20230512PARTIAL Fullarticle

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 23

The current issue and full text archive of this journal is available on Emerald Insight at:

https://www.emerald.com/insight/2040-2317.htm

The partial fire protection of steel The partial fire


protection
members: a comparative study

Jakub Sejna, y and Frantisek Wald
Patrik Dobrovoln
Department of Steel and Timber Structures, Czech Technical University in Prague,
Prague, Czech Republic
Received 2 January 2023
Abstract Revised 14 March 2023
Purpose – This paper provides a summary of the issues in the passive fire protection of steel structures. Types 27 March 2023
of passive fire protection and the material properties of protection members and steel members are described. Accepted 17 April 2023
The paper deals with the possibility of partial fire protection for secondary steel beams, in cases where, due to
possible membrane action, it is not necessary to apply passive protection to the entire beams.
Design/methodology/approach – Studies of partially fire-protected steel structures are compared, and
results from studies with different input data are summarized. A fire experiment was conducted to investigate
the effect of partial passive protection in a small-scale furnace. Based on the findings of the experiment,
numerical models were prepared using Ansys Mechanical.
Findings – The results are summarized, and a partial fire protection length of 500 mm is recommended.
Various partial fire protection lengths were compared, and the temperature development of the steel contactors
was compared using a protection length of 500 mm. At the end of the paper, options for partial passive
protection of steel beams are presented.
Originality/value – Extended paper from ASFE2021 based on selection.
Keywords Fire, Steel joints, Passive fire protection, Finite element method, Partial fire protection, Coatback
Paper type Research paper

1. Introduction
1.1 Fire protection
Many methods and various types of protection are used in the passive protection of steel
structures; see Figure 1. The most common methods involve protecting each element of the
structure with protective coatings, spraying or claddings. However, it is often sufficient to
protect only the main elements of the structure. These are mainly columns and primary
beams. Several studies on partial fire protection of steel structures are available (Ferreira
et al., 2007; Kordosky et al., 2020), but more documentation is needed (Zaharia et al., 2013; Dai
et al., 2010; Wang et al., 2016). Most studies deal mainly with intumescent coatings and their
effect on heat transfer from an unprotected structure to a protected structure. Heat transfer
moves from a warmer environment to a colder environment (Munoz-Garcia, 2016; Imran et al.,
2018; Hautala et al., 2021). It is a well-known phenomenon, and mathematical or analytical
methods for determining it have already been fully developed. However, the chosen input
data have a great influence on the results, the boundary conditions and the thermal properties
of the material. For materials that decompose, that contain water or both at the same time, it is
necessary to include the change in mass during heat propagation. For concrete spalling,
burning of wooden structures, decomposition of materials, or foaming of intumescent
coatings, it is also necessary to include the mass change in the thermal analysis. For most
materials exposed to fire, thermal analysis is sufficient. During the analysis, some of the
thermal material properties is modified to include the effect of the mass change during the
analysis. The three basic mechanisms of heat transfer are conduction, convection and
radiation. Heat transfer is addressed by conduction in a solid material (Wang et al., 2012).
Journal of Structural Fire
Engineering
The support of CTU grant SGS22/141/OHK1/3T/11 Safety and sustainability of timber and steel © Emerald Publishing Limited
2040-2317
structures exposed to fire is gratefully acknowledged. DOI 10.1108/JSFE-01-2023-0001
JSFE

Figure 1.
Options of passive fire
protection systems

The duration of the fire resistance depends on the ratio between the exposed area of the
element that is affected by the temperature during a fire and by the cross-sectional area. This
ratio is given in the dimensioning tables to determine the applicability of protective materials
to a particular element (Xiong et al., 2017; Tunca et al., 2018; Singh and Samanta, 2022).
Depending on the material and the method of application, fire protection cladding can be
divided into board cladding and glued cladding. Board cladding is placed directly on the
protected member or on an auxiliary structure such as a steel grid. The function of the
cladding is limited by the smallest details, for example, due to poor design and joints between
boards. For these reasons, it is essential when applying fire protection to follow the system
manufacturer’s instructions exactly. A basic parameter in the selection of the board material
for fire protection cladding is the relative humidity in the fire compartment where the
cladding is to be applied. High humidity values may cause the board to crack during a fire.
Claddings that are susceptible to breakage when exposed to high humidity levels for long
periods include vermiculite-based boards, which become degraded and lose strength in such
an environment. During the application of passive fire protection, it is advantageous to
distribute the required thickness of the cladding among multiple layers of boards on top of
each other, rather than applying only a single board of a given thickness. An aggressive
environment is created in the area under the cladding, which can lead to corrosion of the
protected elements and can thus reduce their service life. It is therefore important to
thoroughly treat the protected element with an anti-corrosion coating before installing the
protective panels. The same problem can be encountered in corrosive environments and on
industrial construction sites, where widely used wire fasteners can corrode. As a result, the
joints between the plates may open up or the entire system may break down after only a few
years in such an environment. It is therefore necessary to select the appropriate material not
only for the plate but also for the fastener, depending on the type of environment. In the case
of bonded cladding, a basalt felt board with a reduced adhesive content is bonded to the
protected element (Tamas-Benyei and Santha, 2023; Jong et al., 2021; Yang et al., 2022).
Fire coatings are the oldest type of fire protection, as they were already known in some
form in antiquity. In ancient Greece, lime water solutions, aqueous salt solutions and water
glass were used to protect buildings. Barrier coatings are named after their main function in
preventing flames from reaching the surface of the protected element and preventing it from
igniting. However, even if ignition does occur after a period of time, barrier coatings still
prevent oxygen access and limit the spread of flames across the surface. These coatings are
mainly used to protect combustible materials, such as timber structures or cable canals. Most
often, these coatings are based on inorganic systems, such as easy-to-melt enamels with
additives. During a fire, the polymeric binder of the coating is released, and the inorganic
residue is converted into a solid crust that adheres to the substrate and prevents the spread of The partial fire
flames. These coatings are generally inexpensive, and their function is mainly based on protection
preventing the entry of flames and subsequently preventing the entry of oxygen. The most
widely used coatings are intumescent or foaming coatings. They are usually organically
based. At elevated temperatures, a chemical reaction takes place, which produces a
carbonaceous residue on the surface of the protected steel element. As a result of the super
coolant that is present, this residue increases in volume and forms an insulating layer of
nonflammable foam. The amount of foaming, the size of the individual cells and the structure

