff ,T g
B 24
σb,service,Region1 = bending stress of laminated glass in Region 1 (middle of pane) in the case
of four-point bending at service temperature is:
 3( Ls − Lb ) 
σb ,service, Re gion1 = FTservice ⋅ 2
+ σbG  (B.11.4)
 2 bh eff ,T 
The modified bending stress (Eq. 6.12) imposed by the self-weight at service temperature
is:
DT
3 ρgL2s
σbG = (B.11.5)
4 heff ,T
AR
Table B.11.1 Calculation of effective thickness, heff, at service temperature, Tservice, of
laminated glasses consisting of two or three glass layers with 6 mm thickness and with
KH
EVA or resin interlayer material
Type of Number of Service temperature Calculation of
interlayer glass layers, ranges, the effective thickness,
Eq.
AN
material n Tservice heff
pcs °C mm
2 -20 ≤ T ≤ + 60 °C -0.016∙T+12.803 (B.11.6)
EVA
3 -20 ≤ T ≤ + 60 °C -0.023∙T+18.630 (B.11.7)
K.P
2 -20 ≤ T ≤ + 60 °C -0.030⋅T + 13.096 (B.11.8)

Resin
3 -20 ≤ T ≤ + 60 °C -0.066⋅T + 19.067 (B.11.9)
t by
Table B.11.2 Calculation of effective thickness, heff, at service temperature, Tservice, of

laminated glasses consisting of two or three glass layers with 6 mm thickness and with
EVA or resin interlayer material
h
Type of Number of Service Calculation of

yrig
interlayer glass layers, temperature the effective E Modulus,

material n ranges, Eq.
pcs Tservice Eeff
°C N/mm2
2 -20 ≤ T ≤ + 60 °C - 4.4685 ⋅ T 2 + 250.59 ⋅ T + 64550 (B.11.10)
cop
EVA
3 -20 ≤ T ≤ + 60 °C - 6.0495 ⋅ T 2 + 116.84 ⋅ T + 62552 (B.11.11)
2 -20 ≤ T ≤ + 60 °C - 150.594 ⋅ T + 59 886 (B.11.12)
Resin
3 -20 ≤ T ≤ + 60 °C - 150.800 ⋅ T + 51828 (B.11.13)
B 25
100 100 000

heff_E_2_EVA heff_E_3_EVA
heff_F_2_EVA heff_F_3_EVA
90 Eeff_E_2_EVA Eeff_E_3_EVA 90 000
Eeff_F_2_EVA Eeff_F_3_EVA
Polinom. (Eeff_F_3_EVA) Polinom. (Eeff_F_2_EVA)
Effective E-modulus, Eeff (MPa)

80 80 000
E eff,2,EVA = -4.4685 T 2 + 250.59T + 64550 E eff,T
70 70 000
Effective thickness, heff (mm)
60 60 000
E eff,3,EVA = -6.0495 T 2 + 116.84 T + 62552
50 50 000
DT
40 40 000
2 glass layers 3 glass layers
30 30 000
heff,T
heff,3,EVA = -0.023T + 18.630
20 20 000
AR
10 10 000
heff,2,EVA = -0.016T + 12.803
0 0
KH
-30 -20 -10 0 10 20 30 40 50 60 70
Fig. B.11.3 Effective thickness vs. temperature and effective E-modulus vs. temperature
diagrams of EVA foil laminated glasses
AN
K.P
100 100 000

heff_E_2_R heff_E_3_R
heff_F_2_R heff_F_3_R
90 Eeff_E_2_R Eeff_E_3_R 90 000
Eeff_F_2_R Eeff_F_3_R
80 Lineáris (Eeff_F_2_R) Lineáris (Eeff_E_3_R) 80 000
Effective E-modulus, Eeff (MPa)

E eff,T E eff,2,resin = - 150.594 T + 59 886

70 70 000
t by
2
R = 0.998
60 60 000
50 50 000
40 2 glass layers E eff,3,resin = -150.800 T + 51828 40 000
h
2 3 glass layers
R = 0.8964
30 30 000
yrig
heff,T
heff,3,resin = - 0.066T + 19.067
20 20 000
10 10 000
heff,2,resin = - 0.030T + 13.096
0 0
cop
-30 -20 -10 0 10 20 30 40 50 60 70

Fig. B.11.4 Effective thickness vs. temperature and effective E-modulus vs. temperature
diagrams of resin laminated glasses
B 26
Example: Calculation of safety laminated glass consisting of three glass layers with EVA
foil at service temperature, two sided supported, loaded in four-point bending.
- Dimensions and type of laminated glass:

b = 358 mm; L = 1100 mm; h = 5.87 mm glass layer, hint=0.4 mm EVA foil interlayer.
- Service temperature range and material properties:
Tservice = from +23 °C to +60 °C; Tg,EVA = -28 °C; Td,EVA = + 92 °C;
Tm,EVA = + 76-78 °C; EEVA,Tg= 1.5 MPa; Eglass = 70000 MPa.
DT
- Type of support and allowed deformation:
two sided supported (at shorter sides); Ls = 1000 mm, Lb = 200 mm; yallowed = Ls/200 =
5 mm.
AR
Solution:
Flexural stiffness of laminated glass at Td (“lower limit”):
KH
- heff,Td = 8.47 mm, calculated with Eq. (11.5)
- EgIg= (EgIg)Td = 1.27 × 109 Nmm2, calculated with Eq. (11.4)
AN
Flexural stiffness of laminated glass at Tg (“upper limit”):
- Sandwich model (Hegedűs, 1983) can be applied only when the interlayer is in glassy
state (see Fig. B.11.2)
K.P
- Effective thickness of layered sandwich at Tg:

n
∑ E h (h i i
2
i + 12zi2 )
heff ,Tg = i=1
n
= 18.68 mm (B.11.14)
∑E h
t by
i i
i =1
- Effective Young’s modulus of layered sandwich at Tg:

n
∑E h i i
Eeff ,Tg = i=1
h
= 65989 N/mm2 (B.11.15)

heff,Tg
yrig
- (EI)Tg = 12.83 × 109 Nmm2, calculated with sandwich model (Hegedűs, 1983)
Flexural stiffness of laminated glass at service temperature, Tservice:

- determine coupling parameter κ:
cop
o κ3,EVA = -0.0001T2 - 0.0018T + 0.9752 , calculated with Eq.(11.7)

o κ+23°C = 0.8809
o κ+60°C = 0.5072
B 27
- modified flexural stiffness at + 23 °C:

o κ3,EVA,+23 °C ×(EcIc)=0.8809×( 12.83 ×109 -1.27 × 109)=10.19×109 Nmm2,
calculated with Eq. (11.3)
- modified flexural stiffness at + 60 °C:
o κ3,EVA,+60 °C ×(EcIc)=0.5072×( 12.83 ×109 -1.27 × 109)=5.87×109 Nmm2,
calculated with Eq. (11.3)
Flexural stiffness of laminated glass at +23 °C service temperature:

(EI)+23°C = 1.27 × 109 Nmm2+ 10.19×109 Nmm2= 11.46×109 Nmm2
DT
Flexural stiffness of laminated glass at +60 °C service temperature:
(EI)+60°C = 1.27× 109 Nmm2+ 5.87×109 Nmm2= 7.13×109 Nmm2
AR
Calculation of force for deflection yLs/200 ≤ Ls/200 by four-point bending
(EN 1288-3:2000):
( y allowed − y G ,Tservice ) ⋅ 16 ⋅ (EI) Tservice
FTservice , y = (B.11.16)
 LS3 L b 3 L S L b 2 
KH
allowed
 + − 
 3 6 2 
where, yallowed is the allowed deflection e.g. yallowed = Ls/200 and yG is the deflection
AN
imposed by self weight at service temperature:
5 q ⋅ L4s
yG,Tservice = ⋅ (B.11.17)
384 (EI)Tservice
K.P
Maximal allowed force to allowed maximal deflection Ls/200

at service temperature +23 °C:
FLs/200, +23°C = 2808 N = 2.81 kN
t by
Maximal allowed force to allowed maximal deflection Ls/200

at service temperature +60 °C:
FLs/200, +60°C = 1709N = 1.71 kN
h
Table B.11.3 Comparing of measured and calculated values of FLs/200

yrig
Force EVA Resin

Control
(kN) -20 °C 23 °C 60 °C -20 °C 23 °C 60 °C
Consisting Calculated 0.98 0.96 0.70 1.17 0.81 0.49
cop
of two Measured E: 1.07 0.98 0.76 1.2 0.84 0.52

