Professional Documents
Culture Documents
BS 5400 Part 10 Fatigue - Steel Concrete and Composite Bridges PDF
BS 5400 Part 10 Fatigue - Steel Concrete and Composite Bridges PDF
Part 10 : 1980
ICS 93.040
Summary of pages
The following table identifies the current issue of each page. lssue 1 indicates that a page has been introduced
for the first time by amendment. Subsequent issue numbers indicate an updated page. Vertical sidelining on
replacement pages indicates the most recent changes (amendment, addition, deletion).
1 2 50 blank
2 2 51 2
3 2 52 blank
4 2 53 2
4a 1 54 blank
4b blank Inside back cover original
5 2 Back cover 2
a
BS 5400 : Part 10 : 1 980 Issue 1, March 1999
Contents
Page Page
Foreword 1 8.2.2 Procedure 12
Cooperating organizations Back cover 8.2.3 Adjustment factors for an, class S
Recommendations details only 12
1. Scope 2 8.3 Damage calculation, single vehicle
1.1 General 2 method 12
1.2 Loading 2 8.3.1 General 12
1.3 Assessment procedures 2 8.3.2 Procedure 12
1.4 Other sources of fatigue damage 2 8.4 Damage calculation, vehicle spectrum
1.5 Limitations 2 method 14
1.5.1 Steel decks 2 8.4.1 General 14
1.5.2 Reinforcement 2 8.4.2 Design spectrum 14
1.5.3 Shear connectors 2 8.4.3 Simplification of design spectrum 14
2. References 2 8.4.4 Calculation of damage 14
3. Definitions and symbols 2 9. Fatigue assessment of railway bridges 18
3.1 Definitions 2 9.1 Methods of assessment 18
3.2 Symbols 3 9.1.1 General 18
4. General guidance 3 9.1.2 Simplified procedure 18
I
4.1 Design life 9.2 Assessment without damage calculation 18
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
3
4.2 Classification and workmanship 3 9.2.1 General 18
4.3 Stresses 4 9.2.2 Procedure 18
4.4 Methods of assessment 4 9.2.3 Non-standard design life 18
4.5 Factors influencing fatigue behaviour 3 9.2.4 Multiple cycles 20
5. Classification of details 4 9.3 Damage calculation 20
5.1 Classification 4 9.3.1 General 20
5.1.1 General 4 9.3.2 Design spectrum for standard loading 20
5.1.2 Classification of details in table 17 4 9.3.3 Design spectrum for non-standard
5.2 Unclassified details 4 loading 20
5.2.1 General 4 9.3.4 Simplification of spectrum 20
5.2.2 Post-welding treatments 4 9.3.5 Calculation of damage 20
5.3 Workmanship and inspection 4 10. Fatigue assessment of bridges carrying
5.3.1 General 4 highway and railway loading 20
5.3.2 Detrimental effects 4 11. The Palmgren-Miner rule 20
5.4 Steel decks 4 11.1 General 20
6. Stress calculations 4 11.2 Design 0 , - N relationship 22
6.1 General 4 11.3 Treatment of low stress cycles 22
6.1.1 Stress range for welded details 4 11.4 Procedure 22
6.1.2 Stress range for welds 4 11.5 Miner's summation greater than unity 22
6.1.3 Effective stress range for non-welded Appendices
details 4 A. Basis of a,-N relationship 23
6.1.4 Calculation of stresses 4 B. Cycle counting by the reservoir method 25
6.1.5 Effects to be included 5 C. Derivation of standard highway bridge
6.1.6 Effects to be ignored 5 fatigue; loading and methods of use
6.2 Stress in parent metal 25
5 D. Examples of fatigue assessment of
6.3 Stress in weld throats other than
highway bridges by simplified methods
those attaching shear connectors
30
5 E. Derivation of standard railway load
6.4 Stresses in welds attaching shear spectra
connectors
34
6 F. Examples of stress histories and cycle
6.4.1 General 6 counting procedure
6.4.2 Stud connectors 38
6 G. Testing of shear connectors 41
6.4.3 Channel and bar connectors 6 H. Explanatory notes on detail
6.5 Axial stress in bolts 6 classification 41
7. Loadings for fatigue assessment 6
7.1 Design loadings 6 Tables
7.2 Highway loading 6 1. Annual flow of commercial vehicles
7.2.1 General 6 (nc x 108) 8
7.2.2 Standard loading 6 2. Standard load spectra for RU loading 11
7.2.3 Application of loading 6 3. Standard load spectra for RL loading 12
7.2.4 Allowance for impact 8 4. Values of k 3 for RU loading of
7.2.5 Centrifugal forces 8 railway bridges 19
7.3 Railway loading 8 5. Values of k 4 for railway bridges 19
7.3.1 General 8 6. Values of k 5 for railway bridges 19
7.3.2 Application of loading 8 7. Values of k 6 for R L loading of
7.3.3 Standard load spectra 10 railway bridges 19
8. Fatigue assessment of highway 8. Design 0,-N relationships and constant
bridges 12 amplitude non,propagating stress
8.1 Methods of assessment 12 range values 22
8.1.1 General 12 9. Mean-line or -N relationships 23
8.1.2 Simplified procedures 12 10. Probability factors 23
8.2 Assessment without damage 11. Typical commercial vehicle groups 27
calculation 12 12. Proportional damage from individual
8.2.1 General 12 groups of typical commercial vehicles 28
d 0 BSI 03-1999
Issue 2,March 1999 BS 5400: Part 10: 1980
Page Page
13. Typical commercial vehicle gross 12. Typical point load influence line 17