of the foam are important for the function of the coating (Celko et al., 2017; Jo et al., 2022; Li
et al., 2022; Suzuki et al., 2022; Vagge and Ghogare, 2022).
Intumescent coatings can be considered as a complete system consisting of a primer, a
functional layer and a topcoat. Because intumescent coatings are a group of substances that
must interact with each other under precise conditions, the individual layers is applied in the
required quality and with the required thickness, see Figure 2. The primer must therefore
show sufficient adhesion to the surface that is to be protected and must demonstrate stability
up to the point of foaming of the functional layer. At the same time, the primer must show
sufficient corrosion protection to protect the steel element. The topcoat does not have to
perform only an aesthetic function. Its other functions can be, for example, protection against
surrounding moisture, UV radiation and possible minor mechanical damage (Fan et al., 2022;
Lucherini et al., 2023). In addition, it must burn off before it is foamed (Azmi et al., 2022;
Piperopoulos et al., 2022; H€aßler and Hothan, 2022).
The principle of sublimating coatings is that the additives in the polymer epoxy binder
evaporate at high temperatures. This phenomenon causes the flame to be drawn away from
the protected structure, thereby cooling it. A significant advantage of these coatings is their
durability and their resistance to external influences, including mechanical stresses.
However, the comparatively high cost of sublimating coatings is problematic (Milionis
et al., 2017).
Fire protection coatings have a long history. The first modern spraying materials
appeared at the end of the nineteenth century. In terms of composition, spraying materials are
divided into lime cement, gypsum with expanded perlite and vermiculite. In the past, spray
coatings containing large amounts of asbestos, often up to 75%, were used. Coatings offer
many advantages over paints, but there are also disadvantages. a significant disadvantage of
coatings is that they have an uneven surface, and paints have a better appearance.

Figure 2.
3D scheme of the
intumescent coating
JSFE The surface of spray-on coatings can be smoothed, but it is a very expensive process. Another
significant problem is inconsistency on the substrate, which may not become apparent until
several years after application. Due to the need to apply a relatively thick layer of spray, of the
order of 20–40 mm, and due to the high bulk weight of the spray, the weight load on the
structure of the fire protection is not negligible. However, a significant advantage of sprayed
coatings is the possibility to achieve up to 240 min fire resistance of the structure. At the same
time, if the procedure for applying the coating is correctly followed and a suitable
environment is maintained, the coatings age very slowly and their properties hardly change.
The correct functionality of the coatings is influenced by many factors, both in the
preparation of the material itself and in its subsequent application to the protected element.
It is therefore important, both during production in the manufacturing company and during
mixing on site, to keep to the precise proportions of the raw materials, the precise method for
adding them and the correct mixing of the mixture. A poorly mixed mixture can lead to
various unacceptable situations: wrong adhesion to the substrate, inadequate overall
strength, inadequate spray consistency and inadequate surface hardness. Another problem
that can occur is an excess of certain admixtures, such as cement. When excessively large
amounts are used, the insulating properties of the spray can deteriorate, and microcracks can
form immediately after application of the spray. These cracks enlarge with time and can
cause entire sections of the spray to fail (Jaya Kumar et al., 2022; Dai et al., 2022).
The beam can be designed to be completely under the concrete slab or as a shallow floor
beam that is protected by the surrounding concrete mass. Alternatively, the lower part of the
steel beam may be visible in these cases, only the upper part will be fire protected. Research
into this option has been carried out by the collective around Maraveas (Maraveas et al., 2014;
Maraveas et al., 2017; Alam et al., 2021; Asteris et al., 2022).

1.2 Partially fire-protected structures


In ceiling structures with the secondary unprotected beams is heat transferred by conduction
to the main protected connections and beams due to the high thermal conductivity of the steel.
A solution is to apply the partial fire protection for the secondary beams. This protection is
called coatback in British English (Podolski et al., 2017; Breunese, 2019; Vorosin and
Pareenenko, 2021). For intumescent coatings, the temperature dependence of thermal
conductivity is not fixed, due to their chemical reactions and foaming during heating. The
behavior of the coating is determined and then the amount of foaming can be predicted. The
effective thermal conductivity is expressed in a simplified form in equation (1) (Podolski
et al., 2017).
Δθa;t dp AVp ca ρa
λp;t ¼ (1)
ðθt  θa;t ÞΔt

where λp,t is the thermal conductivity [W/(m$K)], Ap/V is the mass of the profile per meter
[1/m], dp is the thickness of the protective material [m], ca is the heat capacity of the steel
[J/(kg$K)], ρa is the density of the steel [kg/m3], θt is the gas temperature [8C], θa,t is the
temperature of the steel [8C], and Δt is the time interval t [s].
During a fire, intumescent coatings are divided into several layers, which have different
temperatures and different thermal conductivities. Therefore equation (1) does not express
the exact relationship between the temperature and the thermal conductivity of the coating.
For this reason, it is preferred to use the fire test results of thin protective coatings and to
derive the basic relationship between thermal conductivity and temperature from them.
Figure 3 shows schematically the connection of a secondary beam that is partially protected
from fire to the main beam, which is fully protected.
The partial fire
protection

Figure 3.
Scheme of a secondary
beam connected to a
primary beam with
partial fire protection

The critical temperature of the steel before loss of load-bearing capacity ranges from 580 8C to
750 8C according to BS5950 (1990). In the case of critical temperature range determination
according to EN 1993-1-2 (2005), a critical temperature range of 498–1135 8C is given. The
final critical temperature of the steel member is influenced by the load ratio, the slenderness
and the effect of lateral supports. The origins for determining the critical temperature of steel
elements were presented by Rubert and Peter (1988) and were compared with new findings in
terms of non-uniform heating (Nguyen et al., 2022). By modeling the thermal action using a
nominal standard temperature curve, it is given by (2). If equation (2) is not used, the effect of
convection and radiation in equation (3) is considered when calculating the gas temperature.
θg ¼ 20 þ 345 3 log ð8 t þ 1Þ (2)

where θg is the gas temperature near the protected steel beam [8C] and t is the time [in min].
The heat flow within the secondary elements is a three-dimensional effect, especially near the
connection to the main loadbearing member. In the fire-rated section, the secondary beam is
protected according to the fire protection requirements. The assumption is that the entire
perimeter of the unprotected element is evenly exposed. This idealization is consistent with
the assumption that the beam acts as a member. The member is insulated with protective
material at the point of connection to the primary member. The heat flux from the surface can
be expressed as an average value. The heat flux through the member is based on Fourier’s
law of conduction. The total heat entering the differential volume can be expressed according
to (3) (Yasseri, 2002).
 