glass F: 1.05 1 0.7 1.13 0.88 0.55
layers Error (%) 6.67% 2.04% 0.00% -3.54% 3.57% 5.77%
Consisting Calculated 3.07 2.81 1.71 3.3 2.04 0.96
of three Measured E: 3.23 2.82 2.01 3.78 1.92 1.12
glass F: 3.07 2.86 1.71 3.37 2.04 1.00
layers Error (%) 0.00% 0.35% 0.00% 2.08% -6.25% 4%
Appendix B Tables B.10.4 and B.10.5 give measured force values to given deflections
(Ls/200 = 5 mm) as a comparison.
B 28
Maximal resisting force of EVA laminated safety glass consisting of 3 glass layers at
service temperature
30
heff_E_2_R heff_E_3_R
heff_F_2_R heff_F_3_R
heff_E_2_EVA heff_E_3_EVA
25 heff_F_2_EVA heff_F_3_EVA 20 mm/min
Lineáris (heff_E_3_R) Lineáris (heff_E_2_R)
heff,3,resin = - 0.066T + 19.067
heff,3.EVA = - 0.023T + 18.630
20 3 glass
layers
EVA Resin
15
2 glass layers
DT
10
heff,2,resin = - 0.030T + 13.096
heff,2,EVA = - 0.016T + 12.803
AR
5
0
-25 -20 -15 -10 -5 0 5 10 15 20 25 30 35 40 45 50 55 60 65
KH
Fig. B.11.5 Effective thickness vs. temperature of EVA foil and resin interlayer laminated
glasses
AN
- at +23 °C service temperature:
- heff ,3,EVA = -0.023×23 + 18.630 = 18.10 mm
- k1= 0.93 (Fig. 11.5)
K.P
- σb,G,T=23°C = 1.04 N/mm2, calculated with Eq. (B.11.5)

- calculation with use of Eq. (B.11.4) follows:
 3 ⋅ (1000 − 200 ) 
150 ⋅ 0 .93 =  FRd , Tservice ⋅ + 1 .04  , (B.11.18)
 2 ⋅ 358 ⋅ 18 .10 2

t by
FRd,T=23°C = 13532 N = 13.53 kN

With measured surface strength (161.9 N/mm2) instead of 150 N/mm2:
FRd,T=23°C= 14614 N = 14.61 kN
Measured value: 15.01 kN
h
Fy=Ls/200, +23°C = 2.78 kN < FRd,T=23°C= 13.53 kN

yrig
- at +60 °C service temperature:

- heff ,3,EVA = -0.023×60 + 18.630 = 17.25 mm
- k1= 0.92 (Fig. 11.6)
- σb,G,T=60°C = 1.09 N/mm2, calculated with Eq. (B.11.5)
cop
- calculation with use of Eq. (B.11.4) follows:

 3 ⋅ (1000 − 200 ) 
150 ⋅ 0 . 92 =  FRd ,Tservice ⋅ + 1 . 09  , (B.11.19)
 2 ⋅ 358 ⋅ 17 .25 2

FRd,T=60°C = 12153 N = 12.15 kN
With measured surface strength (158.4 N/mm2) instead of 150 N/mm2:
FRd,T=60°C= 12840 N = 12.84 kN
Measured value: 13.01 kN
Fy=Ls/200, +60°C = 1.58 kN < FRd,T=60°C= 12.15 kN
B 29
Calculation of stresses
Stress of lower surface of glass layer in Region 2 (edge) at +60 °C service temperature by
allowed deflection (Ls/200 = 5mm):
- heff ,3,EVA,T = -0.023×60 + 18.630 = 17.25 mm
- σb,G,60°C = 1.09 N/mm2
- calculation with use of Eq. (11.10) follows:
 3 ⋅ (1000 − 200 ) 
1580 ⋅ + 1 .09 
2 ⋅ 358 ⋅ 17 .25 2
 = 20.53 N/mm2
σ Ls / 200 , + 60 ° C = (B.11.20)
0 .92
DT
By comparing this value with given values of design bending strength of laminated glass
in Table 6.1 (by TRLV) it can be stated, that this kind of glazing with the applied loading
condition can not be used in overhead areas (σbB=15 N/mm2), only in vertical glazing
(σb,B=22.5 N/mm2) at service temperature of +60 °C.
AR
σbB, = bending strength of outer layer, in the case of four-point bending
(EN 1288-3: 2000) of float glass with thickness of 6 mm is about 45 N/mm2 and it is
about 150 N/mm2 in the case of tempered glass.
KH
Control with 2D FEM model:
With applied effective thickness and effective E-modulus at service temperature
Tservice= +60 °C; heff,T=60°C = 17.25 mm and
AN
Eeff,T=60°C = 42979 N/mm2 (see Figs. B.11.3 and B.11.4)
the calculated bending stress at 5 mm deflection in the outer layer of lower surface is:
Sxx =19.31 N/mm2 ≈ σLs/200, +60°C = 20.53 N/mm2
K.P
eZ [mm]
-0,500
-1,000
-1,500
-2,000
-2,499
-2,500
-3,000
t by
-3,500
-4,000
-4,500
0
2,00
4,00
6,00
8,00
10,00
12,00
14,00
16,00
18,00
Sxx B [N/mm^2]
9,49
h
Fig. B.11.6 Calculated surface stress (Sxx=19.31 N/mm2) at deflection yLs/200 = 5mm
yrig
cop
Fig. B.11.7 Distribution of maximum Fig. B.11.8 Simply supported pane with
principal stresses, simply supported pane distributed load, membrane effects and
(quarter of a pane) (Vallabhan et al., 1993) transverse shear of the strip-wise plate
(Czarnecki et al., 2008)
B 30
B.6 to Chapter 12 Residual load bearing capacity of laminated glass

panes
DT
AR
Fig. B.12.1 Stages during fracture process of laminated glass and numbering of stages of
change in force a) three glass layers, b) two glass layers
Stage A: When the laminated glass is unbroken, the hypotheses of Bernoulli can be adopted
KH
for the glass layers. The modulus of elasticity of the laminate depends on the service
temperature, loading rate, etc., as it was mentioned in Chapter 11. Fracture occurs when the
ultimate strength of the bottom glass layer under highest strain is reached, then the force drops
to the lower level e.g. Fu23 immediately (Fig. B.12.1).
AN
Stage B: The load has to be carried by the non-fractured layers of laminated glass.
Stage C: The interlayer material serves two purposes: (1) the adhesion of the fragments of the
fractured glass layer to the non-fractured glass layer and (2) to transfer forces between the
glass layers. Therefore, the force starts to increase until the ultimate strength of the next layer
K.P
is reached again. When two glass layers are fractured and the interlayer material remains
unbroken, it can help as reinforcement for the last non-fractured glass layer in the case of
laminated glass consisting of three glass layers.
7.50
Fu3 - Fu2
t by
7.00
6.50 No. 3. 20 mm/min
Force during fracture process, F (kN)
Fu2 -Fu1 No. 2.

6.00 No. 1.
5.50 Fu1
5.00
2.73
4.50
1.79
2.49
h
4.00
1.49
7.21
1.16
3.50
yrig
3.00
1.06
2.50
2.38
1.91
2.38
2.15
2.14
2.00
2.14
23 0.71 0.56 0.13

1.50
1.44
1.62
1.18
1.45
2.66
1.50
60 0.80 0.86
23 0.72 0.29
1.00
1.00
-20 0.95
cop
23 0.87
-20 0.83
23 0.85
60 0.81
-20 0.80
23 0.74
23 0.68
60 0.68
0.50
-20 0.61
0.00
23
23
60
12mm F_2_R F_2_EVA F_2_D 19mm F_3_R F_3_EVA F_3_D

Fig. B.12.2 Forces during fracture process of laminated float glasses at temperatures of
-20 °C, +23 °C and +60 °C
Blue colour indicates the ultimate force, Fu1, of the bottom glass layer (No.1) as well as the
ultimate force of single glass layers with thicknesses 12 mm and 19 mm for comparison.
Green colour indicates the difference between ultimate forces, Fu2 and Fu1. Beige colour
indicates the difference between ultimate forces, Fu3 and Fu2. Laminated glasses with non-
bonded layers are also indicated
B 31
20.00
Fu3 - Fu2
dF32
18.00 No. 3.
Force during fracture process, F (kN)