weight spectrum 29 13. Simplification of a spectrum 18
14. Typical commercial vehicle axle 14. Summary of design or- N curves (mean
weight spectrum 29 minus two standard deviations) 21
15. RU loading : annual traffic tonnage for 15. Summary of mean-line o r - N curves 23
standard traffic types 34 16. Typical G-,- N relationship 24
16. R L loading : annual traffic tonnage and 17. Multiple paths 28
composition of standard traffic mix 34 18. Typical Miner's summation adjustment
17. Classification of details curve 29
17(a). Non-welded details 49 19. Trains included in table 2 spectra 35
17(b). Welded details other than at end 20. Trains included in table 3 spectra 37
connections of a member 51 21. Typical example of stress concentrations
17(c). Welded details at end connections of due to geometrical discontinuity 41
22. Stress concentration factors 42
member 53 23. Failure modes at weld ends 43
Figures 24. Edge distance 43
26.
Effective width for wide lap
connections
Type 3 failure modes
44
45
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
Foreword
B S 5400 is a document combining codes of practice to Part 8 Recommendations fot materials and
cover the design and construction of steel, concrete and workmanship, concrete, reinforcement and
composite bridges and specifications for the loads, prestressing tendons
materials and workmanship. It comprises the following Part 9 Bridge beatings 1
1 Parts and Sections:
Part 1 General statement bridge bearings
Part 2 Specification for loads Section 9.2 Specification for materials,
1 Part 3 Code of practice for design of steel bridges manufacture and installation of
Part 4 Code of practice for design of concrete bridges bridgebearings
Part 5 Code of practice for design of composite
Part 10 Code of practice for fatigue
bridges
Part 6 specification for materials and workmanship,
steel
Part 7 Specification for materials and workmanship,
concrete, reinforcement and prestressing
tendons
0 BSI 03-1999 1
BS5400:Part10:1980 Issue 2,March 1999
1 . 1 General. This Part of this British Standard recommends 3.1 Definitions. For the purposes of this Part of this British
methods for the fatigue assessment of parts of bridges which Standard the following definitions apply.
are subject to repeated fluctuations of stress. 3.1.1 fatigue. The damage, by gradual cracking of a
1.2 Loading. Standard load spectra are given for both structural part, caused by repeated applications of a stress
highway and railway bridges. which is insufficient to induce failure by a single application.
1.3 Assessment procedures. The following alternative 3.1.2loading event. The approach, passage and
methods of fatigue assessment are described for both departure of either one train or, for short lengths, a bogie or
highway and railway bridges : axle, over a railway bridge or one vehicle over a highway
(a) simplified methods that are applicable to partsof bridge.
bridges with classified details and which are subjected to 3.1.3load spectrum. A tabulation showing the relative
standard loadings; frequencies of loading events of different intensities
experienced by the structure.
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
2 0 BSI 03-1999
Issue 2, March 1999
loading
of connector Nominal ultimate tensile strength, to be taken as
OU
d Number of standard deviations below the mean 1.1 G~ unless otherwise specified
line gr-Ncurve b V Value of Gr under loading from the standard
n Life time damage factor (Miner's summation for
di 2 0 fatigue vehicle (highway bridges)
120 million repetitions of a stress range 0, in a QV max (0,max - GP min) for a highway bridge
highway bridge) Values of G" (in descending order of
a v l , av2
Design stress parameter for bolts
Parameter defining the mean line 'Jr -N
. . . etc magnitude) in any one stress history for one
lane of a highway bridge
relationship The largest value of gvlfrom all stress histories
'JviA
Parameter defining the ur-N relationship for (highway bridges)
two standard deviations below the mean line The second largest value of ov, from all stress
uvi B
Value of ratio avl B / ~ V ~(highway
A bridges)
histories (highway bridges)
Miner's summation adjustment factor (highway
Qx. Qy Coexistent orthogonal direct stresses
bridges)
*Y Nominal yield strength
Proportion factor for occurrences of vehicles of
a specified gross weight (320 Kw kN) in any one
r Shear stress coexistent with oxand CJ,,
lane of a highway bridge
4. G e n e r a l g u i d a n c e
KR C Fatigue stress concentration factor for
re-entrant corners 4.1 Design life. The design life is that period in which a
hridge is required to perform safely with an acceptable
KUA Fatigue stress concentration factor for
probability that it will not require repair (see appendix A).