JSFE A
d½qðxÞAðxÞ ¼ h_ ðθðxÞ  θg Þ (3)
V

Where q(x) is the heat flow [W/m2], A(x) is the surface [m2], h h_ is the conductive heat transfer
coefficient in time [W/(m2$K)], A/V is the mass per meter of the steel profile [1/m], and θg is the
temperature of the gas [8C].
By including the dependence of the heat flux on the time and in the element, equation (3)
can be modified as follows (4):
dqðxÞ v2 θ
¼ −KðxÞ 2 dx (4)
dt vx
where K is thermal conductivity [W/(m$K)].
The second order partial derivative in equation (4) is due to the dependence on time and on
temperature.
By modifying equation (4) by the heat capacity c and density ρ, equations (5) – (8) are
modified:
dqðxÞ vθðxÞ _ A
¼ cρAðxÞ þ h ðθðxÞ  θg Þ (5)
dt vx V
vθðxÞ _ A v2 θ
cρAðxÞdx þ h ðθðxÞ  θg Þ ¼ −KðxÞ 2 dx (6)
vt V vx

Combining equations (5) and (6), we can determine equation (7) in the time derivative.
Combining equations (5) and (6) we can determine equation (7) in the time derivative. Or by
further modification, equation (8) can be obtained. Equation (8) already includes the effect of
thermal diffusion, which is represented by cKρ.
vθ A v2 θ
cρ ¼ h_ ðθ  θg Þ þ KðxÞ 2 dx (7)
vt V vx
or
K
vθ K v2 θ cρ A
¼  ðθ  θg Þ (8)
vt cρ vx2 K V

The following ideas for modifying the equations were established by Yasseri (2002).
To achieve a closed-form solution according to equation (7), only the length of the isolated
element is considered; see Figure 4. This implies the assumption that the end of the element
suddenly heats up to the same temperature after passing through the fire. Therefore, the
boundary conditions are for t 5 0 and x 5 x, θ 5 θi is the initial temperature of the beam. For
t 5 0 and x 5 0, θ 5 θg the rapid increase in temperature is at the end of the beam.
Differential equation (7) can be solved using the finite element (FE) method or the finite
difference method, or alternatively, by the method of separation of variables. The shape of the
bar combined with the boundary conditions suggests that the solution can be expressed
using intermediate solutions that already exist. The special shape of a finite-length rod is
understood to mean that it can be formed by common points at the mutually perpendicular
intersection of an infinite plate of thickness L and an infinite rod of circumference PH and
region A. When substituted in equation (8), the shape of Φ can be extracted as equation (9).
When substituting into equation (8), we obtain the form (10).
Φ ¼ θ  θg (9) The partial fire
K _ protection
vΦ K v2 θ cρ
hA
¼  Φ (10)
vt cρ vx2 kV

The boundary conditions change for t 5 0 and x 5 x to Φ 5 θt – θg and for t 5 0 and x 5 0 to


Φ 5 0. If the solution sought is of the form (11) then Φ1 Φ2 must satisfy equation (10) and the
given conditions. After substituting both variables into equation (10) and then adjusting
them, the following equation (12) is obtained.
Φ ¼ Φi ðx; tÞΦ2 ðtÞ (11)
0 1
  _
vΦ1 K vΦ1 @vΦ2  K hA Φ2 A
Φ1  ¼ −Φ 1 (12)
vt cρ vx2 vt cρ KV

Both sides of the equation are functions of one variable, so they is equal to a constant.
Equation (12) is a function of all values, so the value of the parentheses is zero. The following
equations (13) – (17) follow from this fact.
vΦ1 K vΦ1
¼ (13)
vt cρ vx2
vΦ2 _
K hA
¼− Φ2 (14)
vt cρ KV
0 1
Φ B x C
¼ erf B ffiC
@ qffiffiffiffiffi A (15)
Φ1 2 Kt cρ

0 1
Z px ffiffiffiffi
B x C 2
erf B ffiC
K
e−z dz
2

@ qffiffiffiffiffi A¼π
2 cρ t
(16)
2 cKρ t 0

Figure 4.
Boundary condition
drawing with
coordinate system
0 1
JSFE K _
hA
Φ 2 ¼ e@ At (17)
cρ KV

The full solution is the following equation (18).


0 1
θ  θg B x C −K h_ At
¼ erf B ffiC
@ qffiffiffiffiffi Ae
cρ K V (18)
θ i  θg 2 Kt cρ

In equation (18), which in this case considers the added insulation material in h,_ K refers to the
thermal conductivity of the steel. For a prismatic bar, A/V 5 Hp/A, which is the thermal form
factor. The thickness of the fire protection is calculated from A/V 5 Hp/A of the member. The
differential governing equation is equation (18). All the parameters except the length of
protection against fire are known (Yasseri, 2002).

2. Fire experiment
2.1 Description
The aim of the experiment was to obtain data to validate the numerical model. The experiment
consisted of four steel samples that were submitted according to a standard temperature curve
for 60 min. The test was carried out in a small-scale fire furnace at University Center for Energy
Efficient Buildings CTU in Prague (Sejna,  2021). The samples had dimensions of
300 3 60 3 8 mm. Half of the length was treated on three sides with the fire spray. The
thickness of the coating was 10 mm for two of the samples (Sample 3 and 4) and 20 mm for the
other two samples (Sample 1 and 2). The protection thicknesses were chosen based on the
applicability of the technology, where the tolerance in creating the protective surface reflected
the different final protection thicknesses. The front of the sample was also treated with the spray.
Thermocouples MTC10/1xK T2/P1,5/N3000/K-MM were placed on each sample on protected
and unprotected parts. The geometry of the samples is shown in Figure 5.