F No. 2. 20 mm/min
u2 -Fu1
dF21
16.00 No. 1.
F1u1
F
9.64
14.00
12.00
7.41
6.48
4.93
5.41
10.00
15.52
5.14
8.00
7.29
6.00
5.55
5.68
5.72
5.47
4.78
5.43
4.76
4.70
23 1.550.82 1.62
4.52
7.94
3.45
3.40
4.00
DT
23 1.58 1.29
2.00
-20 2.46
60 2.40
-20 2.38
-20 2.09
23 2.13
60 1.90
23 1.88
60 1.88
-20 1.92
23 1.93
60 1.86
0.00 23 1.78
23
23
AR
12mm E_2_R E_2_EVA E_2_D 19mm E_3_R E_3_EVA E_3_D
Fig. B.12.3 Forces during fracture process of laminated tempered glasses at temperatures of
-20 °C, +23 °C and +60 °C
KH 300%
AN 228%
277%
250%
186%
230%
K.P 222%
221%
200%
146%
171%
196%
134%
173%
180%
165%
156%
161% 193%
166%
190%
166%
150%
168%
142%
141%
144%
149%
133%
144%
137%
159%
160%
139%
131%
143%
125%
120%
100%
113%
t by
103%
50%
h
-20 0%
F_3_F_C2
23
E_3_F_C2
F_3_R_C2
E_3_R_C2
F_3_F_C1
60
yrig
E_3_F_C1
F_3_R_C1
E_3_R_C1
F_2_F_C1
E_2_F_C1
F_2_R_C1
E_2_R_C1
Temperature,
T, °C
cop
Fig. B.12.4 Increase of force during fracture process after fracture of a glass layer in
laminated glass, at Stages C1 (…_C1) and C2 (…_C2) vs. temperature, in percent (yellow
and orange colours indicate with EVA foil _F_, blue and green colours with resin _R_
laminated glasses)
B 32
20.00
Change of force during fracture process,

5 Fu3-Fu2
18.00
20 mm/min
16.00 4 Fu2-Fu23
14.00 3 Fu23-Fu1 (Fu2-Fu1) 5
12.00 2 Fu1-F u12

5 5
∆F (kN)
5
10.00 5
1 Fu12 (Fu1)
8.00 1 4 5
4
4 4 4
6.00
3 3 4
4.00 1 3 3 3 3 3
3 3 3 3 5
3
3 4
2.00 2 2 2 2 2 2 2 2 2 2 2 3
2 2 2
DT
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
0.00 1
-20°C
23°C
60°C
23°C
-20°C
23°C
60°C
-20°C
23°C
60°C
23°C
23°C
-20°C
23°C
60°C
-20°C
23°C
60°C
23°C
6mm 12mm E_2_R E_2_EVA E_2_D 19mm E_3_R E_3_EVA E_3_D
AR
Fig. B.12.5 Change of force during fracture process of laminated tempered glasses at
KH
AN
K.P
Fig. B.12.6 Schematic force versus deflection diagram of laminated glass in fracture process
b
Fb = Fa (B.12.1)
t by
Table B.12.1 Ratio of residual force, Fb, to maximal force, Fa

Float Tempered
h
No. of
Type Ratio
layers -20 °C +23 °C +60 °C -20 °C +23 °C +60 °C
yrig
2 b2/a2 0.1564 0.2493 0.3012 0.1165 0.1460 0.2206

Resin b3/a3 0.4066 0.4742 0.5549 0.3347 0.3861 0.4460
3
b2/a2 0.1549 0.2167 0.3099 0.1064 0.1576 0.2471
cop
2 b2/a2 0.1870 0.1785 0.2267 0.1201 0.1281 0.1505

EVA b3/a3 0.4180 0.4016 0.4963 0.3347 0.3504 0.4473
3
b2/a2 0.1652 0.1731 0.2749 0.1348 0.1263 0.1770
2 b2/a2 0.6733 0.4774
D b3/a3 0.7429 0.5664
3
b2/a2 0.5118 0.5443
B 33
Float -20 °C Float +23 °C Float +60 °C

Tempered -20 °C Tempered +23 °C Tempered +60 °C
0.80
0.70
0.60
0.50
Ratio (-)
0.40
0.30
0.20
0.10
DT
0.00
b2/a2 b3/a3 b2/a2 b2/a2 b3/a3 b2/a2 b2/a2 b3/a3 b2/a2
2 3 2 3 2 3
Resin EVA D
AR
Fig. B.12.7 Change of ratio of residual force, Fb, to maximal force, Fa, vs. change of
temperature and type of interlayer material or (_D ) without interlayer
KH
AN
K.P
Fig. B.12.8 Calculated areas during the fracture process in force vs. deflection diagram
t by
45 No.3 glass layer

40 No.2 glass layer
No. 3.
12.44
12.22
No.1 glass layer No. 2.

12.99
17.90
h
No. 1.
35
30
yrig
20 mm/min
11.27
13.12
External work, Nm
25
42.79
18.50
18.17
18.61
20
13.95
17.98
6.33 3.18
8.80
23.16
14.50
13.82
15
10.25
8.90
14.14
cop
10.81
10
3.08
11.78
11.42
11.17
10.79
10.26
9.21
8.91
8.23
8.00
5
7.34
7.29
6.92
6.74
5.40
6.16
5.98
4.36
0
-20°C
23°C
60°C
23°C
-20°C
23°C
60°C
-20°C
23°C
60°C
23°C
23°C
-20°C
23°C
60°C
-20°C
23°C
60°C
23°C
F_6mm F_12mm F_2_R F_2_EVA F_2_D 19mm F_3_R F_3_EVA F_ 3_D
Fig. B.12.8 External work during the fracture process of laminated float glasses at
temperatures of -20°C, +23 °C, +60°C, symbols are: F_- float glass; numbers: number of
glass layers, _EVA- EVA foil, _R- resin interlayer material, _D - without interlayer material
as well as single layer glasses with thickness of 6 mm, 12 mm, 19 mm
B 34
B.7 to Chapter 13 Future work
DT
AR
KH
Fig. B.13.1 Evaluation version of “StressOptic” software (© by the Author)
AN
K.P
t by
Fig. B.13.2 Schematic illustration of test arrangement, side and front view
h
yrig
cop
σt=0=72.89 MPa σmax=109.33 MPa

Fig. B.13.3 Captured photos for analysis, changing of fringe pattern by increasing the
loading. Accuracy is 0.263 mm.
B 35
DT
Fig. B.13.4 Four-point bending of single float and tempered laminated glasses with carbon
AR
stripe reinforcement
KH
AN
K.P
Fig. B.13.5 Four-point bending of single float and tempered laminated glasses with
self-manufactured carbon T-profiles
h t by
yrig
cop
B 36
K. Pankhardt Load bearing glasses Appendix C
Properties of interlayer materials
APPENDIX C
Contents
APPENDIX C C1
1. Properties of interlayer materials C1
2. Testing of interlayer materials and informative results C2
2.1 Differential scanning calorimetric analysis (DSC) of cured resin C2
2.2 Informative tensile tests of interlayer materials C3
2.2.1 Conclusions of results of informative tensile tests of EVA interlayer material C5
2.2.2 Conclusions of results of informative tensile tests of resin interlayer material C7
2.3 Informative shear test of resin interlayer material C8
DT
3. Photo documentation C9
3.1 Tensile tests of EVA interlayer C9
3.2 Tensile tests of resin interlayer C11
3.3 Shear tests of resin interlayer C13
AR
4. References C13
KH
1. Properties of interlayer materials
Table C.1.1 Typical properties of cured EVA interlayer (Technical datasheet, Bridgestone)
AN
Properties of EVA Units Type
SG1451U38
Thickness mm 0.4
K.P
Tensile strength MPa 22.1

Ultimate strain % 530
Young’s modulus at 23°C kg/cm2 12
Poisson ratio - 0.32
3
Specific gravity g/cm 0.95
t by
Hardness Shore A 82
Stress transition temperature °C -28
Melting temperature range °C 76-79
Volume resistivity Ωcm 5.4×1015
h
Dielectric constant (1 kHz) - 3.4

yrig
Breakdown voltage kV/mm 19

Optical transmission % 91.7
Haze value - 0.2
Refractive Index - 1.491
cop
UV cutoff wavelength nm 380

Specific heat cal/°C g 0.55
Thermal conductivity kcal/mh°C 0.1
Thermal expansion 1/°C 3.5×10-4
Water absorption (°C×24 hours) % <0.01
Water permeability (°C×90% RH) g/m2×24 hours 64.3
Gel content % 95
C1
DT
Fig. C.1.1 Viscosity of different adhesives EVA type SG50 and SG51 and PVB (Technical
datasheet given by Bridgestone)
2. Testing of interlayer materials and informative results
AR
2.1 Differential scanning calorimetric analysis (DSC) of cured resin
KH
AN
Fig. C.2.1 From left to right: mixed Fig. C.2.2 left: viscosity of liquid resin decreases,
liquid resin, activator, UP resin right: solid phase during curing process
K.P
21
Heat flow (mW) (endo down )
1th heating
20
t by
19
Tg: -38°C Tm: +109 °C
18
h
-50 -35 -20 -5 10 25 40 55 70 85 100 115 130 145 160 175 190 205 220
yrig
Fig. C.2.3 Heat flow vs. temperature in DSC analysis of cured VIAPAL VUP 4808 B/62 resin,
where glass transition temperature, Tg: -38 °C and melting temperature, Tm: + 109 °C is
indicated
cop
23
Heat flow (mW) (endo up)
22 2nd heating
21 Tg: -40 °C
20
19
18
-46 -26 -6 14 34 54 74 94 114 134 154 174 194 214
Fig. C.2.4 Heat flow vs. temperature in DSC analysis of cured VIAPAL VUP 4808 B/62
resin, where glass transition temperature, Tg: -40 °C is indicated in second heating process
C2
21
Heat flow (mW) (endo down)

Cooling
20
Tg: -42 °C
19
18
-60 -45 -30 -15 0 15 30 45 60 75 90 105 120 135 150 165 180 195
DT
Fig. C.2.5 Heat flow vs. temperature in DSC analysis of cured VIAPAL VUP 4808 B/62 resin,
where glass transition temperature, Tg: -42 °C is indicated in cooling process
AR
2.2 Informative tensile tests of interlayer materials
The influence of interlayer on the stiffness and thus on the load bearing capacity of laminated
glass was shown in the previous Chapters of this Thesis (Chapters 9 to 12). Tensile tests were
KH
carried out to study the effect of loading rate and temperature on the stress – strain diagram of
EVA and resin interlayer materials. The properties of interlayer materials are given by the
manufacturer (Chapter 7, Table 7.1) and were supplemented with my own informative results
(Table C.2.1). The test results are only informative, because of the low number of tested
AN
specimens.
Table C.2.1 Physical properties of cured resin and cured EVA foil interlayer (data of
Technical datasheets and supplemented with the results of the Author’s measurements*)
K.P
Resin EVA foil

Properties Unit
(UP) (ethylene vinyl acetate)
Specific gravity g/cm3 1.16 (*1.191) 0.95 (*1.063)
Thickness mm 1 0.4
2 (*0.65 at 14 to 22 (* 19.22 at
N/mm2
t by
Tensile strength at 23±2 °C

20 mm/min ) 20 mm/min)
(~160) > 60 (* 54 at > 465 (*157.1 at
Ultimate strain at 23±2 °C %
20 mm/min) 20 mm/min after 5 years)
~ 12 (*1.18 at > 1.2 (* 547.38 at
Young’s modulus at 23±2 °C N/mm2
h
20 mm/min) 20 mm/min after 5 years)

Poisson ratio at 23±2 °C - ~ 0.40 0.32
yrig
~ -28 (depending on the

Glass transition temperature °C -42 to -38
content of softeners)
Melting temperature range °C 109 (*80 to 140) ~ 76 to 79
Water absorption (°C×24Hours) % - <0.01
cop
Styrene monomer
Other Gel content 95%
content 61.6 m %
Cured EVA interlayer – cured in the laminated glass manufacturing process – was stored at
room temperature for five years. The resin interlayer material was prepared with 1 mm
thickness as shown in Figs. C.3.9 to C.3.12.
The applied testing temperature of the polymer was equal to the temperature of the tested
laminated glasses (-20°C, +23°C, +60°C). The loading rate was 2 mm/min at each
temperature and further loading rates 0.2 mm/min, 20 mm/min and 200 mm/min were applied
at +23 °C. Types of applied tensile testing equipment were: Zwick Z020 and Zwick Z005 for
room temperature tests and Zwick Z050 for -20 °C and +60 °C temperature tests.
C3
The tensile test results are summarised in Table C.2.2 for EVA and Table C.2.3 for resin
materials. The results of tensile tests are only informative. Fig. C.2.6 indicates the typical
stress-strain diagrams of polymers given in MSZ EN ISO 527-1:1999.
DT
AR
KH
a – rigid materials
AN
b, c – ductile materials with yielding (σy)
d – ductile materials without yielding
Fig. C.2.6. Typical stress-strain diagrams (MSZ EN ISO 527-1:1999)
K.P
The yield properties depend on the polymer crystallinity and the polymer morphology. The
yield behaviour also depends on the test conditions used. The yield properties vary with both
the test temperature and the rate of loading.
h t by
yrig
Fig. C.2.7 Effect of temperature on the stress-strain properties of polymers. The material
cop
becomes softer and more rubbery but loses its tensile strength with the increase of
temperature (Scheirs, 2009).
In addition to the length of the polymer chains (i.e. the molecular weight), the mechanical and
physical properties of plastics are also influenced by the bonds within and between chains,
chain branching and the degree of crystallinity (see Chapter 4). The ordered and aligned
portions of the polymer chain form small regions that are called crystallites. The non-ordered
regions are called amorphous. These amorphous regions that are not crystalline and contain
more random orientation of the polymer chains. The proportion of crystalline (ordered and
tightly packed) regions to the amorphous (disordered) regions is expressed as the degree of
crystallinity of the polymer (Scheirs, 2009). The degree of crystallinity has a pronounced
effect on the performance properties of the polymer, especially the mechanical properties and
C4
chemical resistance (Fig. C.2.7). The increase of degree of crystallinity makes polymers
strong, but also lowers the impact resistance.
The semi-crystalline microstructure of polymers imparts excellent chemical resistance and
high strength, however it also makes them susceptible to environmental stress cracking
(ESC). The combined action of stress and a chemical agent is referred to as environmental
stress cracking. Polymers with low crystallinities are more flexible and not susceptible to
environmental stress cracking (see Fig. C.2.8).
2.2.1 Conclusions of results of informative tensile tests of EVA interlayer

material
DT
Fig. C.2.8 indicates the stress – strain curves of tensile test results of EVA specimens.
35
AR
EVA_1_23°C_2mm/min EVA_2_23°C_2mm/min
30 EVA_9_23°C_200mm/min EVA_2_+60°C_2mm/min
EVA_10_23°C_0.2mm/min EVA_1_-20°C_2mm/min
EVA_2_-20°C_2mm/min
KH
25
-20°C; 2 mm/min +23°C; 20 mm/min
Stress, σ, (MPa)
20 +23°C; 200 mm/min