Figure 5.
Sample geometry
Subsequently, all samples were left to mature at 20 8C for 28 days. Five thermocouples The partial fire
were applied to two samples, and then only the two outermost thermocouples were applied to protection
the other two samples, as shown in Figure 5. The thermocouples were carried out of the
furnace through holes in the ceiling structure and were then connected to the recording
device.
The time of the experiment was 60 min. At the end of the experiment, the burner was
turned off and data continued to be measured for another 15 min. During the test, at a time of
35 min, the pressure in the gas cylinders dropped and the burner power was reduced due to a
lack of gas. The prepared samples are shown in Plate 1, and the samples after the experiment
are shown in Plate 2.

2.2 Temperatures
A comparison between the average temperature in the upper part of the furnace and the
standard temperature curve is shown in Figure 6. During the test, thermocouples No. 9 and
No. 13 showed large differences in measured temperatures, so their values were not used for
validating the numerical model. Comparing all the results, samples with the same thickness of
protection showed similar measured values. Samples with different protection thicknesses
had differences in temperatures of up to 100 8C on the protected part. The measured
temperatures are shown in Figures 7–10.

Plate 1.
Prepared samples
before the experiment

Plate 2.
Samples after the
experiment
JSFE

Figure 6.
Temperature
according to the
standard temperature
fire curve and the
furnace temperature in
the experiment

Figure 7.
Sample 1 – steel
temperatures

Figure 8.
Sample 3 – steel
temperatures
The partial fire
protection

Figure 9.
Sample 2 – steel
temperatures

Figure 10.
Sample 4 – steel
temperatures

3. Numerical model
3.1 Material characteristics
The thermal conductivity and the heat capacity of the steel were taken from Li and
Wang (2013). The most important characteristic of fire protection materials is their
thermal conductivity. The value of this variable has the greatest influence on the
temperature of the protected element due to the light weight of the protective materials.
This value must therefore be determined as accurately as possible. However, the
problem is the change in the thermal conductivity of materials at elevated temperatures.
Using a constant value undervalues the resulting temperature of the protected element
and thus reduces the safety of the design. For this reason, a constant value cannot be
used. There are several methods for measuring thermal conductivity, but their
disadvantage is their high cost. Thermal conductivity can be obtained using a
theoretical model, but its use is limited. The values of the material characteristics used
in the models are listed in Table 1.
JSFE Steel
Temperature [8C] 20 700 735 765 900 1000
Density [kg/m3] 7850
Heat capacity [J/(kg$K)] 440 1008 5000 1069 650 650
Thermal conductivity [W/m2] 54 30 29 28 27 27
Coatback
Temperature [8C] 20 200 400 600 800 1000
Density [kg/m3] 440
Heat capacity [J/(kg$K)] 924 968 1032 1112 1208 1320
Table 1. Thermal conductivity [W/m2] 0.079 0.096 0.137 0.212 0.33 0.5
Material characteristic Source(s): Author’s own creation

3.2 Model
Ansys Mechanical is a program that provides a mechanical and thermal analysis of
problems using the FE method. The basic principle of the FE model is the partitioning of
the system geometry into small points called nodes. In each node, there is a set of degrees
of freedom such as temperature, displacement, etc. These degrees of freedom vary
depending on the input to the system. For some elements, e.g. beams, the solution is
known; for other elements, the interaction of the degrees of freedom is estimated by
integration over the element. To increase the accuracy of the analysis, a sufficient
number of elements is used to solve the FE model. The element type was defined, for
example, Plane 55 for a 2D model and Solid 70 for a 3D model for thermal analysis. The
incremental time step was set to a maximum value of 100 s for the first step of the
calculation, and the average time increment was chosen to be 1 s. The convergence
criterion was chosen to be 103. Implicit analysis was chosen for the solution method.
Then, the properties of the material such as density, specific heat capacity and
thermal conductivity were defined. Then the geometry of the model created an FE
model using the MESH command. The loads and constraints were applied, the
solution conditions were applied, and the problem was solved. The elements were
identified by the name of the group that they belonged to and then by a number. This
number was the specific identifier of the element routine. Each element had many
properties based on its name. These included the characteristic and degenerate shape,
the number of nodes, the degrees of freedom, the real constants, the key options, the
material properties, the allowable loads and the special properties (Stolarski et al., 2018;
Stolarski et al., 2019).
The first boundary condition represented the constraints. This was a pair of input/output
values that defined the value for a given degree of freedom. It specified the value of the
solution for a given independent variable at a particular location. The constraints were based
on the physical situation. For example, the boundary condition indicated zero displacement
for the clamped end of a beam or a known deflection at the opposite end. Similarly, the
boundary condition indicated the temperature during thermal analysis. The second
boundary condition represented the loads. These were deviations in pressure, temperature,
etc. applied to the system from an external source. The purpose of the simulations was to
determine the behavior of the system with respect to these loads. The loads were applied in
the time variable analysis before time t 5 0 and did not change with time (Thompson and
John, 2017). The boundary conditions can be organized according to physics, according to an
element or according to an interaction.
The model resolved the ceiling area at the joint of the beam and the column. In the
modeling, the column was assumed to be perfectly insulated, and all the thermal load of
the column had been transferred from the draft tube and from its fire protection. Heating of The partial fire
the protected and unprotected part of the structure was considered from three sides; the protection
ceiling structure was assumed to be from the top side. The column was protected by a fire
protection system of the same thickness as the layer at the point of connection. A schematic
representation of the model is provided in Figure 11, Figures 12 and 13. The convection and
radiation follow the standard temperature curve. A SOLID 70 element was used, which has
eight nodes and one degree of freedom.
The cross-sections of the airfoil were varied from steel profile IPE 180 to IPE 270. The
column was then varied from the steel profile HEB 200 to HEB 280 cross-section. The length

Figure 11.
3D FE model

Figure 12.
FE model – section
JSFE

Figure 13.
FE model – MESH

of fire protection from the connection point ranged from 300 mm to 700 mm and the thickness
of the protection ranged from 10 mm to 30 mm.

3.3 Sensitivity analysis


The mesh sensitivity study was performed in terms of the number of nodes in the model
under the same boundary conditions. The average temperature at the duct and at the column
connection point was compared. The results are presented in Table 2.
For the accuracy of the model, it was necessary to determine the value of the thermal
conductivity coefficient of fire protection λ. For this purpose, calculations with different
coefficients were calculated and were compared step by step, and the subsequent results were
compared with the numerical model. A spray length of 200 mm with a thickness of 50 mm was
used as a control. The comparison was carried out with a simulation time of 120 min. Individual
results are shown in Table 3. For the purposes of this paper, the thermal conductivity
coefficient λ 5 0.413 W/(m$K) is used. This value corresponds to the 80% quartile.