AN
15
+23°C; 0.2 mm/min

10
+23°C; 2 mm/min
K.P
+60°C; 2 mm/min
5
end of testing
0
0 20 40 60 80 100 120 140 160 180
t by
Strain, ε , %
Fig. C.2.8 Stress – strain diagram of tensile test results of EVA
In the case of +23 °C and lower loading rates (0.2; 2; 20 mm/min), a hardening effect can be
observed. The reason is that long chain molecules adapt to their new state while being
h
permanently stretched. In the case of increased loading rate, from 20 to 200 mm/min, the
yrig
chains have less time to adopt to the direction of tension or in the case of -20 °C the motion of
the chains is reduced. The decrease of temperature around the glass transition temperature of
the interlayer material can cause significant stiffening and reduction in elasticity, see Fig.
C.2.8.
cop
The tensile stress – strain curve of EVA (with loading rate of 2 mm/min) is more affected by
the change of the temperature (-20 °C; +23 °C; +60 °C) than by the change of the loading rate
from 0.2 mm/min to 200 mm/min. With the increase of temperature, the time for chain motion
and adaptation to the loading direction is reduced. In the case of EVA material, sudden
stiffening on the stress – strain curves does not take place with the decrease of the
temperature, such as in the case of resin interlayer material (Figs C.2.11 to C.2.12).
Adhesion capacity, elasticity and water-absorbing capacity are influenced by softeners (see
also Chapter 4), therefore, the results are valid for the tested type of interlayer material.
It should be mentioned that the viscoelastic properties of the polymers will change not only in
time (ageing) but the type of loading will also have a great influence on them. In the case of
cyclic loading, the long chains of molecules will start to adapt to their loading direction and
more and more molecules will adapt under cyclic loading. Therefore, some increase in
C5
stiffening can be observed in time, which affects the load bearing capacity and deflection
behaviour of the laminated glass panes. In the case of increase of stiffening, the laminated
glass behaves more rigidly than in the first years, and with the aging of the interlayer material
the ultimate elongation decreases which affects the residual load bearing capacity of the
laminate. In this field further research of laminated glass is needed.
The tensile strength of EVA given by the manufacturer is 14 to 22 N/mm2 at 23 °C and the
ultimate strain should be higher than 450 % (Table C.2.1). Table C.2.2 indicates that the
ultimate tensile stresses remain between the given values at +23 °C, but after 5 years a
significant decrease in the ultimate strain can be observed (see also Chapter 4.2.3.2).
Table C.2.2 indicates also the types of typical stress-strain diagrams of polymers given in the
DT
MSZ EN ISO 527-1:1999 standard.
Table C.2.2 Tensile test results of aged EVA interlayer material (5 years old, stored at room
temperature); * testing limit for large specimen; ** result of small specimen
AR
Tempe- Loading Ultimate Ultimate Yield Yield Young’s Type
rature rate stress, strain, stress, strain, modulus according
T, σb, εb, σy, εy, E, to
KH
2 2 2
°C mm/min N/mm % N/mm % N/mm EN 527-1
-20 2 22.03 2.47 - - 1195.28 a
+23 200 19.14 12.73 21.12 9.87 508.27 a
AN
+23 20 19.22 157.10 17.94 9.47 547.38 b
+23 2 15.00 94.56 15.48 12.18 486.53 b,c
(55.64)
+23 0.2 15.94 159.06 11.84 14.91 304.16 b
K.P
+60 2 6.01 159.55* 3.25 56.54 61.09 d

682.17**
The change of Young’s modulus versus change of temperature and loading rate is plotted in
Figs. C.2.9 and C.2.10. Fig. C.2.9 indicates the decrease of Young’s modulus of EVA
t by
material with increase of temperature from -20 °C to +23 °C and +60 °C. Fig. C.2.10
indicates no significant increase of Young’s modulus of EVA material with increase of
loading rate from 0.2 to 200 mm/min at +23 °C.
10000 1000
h
E modulus_EVA
-20; 1195.28
loading rate: 2 mm/min 2; 486.53 20; 547.38 200; 508.27
1000
E-Modulus (MPa)
E-Modulus (MPa)
yrig
0.2; 304.16
100
23; 486.53 E modulus_EVA
100 60; 61.09
10
10
cop
1 1
-40 -20 0 20 40 60 80 0.1 1 10 100 1000
Temperature (°C) Loading rate (mm/min)
Fig. C.2.9 Temperature and tensile E – Fig. C.2.10 Loading rate and tensile E –
Modulus of EVA interlayer material, semi- Modulus of EVA interlayer material
logarithmic scale at +23 °C, logarithmic scale
The photo documentation of tensile tests of EVA material can be seen in Chapter C.3.1.
C6
2.2.2 Conclusions of results of informative tensile tests of resin
interlayer material
Fig. C.2.11 indicates the stress - strain diagram of tensile test results of resin specimens.
Fig. C.2.12 is enlarged from Fig. C.2.11.
30
Resin_1_2mm/min Resin_2_2mm/min
25 Resin_1_200mm/min Resin_2_200mm/min
Resin_1_0.2mm/min Resin_2_0.2mm/min
Resin_1_-20°C_2mm/min Resin_2_-20°C_2mm/min
20
Stress, σ , MPa
DT
-20°C; 2 mm/min
15
AR
10
+23°C; 20 mm/min
+23°C; 2 mm/min
5
+23°C; 0.2 mm/min +23°C; 200 mm/min
KH
0
0 20 40 60 80 100 120
Strain, ε , %
AN
Fig. C.2.11 Tensile stress – strain curves of resin
1.8
K.P
1.6 Resin_1_20mm/min Resin_2_20mm/min

1.4 Resin_1_0.2mm/min Resin_2_0.2mm/min
Resin_1_-20°C_2mm/min Resin_2_-20°C_2mm/min
1.2 +23°C; 200 mm/min
Stress, σ , MPa
-20°C; 2 mm/min
t by
1.0
+23°C; 20 mm/min
0.8
+23°C; 2 mm/min
0.6
h
0.4
+23°C; 0.2 mm/min
yrig
0.2
0.0
0 20 40 60 80 100
Strain, ε , %
cop
Fig. C.2.12 Tensile stress – strain curves of resin (detail, enlarged from Fig. C.2.11)
Fig. C.2.11 and Fig. C.2.12 indicate that the tensile stress-strain curves of resin do not have
yielding properties and are nearly linear until fracture of the specimen. The molecules of the
tested type of cured resin have little relative mobility. The molecular structure of the tested
UP resin is rather crystalline (App. A, Fig. A.4.13) while the structure of EVA - as a hot-melt
adhesive - can vary from partially crystalline to amorphous and rubber-like, depending on the
composition (VA content, see Fig. A.4.9). In the case of resin, the Young’s modulus is not
significantly affected by the change of loading rate at +23 °C. A significant increase of
Young’s modulus can be observed with the increase of temperature. (In the case of +60 °C,
very small resistance was measured, therefore only the ultimate strain is indicated in Table
C.2.3).
C7
Table C.2.3 Tensile test results of cured resin interlayer material (tested 24 h after curing)
Tempe- Loading Ultimate Ultimate Yield Yield Young’s Type by
rature rate stress, strain, stress, strain, modulus EN 527-1
T, σb, εb, σy, εy, E,
°C mm/min N/mm2 % N/mm2 % N/mm2
-20 2 20.10 105.78 - - 19.00 a
+23 200 1.07 75.89 - - 1.41 a
+23 20 0.65 54.78 - - 1.18 a
+23 2 0.34 31.26 - - 1.09 a
DT
+23 0.2 0.31 29.76 - - 1.04 a
+60 2 < 0.3 31.92 - - <0.94 a
The tensile strength of cured resin given by the manufacturer is about 2 N/mm2 (Table C.2.1)
AR
and the ultimate strain should be higher than 60 % at 23 °C.
100 10
E modulus_Resin
-20.00; 19.00 E modulus_Resin
E-Modulus (MPa)
KH
loading rate: 2 mm/min
E-Modulus (MPa)
10
23.00; 1.09 60.00; 0.94

1
AN 200; 1.41
-40 -20 0 20 40 60 80 2; 1.09 20; 1.18
0.2; 1.04
1
0.1 0.1 1 10 100 1000
Temperature (°C) Loading rate (mm/min)

K.P
Fig. C.2.13 Temperature and tensile Fig. C.2.14 Loading rate and tensile
E-modulus of resin interlayer material, semi- E-modulus of resin interlayer material
logarithmic scale at +23 °C, logarithmic scale
Fig. C.2.13 indicates the decrease of Young’s modulus of resin with increase of temperature
t by
from -20 °C to +23 °C and +60 °C. Fig. C.2.14 indicates the increase of Young’s modulus of
resin with increase of loading rate from 0.2 to 200 mm/min at +23 °C.
A significant decrease of Young’s modulus can be observed with the increase of temperature
from -20 °C to + 23 °C.
h
Young’s modulus is more affected by the change of temperature in the case of resin than in
yrig
the case of EVA material and that will affect the behaviour of the laminate. Young’s modulus
of EVA is 2.45 times higher at -20 °C than at +23 °C and is 17.43 times higher in the case of
resin material.
The photo documentation of tensile tests of cured resin material can be seen in Chapter C.3.2.
cop
2.3 Informative shear test of resin interlayer material

Shear tests were also carried out with resin laminated glass specimens. Glass specimens were
prepared with an overlapping length of 50 mm and width of 25 mm and a thickness of 1 mm
of cured resin (Fig. C.3.20). Two different types of testing arrangements were applied. One
type of testing arrangement was axial shear-tensile and the other was a pure shear test. The
test facilities are presented in Fig. C.3.21 and Fig. C.3.22. The specimens were moment-free
loaded with a loading rate of 2 mm/min. The force and displacements were continuously
recorded. The results of shear tests are only informative, because of the low number of tested
specimens.
C8
0.60 0.60
shear_tension_resin + 23 °C 2 mm/min shear_tension_resin + 23 °C 2 mm/min
shear_resin_1 shear_resin_1
0.50 shear_resin_2 0.50 shear_resin_2
Shear stress, τ (MPa)
Shear stress, τ (MPa)

0.40 0.40
0.30 0.30
0.20 0.20
0.10 0.10
0.00 0.00
0.00 2.00 4.00 6.00 8.00 10.00 0.00 0.50 1.00 1.50 2.00 2.50
Strain, ε (%) Shear angle, γ, ∆ u/h int (-)
DT
Fig. C.2.15 Shear stress - strain curves of Fig. C.2.16 Shear stress – shear angle of
resin interlayer material resin interlayer material
AR
Table C.2.4 Shear test results of cured resin interlayer material (tested 24 h after curing)
Temperature Loading Shear Ultimate shear angle, Type of testing
rate strength,
T, τb,
KH γb,
AN
°C mm/min 2 - (°)
N/mm
+23 2 0.41 1.34 (53.37°) tension-shear
+23 2 0.40 1.74 (60.11°) pure shear
K.P
The photo documentation of shear tests of cured resin material can be seen in Chapter C.3.3.
t by
3. Photo documentation
3.1 Tensile tests of EVA interlayer
h
yrig
cop
Fig. C.3.1 EVA specimens for tensile tests Fig. C.3.2 Tested EVA specimens
C9
DT
Fig. C.3.3 Tensile testing of EVA at temperature of +23°C, 2 mm/min
AR
KH
AN
K.P
Fig. C.3.4 Tensile testing of EVA at Fig. C.3.5 Tensile testing of EVA at
temperature of +23°C, 2 mm/min temperature of +23°C, 20 mm/min
h t by
yrig
cop
Fig. C.3.6 Tensile test of EVA at +60°C Fig. C.3.7 Tensile test of EVA at -20°C
C 10
DT
AR
Fig. C.3.8 Tensile test of EVA at -20°C, cracks in fractured specimen
3.2 Tensile tests of resin interlayer
KH
AN
K.P
Fig. C.3.9 Preparation of 1 mm thick resin Fig. C.3.10 Cured 1 mm thick resin foil for
t by
foil for tensile tests tensile tests

h
yrig
cop
Fig. C.3.11 Cured transparent 1 mm thick Fig. C.3.12 Specimens for tensile tests of
resin foil for tensile tests cured resin
C 11
DT
Fig. C.3.13 Tensile testing of cured resin at temperature of +23°C, 2 mm/min
AR
KH
AN
K.P
Fig. C.3.14 Tensile testing of resin at Fig. C.3.15 Tensile testing of resin at
t by
temperature of +23°C, loading rate of temperature of +23°C, loading rate of

20 mm/min 20 mm/min
h
yrig
cop
Fig. C.3.16 Tensile test of resin at +60 °C Fig. C.3.17 Tensile test of resin at -20 °C
C 12
DT
Fig. C.2.18 Tensile tested resin specimens Fig. C.2.19 Tensile tested resin specimens
3.3 Shear tests of resin interlayer
AR
KH
AN
K.P
Fig. C.3.20 Prepared specimens for shear tests Fig. C.3.21 Tensile-shear test
(overlapping 50 mm, width: 25 mm)
h t by
yrig
cop
Fig. C.3.22 Shear test Fig. C.3.23 Shear tested specimen
4. References
MSZ EN ISO 527-1:1999. Plastics. Determination of tensile properties. Part 1: General principles (ISO
527-1:1993 including Corr 1:1994), CEN, Brussels
Scheirs, J. (2009), A Guide to Polymeric Geomembranes: A Practical Approach, Ch 1, John Wiley &
Sons, Ltd., Chichester, UK.
Scheirs, J. (2000), Compositional and Failure Analysis of Polymers: A Practical Approach, John Wiley
& Sons Ltd., Chichester, UK.
C 13
K. Pankhardt Load bearing glasses Appendix D
Definitions
APPENDIX D
Definitions
acid etching: This process for the decoration of glass involves the application of hydrofluoric
acid to the glass surface. Hydrofluoric acid vapours or baths of hydrofluoric acid salts may be
used to give glass a matt, frosted appearance (similar to that obtained by surface
sandblasting), as found in lighting glass. Glass designs can be produced by coating the glass
with wax and then inscribing the desired pattern through the wax layer. When applied, the
acid will corrode the glass but not attack the wax-covered areas, (Banks, 2008).
DT
allowable stress (EN 13474-1): the effective stress which should not be exceeded by the
stress calculated from actions on the glass pane. Note: The allowable stress sometimes called
the design stress.
AR
annealing: is the process of removing all the residual strains from glass by reheating the
finished piece to a precise temperature range (annealing point) depending on the composition
of the glass. While annealing removes the residual stresses induced during the production
KH
process, tempering actually induces a pre determined amount of stress into the glass
batch: A term used to refer to the raw materials required to produce the desired type of glass
once they have been weighed and mixed, and are ready for melting.
AN
cast-in-place lamination process (EN ISO 12543-1, Ch. 3.16): Lamination process where
the interlayer is obtained by pouring a liquid between the plies of glass or plastics glazing
sheet material and is then chemically cured to produce the final product.
K.P
coloured glass: Most glass is coloured by the addition of specific minerals, typically metal
oxides, during founding. Glass of all types can be coloured by the addition of metals, metal
oxides or other compounds to the melt. The colouring agent will either be suspended or
dissolved in the glass.
t by
Table D.1 Metal oxides and colour of glass (Banks, 2008)

Compound added to the melt Colour of the glass
cobalt oxide blue
magnesium oxide violet
h
gold or selenium red

yrig
uranium, iron, or silver oxides yellow

ferric oxide brown
iridium oxide black
copper or chromium oxides green
cop
calcium fluoride or stannic oxide white or opal
cross-linked polymers: Chains are connected by covalent bonds. Often achieved by adding
atoms or molecules that form covalent links between chains. Many rubbers have this structure
(Anderson, 2001).
effective stress (EN 13474-1): A weighted average of the tensile bending stresses, calculated
by applying a factor to take into account non-uniformity of the stress field.
effective thickness (EN 13474-1): The thickness which, when used in stress or deflection
formulae appropriate monolithic (i.e. not laminated) glass, results in a sufficient accurate
value to describe the deflection of laminated glass or the stress in the individual plies of glass
in the laminated glass.
D1
Definitions
flat (float) glass: „Flat glass is glass manufactured in flat sheets (float, sheet and rolled),
which may be further processed (toughened, laminated, coated, etc.). Excludes bottles,
containers, fibreglass, rods, and tubes. Today, almost all flat glass is made using the "float"
process. In this process sand, limestone, soda ash, dolomite, iron oxide and salt cake are
mixed together and fed into a large furnace, in which they combine, in a process of chemical
and physical transformation, to form molten glass. A continuous ribbon of molten glass is fed
out of the melting furnace through a "delivery canal" and onto the surface of an enclosed bath
of molten tin. The molten glass literally floats on top of the tin, and as it flows along the
surface of the tin bath away from the delivery canal, its temperature decreases until it
becomes a solid sheet that can be lifted from the tin onto rollers for further cooling, and
DT
cutting. This process produces glass sheets with a uniform thickness and perfectly smooth
surfaces that need no further polishing. The resulting glass will then be further treated in
various ways to incorporate one or several of the advanced technologies applied to flat glass
AR
today, depending on the end-product and application for which it is destined.”, (Glass for
Europe, Retrieved 2008).
folio lamination process (EN ISO 12543-1, Ch. 3.15): Lamination process where the
KH
interlayer is a solid film which is placed between the plies of glass or plastics glazing sheet
material and is then subjected to heat and pressure to produce the final product.
glass: Glass is a homogeneous material with a random, liquidlike (non-crystalline) molecular
structure. The manufacturing process requires that the raw materials be heated to a
AN
temperature sufficient to produce a completely used melt, which, when cooled rapidly,
becomes rigid without crystallizing, (Schittich, 1999).
glass type factor (GTF): This is defined as a multiplying factor for adjusting the load
K.P
resistance of different glass types. The key term is used in ASTM E 1300-04 (Block, 2002).
heat distortion temperature (HDT) of interlayer materials: Heat distortion temperature
(HDT) and the glass transition temperature (Tg) is used to measure the softening of a cured
polymer e.g. resin. The heat deflection temperature (HDT) test applies a load to force a
t by
deflection in the specimen, while the Tg test only applies heat to the specimen (Fig. D.1).
h
yrig
Fig. D.1 The deflection temperature is the temperature at which a test bar, loaded to the
cop
specified bending stress, deflects by 0.25 mm, (Mathweb, 2009).