Number of nodes
5723 12,599 19,652 28,728 31,647

Time [min] Temperature [8C]


0 20.0 20.0 20.0 20.0 20.0
15 38.0 37.4 37.5 37.1 37.0
30 76.7 76.4 76.5 75.8 75.8
Table 2. 45 115.8 111.4 115.5 114.8 114.6
Mesh sensitivity 60 151.2 150.8 150.8 150.1 150.0
analysis Source(s): Author’s own creation
3.4 Parametric analysis The partial fire
The results of more than fifty simulations are presented in the following parts, in which the protection
main variables are changed and their influence on the resulting temperature at the
connection point of the duct and column is investigated. The resulting temperature is
taken as the average temperature between the temperature of the bottom flange and of
the top flange and the temperature at the beam stand. In the simulation, the connection of
the beam and column is assumed to be by welds; these welds are ignored during the
simulation.
The first parameter is the variation of the fire protection length. The effect of changing the
fire protection length is investigated on an IPE 400 beam with a constant fire protection
thickness of 30 mm. The resulting temperatures of the model with the column are shown in
Figure 14. A comparison of the calculated temperatures shows that the effect of the unheated
column on the resulting temperature at the connection point is significant. However, this
simplifying effect contained in the model does not affect the other parameters investigated
and their trends.
The second parameter is the thickness of the fire protection; for this purpose, a model case
is created with IPE 200 with a fire protection length of 400 mm and thickness varying from 10
to 30 mm. The results are shown in Figure 15. As the thickness of the fire protection increases
the temperature at the connection point decreases. The difference between the smallest and
the greatest thickness tested is 121 8C. During the parametric studies, the influence of the
beam profile, which varies from IPE 180 to IPE 270, was also investigated. All cross-sections
have the same thickness and length of fire protection. From the temperature histories, it can
be observed that the smaller the cross-section, the lower the temperature achieved during the
simulation. For example, the difference between the cross-section of IPE 180 and IPE 270 is
150 ᵒC. For an IPE 200 beam with a fire protection length of 300 mm and a thickness of 10 mm,

Thermal conductivity [W/(m$K)]


0.103 0.170 0.297 0.331 0.413 0.500 ANSYS

Time [min] Temperature [8C] Table 3.


120 584.3 589.6 599.5 602.1 608.2 614.1 618.0 Thermal conductivity
Source(s): Author’s own creation sensitivity

Figure 14.
Temperatures in the
center of column face,
the fire protection
thickness 30 mm and
length L 5 100, 200,
400 and 7000 mm,
material properties of
coatback by Table 1,
beam IPE 400 and
column HEB 200
JSFE the influence of the column cross-section was investigated when a cross-section from HEB
200 to HEB 280 was considered. For example, the temperature difference between HEB 200
and HEB 280 is 50 ᵒC.

3.5 Required length of the partial fire protection


An analytical approach was used for the possible design of the required partial fire protection
length, which was compared with the numerical solution. An integral objective was to verify
the precision of the generally accepted statement of a fire protection length of 500 mm. The
resulting use of the numerical model has two phases; in the first step, equation (18) is applied
and, according to this equation the required fire protection length for the required resistance
of 120 min is determined, during which the average temperature of the element at the
connection point must not exceed 500 8C. In the second step, the temperatures are determined
for a model case of a beam with varying lengths of fire protection. The analytical model is
based on assumptions published in Yasseri (2002). The steel input parameters are given in
Table 1. The temperature of the element at time 0 is 20 8C. The fire temperature is determined
according to equation (2).
By comparing the results of the analytical model with numerical experiments, fire
protection lengths of 100 mm–700 mm were monitored. The analytical model considers the
use of an ideal thickness of fire protection using the cross-sectional area factor Am/V. The
simulations were compared with the spray thicknesses according to the dimensioning tables
for PROMASPRAY® P300 and PROMASPRAY® F250 (Breunese, 2019). The simulation
uses the spray thickness corresponding to fire resistance of 120 min for the IPE 400 profile
loaded on three sides by fire according to equation (2). The results of the individual
simulations are shown in Table 4.
Table 4 shows that both models have a partial fire protection distance of up to 400 mm,
corresponding to PROMASPRAY® F250. For greater protection distances, the thickness is
increased additionally. It is necessary to take into consideration that the original equation (18)
works with the design temperature to which fire protection is designed. Therefore, for a
protection distance of 700 mm, it is necessary to use the dimensioning tables for a
temperature of 200 8C. Verifying the requirement of the partial fire protection sizing table at
70 mm, the temperature in the numerical model comes out at 208 8C, which confirms the
analytical method.

Figure 15.
The steel temperature
in column face during
heating, the fire
protection thickness
d 5 10; 15, 25 and 30
and length 500 mm,
material properties of
coatback by Table 1,
beam IPE 200, column
HEB 200 and fire
protection
length 500 mm
Two major studies were carried out to finally verify the required length of the partial fire The partial fire
protection. The first study investigated the effect of partial fire protection thickness on protection
temperature development. For the first study, an IPE 220 beam profile, an HEB 300 column
profile and a protection length of 500 mm were chosen. The numerical model was
investigated for a fire duration of 120 min. The temperature development of the steel
profile along the length of the partial fire protection is shown in Figure 16, Figures 17
and 18.
In the second study, the effect of a protective coating length of 500 mm and thickness
of 30 mm was verified for different IPE cross-sections exposed to fire from three sides
for 120 min. The temperature history of the gas acting on the beam in the FE model
follows equation (2). The resulting values for each cross-section are summarized in
Table 5.
It is necessary to consider not only the cross-section of the protected element but also other
variables when designing partial fire protection. The fully protected element to which the
partially protected element is connected plays an important role in the resulting length of fire
protection. Based on the simulations that were performed, it can be predicted that the larger
the protected element, the more heat it will deliver to the point of connection when exposed to
fire. It is also necessary to include in the design the expected sides of the structure that are
exposed to fire. A structure presented on three sides has different values from the same
structure presented on four sides. The type of fire protection, especially its thickness, is very
important. If the thickness of the protection is wrongly designed, the structure will overheat