The heat deflection temperature or heat distortion temperature (HDT, HDTUL, or DTUL) is
the temperature at which a polymer specimen deflects under a specified load, defined in ISO
75-1:2004. This property is applied in many aspects of product design, engineering, and
manufacture of products using thermoplastic components. Heat deflection temperatures are
usually lower (can be about 20 °C lower) than the glass transition temperatures (Tg) in the
case of resins. The HDT test results are useful to determine service temperature for a polymer
when used in load-bearing elements. However, the deflection temperature test is a short-term
test and should not be used alone for product design. Other factors such as the time of
exposure to temperature, the rate of temperature increase, and the part geometry also affect
the HDT.
D2
Definitions
heat soaked thermally toughened alkaline earth silicate safety glass (EN 15682-1,
Ch. 3.1): Glass within which a permanent surface compressive stress has been induced in
order to give it greatly increased resistance to mechanical and thermal stress and prescribed
fragmentation characteristics and which has a known level of residual risk of spontaneous
breakage due to the presence of critical nickel sulphide (NiS) inclusions.
insulating glass (IG): Glazing units made up of two or three sheets of glass (known as
"double glazing" or "triple glazing" respectively) held in a metal frame and sealed (normally
with butyl) to create an airtight space of 9-16 mm between them. A desiccant substance in the
airtight cavity prevents the formation of condensation.
DT
AR
KH
Fig. D.2 Edge seal for IG, (Schittich, 1999)
AN
interlayer (EN ISO 12543-1, Ch. 3.10): One or more layers or materials acting as an adhesive
and separator between plies of glass and/or plastics glazing sheet material. It can also give
additional performance to the finished product e.g. impact resistance, resistance to fire, solar
K.P
control, acoustic insulation. The interlayer itself may also encapsulate non adhesive films and
plates, wires, grids etc.
laminated glass (EN ISO 12543-1, Ch. 3.1): Assembly consisting of one sheet of glass with
one or more sheets of glass and/or plastics glazing sheet material joined together with one or
more interlayer.
t by
Table D.2 Glass types used as components for the production of laminated glass and
laminated safety glass (EN 14449:2005(E) - Table 1)
Basic glass products Reference
h
Special basic glass products: EN 572-1

Borosilicate glasses EN 1748-1-1
yrig
Glass ceramics EN 1748-2-1

Heat strengthened soda lime silicate glass EN 1863-1
Thermally toughened soda lime silicate safety glass EN 12150-1
Chemically strengthened soda lime silicate glass EN 12337-1
cop
Thermally toughened borosilicate safety glass EN 13024-1

Alkaline earth silicate glass products EN 14178-1
Heat soaked thermally toughened soda lime silicate safety glass EN 14179-1
Thermally toughened alkaline earth silicate safety glass EN 14321-1
laminated safety glass (EN ISO 12543-1, Ch. 3.2): Laminated glass where, in the case of
breakage, the interlayer serves to retain the glass fragments, limits the size of opening, offers
residual resistance and reduces the risk of cutting or piercing injuries.
load resistance (LR): This is the uniform lateral load that a glass construction can sustain
based on a given probability of breakage and load duration. Load resistance is associated with
failure probability less than or equal to 8 specimens per 1000 (from tables). The key term is
used in ASTM E 1300-04 (Block, 2002).
D3
Definitions
load share factor (LSF): This is a multiplying factor derived from the load sharing between
the two glass layers, of equal or different thicknesses and types, also in a sealed IG (insulating
glass) unit. The key term is used in ASTM E 1300-04 (Block, 2002).
musical glasses: Experiments with glasses have produced sound either by rubbing or striking.
glass harmonica: An instrument invented in the 1700s made of various sizes of glass
bowls played by rubbing around the rim with a wet finger (Fig. D.3). It was invented by
Benjamin Franklin (1761) and was derived from the vérillon (musical glasses), a set of
glasses, holding different amounts of water and thus yielding different notes, placed on a
soundboard and rubbed by moistened fingers or, rarely, struck with rods. In 1791 the
DT
Quintet for Glass Harmonica was composed by Mozart for the virtuoso Marianne
Kirchgässner. (Encyclopaedia Britannica Online, 2009).
glass harp, verrophone (French: verre = glass), consisting of vertically arranged glass
tubes, played with moist fingers on the upper edge as with the "musical glasses" (Hilmer,
AR
2009). From 1929, Bruno Hoffmann of Stuttgart started arranging musical glasses which
he referred to as "glass harp". In 1983 S. Reckert invented the Tube-Verrophone (made of
glass tubes) (Fig. D.4).
KH
AN
K.P
Fig. D.3 Glass armonica, glass harmonica, Fig. D.4 Glass harp, verrophone
armonica, harmonica, Glasharmonika, harmonika, (Hilmer, 2009).
t by
harmonica de verre, glassychord (The Bakken

Library and Museum, 2007)
Today, composers are re-discovering “musical glasses” and use them in various styles: ballet
music, songs, movie music, theater-music, contemporary music, open air shows, etc. The
h
master glassblower Gerhard Finkenbeiner rediscovered them in the 1960's. Thomas Bloch is
one of the very few professional glassharmonicist, in the world presently (Bloch, 2009).
yrig
non-factored load (NFL): This is the three-second duration uniform load associated with a
probability of breakage of less than or equal to eight specimens per 1.000 for monolithic
glass. The key term is used in ASTM E 1300-04 (Block, 2002).
cop
sand blasting: A method for giving glass surfaces a matt finish either for decoration or to
reduce transparency. Compressed air forces the abrasive material through the nozzle of a
sandblasting gun and onto the glass surface. Although sand can be used, more effective
abrasives with less toxic effects are now available. Silicon carbide is commonly used, as is
electro-corundum (aluminium oxide). The glass is normally placed inside a special cabinet
with arm holes, a viewing window and dust extraction facilities, (Banks, 2008).
silica: Silicium-dioxide is the main ingredient of glass. The most common form of silica used
in glassmaking has always been sand.
silicones: Polymeric compounds which contain chains of silicon atoms alternating with
oxygen atoms and linked to organic groups. Originally used for container glass coatings
because of their good workability and scratch resistance. The main drawback was that silicone
D4
Definitions
coatings are water-repellent, making labelling difficult. Silicones are now used principally as
adhesives and setting materials, particularly where plasticity or water-repellent characteristics
are required.
soda lime silicate glass: Soda lime silicate glass is low in cost and can be easily worked at
reasonable temperatures. Soda lime silicate glass is used for light bulbs, bottles, fibreglass,
building blocks, windowpanes and other applications where cost is a factor (Banks, 2008).
specified design load: This is the magnitude, type of load (wind or snow), and duration of the
load given by the specifying authority. The key term is used in ASTM E 1300-04 (Block,
2002).
DT
surface treated glass: Glass can be coloured without the addition of specific minerals, a thin
layer can also cover its surface. Coatings may be applied to glass for appearance or
performance of the product in question e.g. anti-reflective coatings applied to auto mirrors to
AR
aid vision, coatings with photocatalytic and hydrophilic properties to make self-cleaning
windows or colouring with foil (Banks, 2008).
tempered (thermally toughened) soda lime silicate safety glass (EN 12150-1:2000, p. 5.):
KH
Glass within which a permanent surface compressive stress has been induced by a controlled
heating and cooling process in order to give it greatly increased resistance to mechanical and
thermal stress and prescribed fragmentation characteristics. Thermally toughened soda lime
silicate safety glass is made from a monolithic glass generally corresponding to standards
AN
EN 572-1,2-5. It fractures into numerous small pieces, the edges are generally blunt.
tempered (FT fully tempered) glass (ASTM E 1300-04, p. 2.): Flat, monolithic, glass layer
of uniform thickness that has been subjected to a special heat treatment process where the
K.P
residual surface compression is not less than 69 MPa or the edge compression is not less than
67 MPa.
Vickers indentation: The procedure is based on the principle of indentation fracture, in
which the scale of microcracking around hardness impressions provides a measure of
t by
resistance to cracks extension. The test surface is loaded with a standard diamond pyramid
indenter (Marshall and Lawn, 1977).
viscosity points, glass forming temperatures:
melting point: above this temperature glass is liquid;
h
working point: glass is easily deformed;

yrig
softening point: maximum temperature at which a glass piece maintains shape for a long
time;
annealing point: internal stresses relax (diffuse);
strain point: above this, viscosity fracture occurs before plastic deformation
(Anderson, 2001).
cop
D5
K. Pankhardt CV english
PROFESSIONAL CURRICULUM VITAE OF

Kinga PANKHARDT
Address: Kinga PANKHARDT

associate prof.
University of Debrecen
Department of Civil Engineering
H - 4028 Debrecen, Ótemető 2-4., Hungary
PhD research at
Budapest University of Technology and Economics,
DT
H - 1111 Budapest, Műegyetem rkp. 3., Hungary
Tel/Fax: (36-1) 463-3451
AR
Home address: H - 1225 Budapest
Nagytétényi út 251/a., Hungary
Tel/Fax: (36-1) 362-0035
KH
Born: Hungary, Budapest, January 15, 1975.
AN
Diplomas, degrees:
MSc Civil Engineer, Technical University of Budapest and
University of Karlsruhe, Germany 1998
K.P
Languages:
English, state examination, intermediate level, 2008
German, state examination, intermediate level, 1991
Employment:
t by
UNIVERSITY OF DEBRECEN
Associate Professor (2007- )
h
SK Engineering Office (2003-2006)

yrig
Civil engineer, lecturer at University of Debrecen (2006)
GLASMETAL Ltd. (2001-2003)

Civil engineer for special glass structures
cop
GKSS RESEARCH CENTRE, Geesthacht (Hamburg), Germany

Department of Material Modelling (WMS)
Research fellow (2000)
TECHNICAL UNIVERSITY OF BUDAPEST

Department of Building Materials
PhD Student (1998-2001)
CV-ENG-1/3
University teaching experience:

University Debrecen (2006- )
- Building Materials 1. and Building Materials 2. (lecture and practicum) in
Hungarian
- Diagnostics of buildings
- Concrete Technology (lecture and practicum) in English
Budapest University of Technology and Economics,

DT
for undergraduate and graduate courses invited to different lectures for civil
engineer and architect students (2002- )
- Topics: glasses, recycling of building materials in Hungarian and English
AR
Technical University of Budapest (1998-2001)
Department of Building Materials
- Building Materials 1. and Building Materials 2. (practicum) in Hungarian and
KH
German
Scholarships:
AN
University of Karlsruhe (Diploma) (Baden-Württemberg), three months, 1998
University of Karlsruhe (DIA) (DAAD), five months, 1995
General areas of research:

K.P
Load bearing glass structures, recycling of building materials
Hungarian projects:
BV-MI 01:2005 (H) “Technical guidelines for Concrete and Reinforced concrete;
Production of concrete with recycled demolition, construction and construction material
t by
manufacturing waste” in Hungarian: ”Beton- és Vasbetonépítési Műszaki Irányelv,

Betonkészítés bontási, építési és építőanyag-gyártási hulladék újrahasznosításával”
“Preparation of decree on the detailed regulation of construction and demolition waste
management” in Hungarian, Mandator: Ministry of Environment, and Ministry of
h
Economy, Mandate: Non-profit Company for Quality Control and Innovation in

yrig
Building and Mahill Ltd. in Hungarian (“Építési és bontási hulladékok kezelésének

részletes szabályairól szóló rendelet előkészítése”, 2001. nov. társszerzőkkel együtt,
Megbízó: Környezetvédelmi Minisztérium és Gazdasági Minisztérium, Vállalkozó:
ÉMI Kht. és Mahill Mérnök Iroda Kft.)
cop
International projects:
Gkss Research Centre, Geesthacht (Hamburg), Germany, Deptartment of
Material Modelling (project: offshore)
Publications: 30 publications (11 in English): 17 papers, 13 lecture.
CV-ENG-2/3
Selected publications
Pankhardt, K., (2004), „Load-Bearing glass structures”, Periodica Polytechnica - Civil

Engineering, Vol.48, No.1-2., pp. 157-172.
Pankhardt, K., Balázs, L.Gy. (2006), „New opportunities of structural glazing, loadbearing
glass structures.” Proceedings (Report) "Responding to tomorrow’s
challenges in structural engineering” Report, IABSE Symposium
Budapest 2006; Vol.92, No.34-35, the whole paper on CD with 11
pages.
Pankhardt, K., Nehme, S.G. (2007), „Experimental studies on carbon fibre reinforced white
DT
cement lightweight mortar and concrete elements” Proceedings of
the 3rd Central European congress on Concrete Engineering,
Visegrád, Hungary, pp. 269-274.
AR
Pankhardt, K. (2008), „Investigation on load bearing capacity of glass panes”, Periodica
Polytechnica - Civil Engineering, Vol.52, No.2, pp. 73-82.,
doi:10.3311/pp.ci.2008-2.03
Pankhardt, K., Balázs, L. Gy. (2010), „Study of edge strength of load bearing glasses”,
KH
Építőanyag, Vol. 62, No.1, pp. 15-22.
AN
K.P
h t by
yrig
cop
CV-ENG-3/3
cop
yrig
h t by
K.P
AN
KH
AR
DT

Load Bearing Glass

Uploaded by

Copyright:

Available Formats

You might also like

Load Bearing Glass

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Load Bearing Glass

Uploaded by

Copyright:

Available Formats

Budapest University of Technology and Economics

Faculty of Civil Engineering

LOAD BEARING GLASSES

Doctoral (PhD) Thesis

Kinga PANKHARDT, MSc (CE)

LOAD BEARING GLASSES

Doctoral (PhD) Thesis

Kinga PANKHARDT, MSc (CE)

writes Kengo Kuma (1954) architect in Tokyo, Japan

Erasing architecture means making it transparent itself as well as

This dissertation is dedicated to Dr. Salem Georges Nehme.

© K. Pankhardt, 2010 for this Doctoral Thesis “Load bearing glasses”

This work is connected to the scientific program of the “Development of quality-oriented

Notations and abbreviations ix

Summary of Scientific New results S1–S3

Appendix A – Literature review A 1 – A 26

Appendix B – Test parameters and test results B 1 – B 36

Appendix C – Properties of interlayer materials C 1 – C 13

Appendix D – Definitions D1–D5

Notations and abbreviations

nR Mean refractive index -

§ temperature dependent residual load bearing capacity (Chapter 12).

The following main influencing factors on bending characteristics were studied:

§ effects of interlayer materials (Chapter 9),

2.1.1 Brief history of glass

Europe after the turn of the century.

of glass panes (see, Chapter 3).

Properties Symbol Soda lime silicate glass Unit

Bending strength (characteristic) ft,fl,k

determined amount of stress into the glass.

3.3.3 Chemically strengthened glass

3.4 Recycled glass

in the so-called deacetylation leaving an unsaturated polymer backbone. Polyene is formed

crystalline materials have their glass-transition temperature above or below room

4.3.2 Thermo-mechanical properties of polymers

behaviour) has a greater value than G’ (solid-like behaviour). At extremely low

1+E-14 1+E+14 t, (s)

rheological simple materials), where a change of temperature is equivalent to a shift on the

5.3.3 Strength recovery by aging

5.4 Crack development during fracture of glass

5.4.2 Crack growth in glass

in terms of statistical distribution functions. Appropriate design of glass elements (supports

of heat tempered or heat-strengthened glasses are easier.

6.2.2 Laminated float glass pane

performance of LG in bending. Researchers (Behr et al., 1984; Vallabhan et al., 1987;

elastically until the fracture of the beam (App. A, Fig. A.6.8).

shown in Fig. 6.4.

parameter κ should be between 0 ≤ κ ≤ 1 (App. A, Fig. A.6.9). Parameters κ and Ec were

Table A.6.1 for different types of structural glasses.

EN 1288-3:2000, 70 × 103 N/mm2 is given.

It can be helpful in preliminary architectural design or to make pre-calculations on costs.

specimens in bending tests with different kinds of influencing factors.

7.2.2 Test programme

with two glass layers with three glass layers

-20 C° +23 C° +60 C°

Non heat- Tempered Non heat- Tempered Non heat- Tempered

Fig. 7.2 Schematic diagram of test programme for laminated specimens

7.3.2.1 Force measurement