Analytical model Numerical model


Length of partial fire protection [mm] temperature [8C] temperature [8C] Analytical/Numerical ratio

100 733 733 1.0


200 608 618 0.98
300 494 507 0.97 Table 4.
Comparison between
400 396 436 0.91
the analytical model
500 315 393 0.08 and the numerical
600 252 368 0.68 model, partial fire
700 205 354 0.58 protection
Source(s): Author’s own creation thickness 50 mm

Figure 16.
The steel temperature
along the beam from
the column face in
15 min, the fire
protection thickness
d 5 10; 20 and 30 mm
and length 500 mm,
material properties of
coatback by Table 1,
beam IPE 220 and
column HEB 300
JSFE before the time guaranteed for a correctly designed length. When designing the length of
partial fire protection, it is necessary to include the required fire resistance period.
A comparison of the proposed fire protection lengths for 15, 30 and 60 min is shown in
Table 6. Again, the overpressure is loaded by fire from three sides. The protection thickness is
selected according to the spray characteristics (Breunese, 2019) dimensioning tables for 15, 30
and 60 min.

4. Conclusions
Ways of providing passive protection of steel members, especially partial fire protection,
have been summarized. The rate of temperature loss in the member during partial fire
protection has been investigated. The rate was confirmed to be high, due to the heat
conduction through the steel and the rate of change of the gas temperature during the fire.
The existing formulas have been modified to a time-dependent temperature progression
subroutine. The approach for the theoretical design of partial fire protection of steel members
has been explained.

Figure 17.
Temperature
development under fire
protection when
changing the thickness
Lp of the partial fire
protection, IPE 220
beam, column HEB
300, fire protection
length 500 mm and
temperatures in
time 30 min

Figure 18.
Temperature
development under fire
protection when
changing the thickness
dp of the partial fire
protection, IPE 220
beam, column HEB
300, fire protection
length 500 mm and
temperatures in
time 60 min
Profile height Length of Length of protection/ Average temperature at the point
The partial fire
Profile [mm] protection [mm] profile height of connection [8C] protection
IPE 80 80 500 6.25 453
IPE 100 500 5.00 348
100
IPE 120 500 4.17 266
120
IPE 140 500 3.57 336
140
IPE 160 500 3.13 275
160
IPE 180 500 2.78 292
180
IPE 200 500 2.50 309
200
IPE 220 500 2.27 331
220
IPE 240 500 2.08 359
240
IPE 270 500 1.85 356
270
IPE 300 500 1.67 351
300
IPE 330 500 1.52 377
330
IPE 360 500 1.39 399
360
IPE 400 500 1.25 407
400
IPE 450 500 1.11 413
450
IPE 500 500 1.00 415
500 Table 5.
Beam temperatures at
IPE 550 500 0.91 426
the connection using a
550 partial fire protection
IPE 600 500 0.83 425 length 500 mm and
600 thickness 30 mm and
Source(s): Author’s own creation fire time 120mins

Experiments were carried out to observe the heat conduction behavior under partial fire
protection. Four steel specimens protected by partial coatings were exposed to fire for 60 min.
Two specimens had a protection thickness of 10 mm, and the other two specimens had a
protection thickness of 20 mm. The gas temperature curve in the experimental furnace was
comparable to the standard temperature curve under fire for up to 35 min. The measured
temperature during the experiment has been presented.
Numerical and analytical studies for partial fire protection were performed. The empirical
and analytical formulas for the temperature of steel members under partial fire protection
were summarized, validated and verified using the FE model. The parametric study
investigated the effect of the thickness and the length of partial fire protection for different
steel member profiles. The results were presented using graphs and tables for ease of use. An
analytical study was carried out for the commonly accepted length of 500 mm of partial fire
JSFE Time 15 min 30 min 60 min
Length of Length of Length of
partial fire Temperature partial fire Temperature partial fire Temperature
Profile protection [mm] [8C] protection [mm] [8C] protection [mm] [8C]

IPE 80 100 276 200 409 300 367


IPE 100 269 398 361
IPE 120 261 3866 373
IPE 140 253 375 374
IPE 160 246 365 398
IPE 180 240 355 402
IPE 200 234 347 426
IPE 220 226 337 416
IPE 240 221 328 429
IPE 270 215 320 420
IPE 300 209 312 421
IPE 330 203 304 421
IPE 360 195 291 437
Table 6. IPE 400 189 282 436
Beam temperatures at
IPE 450 181 271 431
the connection for
different partial fire IPE 500 173 260 425
protection lengths and IPE 550 166 249 420
thicknesses, from IPE 600 158 238 413
(Breunese, 2019) Source(s): Author’s own creation

protection for steel members. The resulting average temperatures of the steel elements were
compared.
It can be concluded that, when properly designed, the use of partial fire protection is a sound
practice in assessing the fire resistance of structures. The procedures presented in this paper are
partial fire protection is appropriate when membrane action is taken into account.

References
Alam, N., Maraveas, C., Tsavdaridis, K.D. and Nadjai, A. (2021), “Performance of ultra shallow floor
beams (USFB) exposed to standard and natural fires”, Journal of Building Engineering, Vol. 38,
102192, ISSN 23527102, doi: 10.1016/j.jobe.2021.102192.
Asteris, P.G., Maraveas, C., Chountalas, A.T., Sophianopoulos, D.S. and Alam, N. (2022), “Fire
resistance prediction of slim-floor asymmetric steel beams using single hidden layer ANN
models that employ multiple activation functions”, Steel and Composite Structures, Vol. 44
No. 6, pp. 755-774, doi: 10.12989/scs.2022.44.6.769.
Azmi, Y., Ahmad, F., Kabir, S., Lee, Y.X., Zulfiqar, A., Yeoh, G.H., Qaiser, A. and Masset, P.J. (2022),
“Investigating the mechanical performance of intumescent coating enhanced with magnesium
oxide (MgO) for structural steel application”, ICPER 2020: Proceedings of the 7th International
Conference on Production, Energy and Reliability, Springer Nature Singapore, Singapore, ISBN
978-981-19-1938-1, doi: 10.1007/978-981-19-1939-8_69.
Breunese, A.A. (2019), “Simple method for calculation of coat back lengths”, PROMAT,
Vol. 2019, pp. 1-9.
BS5950 (1990), Structural Use of Steelwork in Building: Part 1. Code of Practice for Design in Simple
and Continuous Construction: Hot Rolled Sections, British Standard, London, Vol. 1.
Dai, X., Qian, J., Qin, J., Yue, Y., Zhao, Y. and Xingwen, J. (2022), “Preparation and properties of
magnesium phosphate cement-based fire retardant coating for steel”, Materials, Vol. 15 No. 12,
p. 4134, ISSN 1996-1944, doi: 10.3390/ma15124134.
Dai, X., Wang, Y. and Bailey, C. (2010), “A simple method to predict temperatures in steel joints with The partial fire
partial intumescent coating fire protection”, Fire Technology, Vol. 46, pp. 19-35, ISSN 0015-2684,
doi: 10.1007/s10694-009-0095-7. protection

Celko,  ak. I., Klakurkova, L. and Obrtlık, K. (2017), “Durability of
L., Jech, D., Dvorak, K., Sul
amorphous and crystalline BMAS thermal barrier coatings produced by plasma spraying”,
Trans Tech Publications, Solid State Phenomena, Vol. 258, ISSN 1662-9779, doi: 10.4028/www.
scientific.net/SSP.258.383.
EN 1993-1-2 (2005), Eurocode 3: Design of Steel Structures - Part 1-2: General Rules - Structural Fire
Design, The European Union Per Regulation EU, Brussels, Vol. 1.
Fan, C., Li, Y., Gao, Y., Yan, L., Zhu, D., Ou, C. and Wang, Z. (2022), “Effect of ambient wind on the
flame retardancy of intumescent coatings”, Journal of Thermal Analysis and Calorimetry, pp.
1-13, ISSN 1388-6150, doi: 10.1007/s10973-022-11593-0.
Ferreira, F.A., Claret, A.M. and Santolin, A. (2007), “A simplified method for assessing the temperature
distribution in partially protected steel sections”, Rem: Revista Escola de Minas, Vol. 60, pp.
645-655.
H€aßler, D. and Hothan, S. (2022), “Performance of intumescent fire protection coatings applied to
structural steel tension members with circular solid and hollow sections”, Fire Safety Journal,
Vol. 131 No. 2022, 103605, ISSN 03797112, doi: 10.1016/j.firesaf.2022.103605.
Hautala, J., Kangashaka, I., Malaska, M. and Pajunen, S. (2021), “Resistance of partially protected steel
beams in fire”, ce/papers 4.2-4, pp. 1849-1855, ISSN 2509-7075, doi: 10.1002/cepa.1495.
Imran, M., Liew, M.S., Nasif, M.S., Gracia, E.M., Danyaro, K.U., Niazi, M.U., Raza, U.L., Mustafa,
M., Bin Othman, I., Latheef, M., Bayu Endrayana, D., Zulaikha, B. and Yusof, N. (2018),
“Thermal and mechanical response of partially protected steel I-beam under fire”, MATEC
Web of Conferences, Vol. 203, EDP Sciences, ISSN 2261-236X, doi: 10.1051/matecconf/
201820306009.
Jaya Kumar, G., Tattukolla, K., Anand, N. and Al-Jabri, K. (2022), “Influence of fire-resistant
coating on the physical characteristics and residual mechanical properties of E350 steel
section exposed to elevated temperature”, Journal of Structural Fire Engineering, ISSN
2040-2317, doi: 10.1108/JSFE-02-2022-0008.
Jo, H.M., Kim, D.H., Lee, S.H. and Lee, J.Y. (2022), “Multi-layer barrier coating technology using
nanofibrillated cellulose and a hydrophobic coating agent”, BioResources, Vol. 17 No. 4, ISSN
19302126, doi: 10.15376/biores.17.4.6222-6233.
Jong, K.Y.J., Yew, M.C., Yew, M.K., Ting, C.H., Saw, L.H., Ching, N.T., Yeo, W.H. and Beh, J.H. (2021),
“Fire resistance and mechanical properties of the fire-resistant board”, Advances in Material
Science and Engineering: Selected Articles from ICMMPE 2020, Springer, Singapore, ISBN 978-
981-16-3640-0, doi: 10.1007/978-981-16-3641-7_29.
Kordosky, A.N., Drury, M.M. and Quiel, S.E. (2020), “Structural fire resistance of partially restrained,
partially composite floor beams, I: experiments”, Journal of Constructional Steel Research,
Vol. 167, 105945, doi: 10.1016/j.jcsr.2020.105946.
Li, G. and Wang, P. (2013), Advanced Analysis and Design for Fire Safety of Steel Structures, Springer
Science and Business Media, Berlin, ISBN 978-3-642-34392-6, doi: 10.1007/978-3-642-34393-3_3.
Li, Q., Hou, P. and Shang, S. (2022), “Accurate 3D thermal stress analysis of thermal barrier coatings”,
International Journal of Mechanical Sciences, Vol. 217, 107024, ISSN 00207403, doi: 10.1016/j.
ijmecsci.2021.107024.
Maraveas, C., Wang, Y.C. and Swailes, T. (2014), “Fire resistance of 19th century fireproof flooring
systems: a sensitivity analysis”, Construction and Building Materials, Vol. 55, pp. 69-81, ISSN
09500618, doi: 10.1016/j.conbuildmat.2014.01.022.
Lucherini, A., Hidalgo, J.P., Torero, J.L. and Maluk, C. (2023), “Numerical heat transfer model for
swelling intumescent coatings during heating”, International Journal of Thermal Sciences,
Vol. 184, 107922, ISSN 12900729, doi: 10.1016/j.ijthermalsci.2022.107922.
JSFE Maraveas, C., Tsavdaridis, K.D. and Ali, N. (2017), “Fire resistance of unprotected ultra shallow floor
beams (USFB): a numerical investigation”, Fire Technology, Vol. 53, pp. 609-627, ISSN 0015-
2684, doi: 10.1007/s10694-016-0583-5.
Milionis, A., Antonini, C., Jung, S., Nelson, A., Schutzius, T.M. and Poulikakos, D. (2017), “Contactless
transport and mixing of liquids on self-sustained sublimating coatings”, Langmuir, Vol. 33
No. 8, pp. 1799-1809, ISSN 0743-7463, doi: 10.1021/acs.langmuir.6b04377.
Munoz-Garcia, E. (2016), “Analysis and design challenges and solutions of 3 sided passive fire
protection PFP on steel beam decks for offshore installations”, Offshore Technology Conference
Asia, OnePetro, doi: 10.4043/26623-MS.
Nguyen, X.T., Nguyen, T.Q., Ho, X.B., Nguyen, H.Q. and Nguyen, T.N.M. (2022), “Numerical
investigation on critical temperature of steel I-beams under non-uniform heating condition—
comparison with eurocode 3”, Computational Intelligence Methods for Green Technology and
Sustainable Development: Proceedings of the International Conference GTSD2022, Springer
International Publishing, Cham, ISBN 978-3-031-19693-5, doi: 10.1007/978-3-031-19694-2_25.
Piperopoulos, E., Grifo, G., Scionti, G., Atria, M., Calabrese, L., Consolo, G. and Proverbio, E. (2022),
“Study of intumescent coatings growth for fire retardant systems in naval applications:
experimental test and mathematical model”, Coatings, Vol. 12 No. 8, p. 1180, ISSN 2079-6412,
doi: 10.3390/coatings12081180.
Podolski, D., Wang, Y.-C. and Yuan, J. (2017), “Numerical investigation of intumescent ‘coatback’on
unprotected secondary beams”, in Applications of Fire Engineering, CRC Press, pp. 161-170,
ISSN 9781315107202, doi: 10.15127/1.309552.
Rubert, A. and Peter, S. (1988), “Critical temperatures of steel columns exposed to fire”, Fire Safety
Journal, Vol. 13 No. 1, pp. 39-44, ISSN 03797112, doi: 10.1016/0379-7112(88)90031-8.

Sejna, J. (2021), “Small furnace experiments for wood burning pyrolysis models”, Civil Engineering
Research Journal, Vol. 12 No. 3, ISSN 25758950, doi: 10.19080/CERJ.2021.12.555838.
Singh, R. and Samanta, A. (2022), “Proposals for critical temperature of cold-formed steel compound
sections under various load patterns”, Asian Journal of Civil Engineering, Vol. 23 No. 4,
pp. 609-633, ISSN 1563-0854, doi: 10.1007/s42107-022-00444-w.
Stolarski, T., Nakasone, Y. and Yoshimoto, S. (2018), Engineering Analysis with ANSYS Software,
Butterworth-Heinemann, Oxford, ISBN 978-0-08-102164-4.
Stolarski, T., Nakasone, Y., Yoshimoto, S. and Winter, W. (2019), “Engineering analysis with ANSYS
software”, Applied Mechanics and Materials, Vol. 887, pp. 21-29, ISSN 1662-7482, doi: 10.1016/
C2016-0-01966-6.
Suzuki, M., Shahien, M., Shinoda, K. and Akedo, J. (2022), “The current status of environmental barrier
coatings and future direction of thermal spray process”, Materials Transactions, Vol. 63 No. 8,
pp. 1101-1111, ISSN 1345-9678, doi: 10.2320/matertrans.MT-T2021003.
Tamas-Benyei, P. and Santha, P. (2023), “Potential applications of basalt fibre composites in thermal
shielding”, Journal of Thermal Analysis and Calorimetry, Vol. 148 No. 2, pp. 271-279, ISSN 1388-
6150, doi: 10.1007/s10973-022-11799-2.
Thompson, M.K. and John, M.T. (2017), ANSYS Mechanical APDL for Finite Element Analysis,
Butterworth-Heinemann, Oxford, ISBN 9780128131107.
gs€oz, A.E. and Çarbaş, S. (2018), “Structural features of cold-formed steel
Tunca, O., Erdal, F., Sa
profiles”, ISSN 2149-8024, doi: 10.20528/cjsmec.2018.02.005.
Vagge, S.T. and Ghogare, S. (2022), “Thermal barrier coatings”, Materials Today: Proceedings, Vol. 56
No. 2022, pp. 1201-1216, ISSN 22147853, doi: 10.1016/j.matpr.2021.11.170.
Vorosin, A.O. and Pareenenko, A.P. (2021), “The research into the heating effect of secondary steel
structures, having no fire proofing, on the fire resistance of fireproof steel beams”,
Pozharovzryvobezopasnost/Fire and Explosion Safety, Vol. 30 No. 3, pp. 16-30, ISSN 2587-
6201, doi: 10.22227/0869-7493.2021.30.03.16-30.
Wang, Y., Burgess, I. and Wald, F. (2012), Performance-based Fire Engineering of Structures, CRC The partial fire
Press, Boca Raton, ISBN 978-0-415-55733-7.
protection
Wang, P., Wang, X., Liu, M. and Zhang, L (2016), “Web-post buckling of fully and partially protected
cellular steel beams at elevated temperatures in a fire”, Thin-Walled Structures, Vol. 98, pp.
29-38, ISSN 02638231, doi: 10.1016/j.tws.2015.02.028.
Xiong, M.-X., Huang, Z.-Y. and Liew, J.Y.R. (2017), “Modified critical temperatures for steel design
based on simple calculation models in eurocode 3”, Fire Technology, Vol. 53, pp. 227-248, ISSN
0015-2684, doi: 10.1007/s10694-015-0522-x.
Yang, W., Abu Bakar, B.H., Mamat, H. and Gong, L. (2022), “A new, green, recyclable fireproof insulation
board for use in integrated composite structure fire protection systems”, Fire, Vol. 5 No. 6, p. 203,
ISSN 2571-6255, doi: 10.3390/fire5060203.
Yasseri, S.F. (2002), “Coat-back length: in passive fire protection of offshore installations”,
International Conference on Offshore Mechanics and Arctic Engineering, Vol. 36126, ISBN
0-7918-3612-6, doi: 10.1115/OMAE2002-28537.
Zaharia, R., Vulcu, C., Vassart, O., Gernay, T. and Franssen, J.M. (2013), “Numerical analysis of
partially fire protected composite slabs”, Steel and Composite Structures, Vol. 14, p. 1, ISSN
1229-9367, doi: 10.12989/scs.2013.14.1.021.

Corresponding author

Jakub Sejna can be contacted at: jakub.sejna@fsv.cvut.cz

For instructions on how to order reprints of this article, please visit our website:
www.emeraldgrouppublishing.com/licensing/reprints.htm
Or contact us for further details: permissions@emeraldinsight.com

You might also like