unreinforced apertures
Kw Ratio of actual : standard gross weights of The standard design life for the purposes of this Part of this
vehicles, trains, bogies or axles in a load British Standard should be taken as 120 years unless
spectrum otherwise specified.
ki- ka Coefficients in the simplified assessment
n procedure for a railway bridge 4.2 Classification and workmanship. Each structural steel
L Bass length of that portion of the point load detail is dassified in accordancewith table 17 (see 5.1.2). This
influence line which contains the greatest shows the maximum permitted dass for different types of
ordinate (see figure 12) measured in the structural detail. The class denoted in table 17 determines the
direction of travel design of U,- N cutve in figure 14 that may be safely used with
Applied bending moments the highest Workmanship standards specified in Part 6 for the
lnverseslope of log *,/log Ncurve detail under consideration.
Number of repetitions to failure of stress range
01
In 5.3.1 is defined the informationto be provided to the
Number of repetitions to failure of stress ranges fabricator, to ensure that the appropriate quality standards for
O r i , C r 2 . . . etc., corresponding ton,, n 2 . .. etc., Part 6 are invoked.
repetitions of applied cycles
n,.n,. Number of applied repetitions of damaging stress 4.3 Stresses. Stresses should generally be calculated in
etc. ranges Or,. U r 2 . . . etc., in a design spectrum accordance with Part 1 of this British Standard but clause6
flc Number of vehicles (in millions per year) of this Part supplements the information given in Part 1 .
- traversing any lane of a highway bridge 4.4 M e t h o d s o f assessment. All methods of assessment
flC Effective value of nc described in this Part of this British Standard are based on
nR Total number of live load cycles (in millions) for the Palmgren-Miner rule for damage calculation (see
each load proportion Kw in a railway bridge clause 11). The basic methods given respectively in 8.4
P.P, Amlied a x i a l forces and 9.3for hiahwav and railwav. bridaes " mav be used at all
P" Basic static strength of the stud . times. The simplified procedures given in 8.2and 8.3 for
Z Elastic modulus of section highway bridges and in 9.2 for railway bridges may be used
Yf Partial safety factor for load(the product Y f , . when the conditions stipulated in 8.2.1,8.3.1 and 9.2.1 are .
n
Yr,.- Y f-3 ,see Part 1) satisfied.
YfL Product of 71, . Yf2 4.5 Factors influencing fatigue behaviour. The best
Ym Partial safety factor for strength fatigue behaviour of joints is achieved by ensuring that the
d Reciprocal of the antilog of the standard structure is so detailed that the elements may deform in their
deviation of log N
Q BSI 03-1999 3
BS 5400: Part 10: 1980 Issue 2,March 1999
intended ways without introducing secondary 5.1.2.4 Class A is generally inappropriatefor bridge work and
deformations and stressesdue to local restraints. Stresses the speaal inspection standards relevant to classes B and C
may also be reduced, and hence fatigue life increased, by cannot normally be achieved in the vicinrty of welds in bridge
increased thickness of parent metal or weld metal. work. (For these and other classifications that should be used
The best joint performance is achieved by avoiding joint only when speaal workmanship is specified see the footnote to
eccenbiaty and welds near free edges and by other controls table 17.)
over the quality of the joints. Performance is adversely affected
by concentrationsof stress at holes, openings and reentrant 5.1.2.5 The classificationsof table 17 are valid for the qualities
corners. Guidance in these aspects is given in table 17 and of steel products and welds which meet the requirements of
Part 6. except where otherwise noted. For certain details the
I appendix H. The effect of residual stresses is taken into
account in the classificationtables. maximum permitted class depends on acceptance criteria
given in Part 6.
5. C l a s s i f i c a t i o n of d e t a i l s 5.2 Unclassified details
5.1. Classification 5.2.1 General. Details not fully covered in table 17 should
5.1 .I General be treated as class G, or class W for load carrying weld
5.1.1.1 For the purpose of fatigue assessment, each part of metal, unless a superior resistanceto fatigue is proved by
a constructional detail subject to fluctuating stress should, special tests. Such tests should be sufficiently extensive to
where possible, have a particular class designated in allow the design or-Ncurve to be determined in the
accordance with the criteria given in table 17. Otherwise manner used for the standard classes (see appendix A).
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
the detail may be dealt with in accordance with 5.2. 5.2.2 Post-welding treatments.Where the
5.1 .I .2The classification of each part of a detail depends classification of table 17 does not give adequate fatigue
upon the following : resistance, the performance of weld details may be
(a) the direction of the fluctuating stress relative to the improved by post-welding treatments such as controlled
detail ; machining, grinding or peening. When this is required the
detail should be classified by tests as given in 5.2.1.
NOTE. Propagationof cracks takes place in a direction
perpendicular to the direction of stress. 5.3 Workmanship and inspection
(b) the location of possible crack initiation at the detail ; 5.3.1 General. Where the classification of a detail is dependent
(c) the geometrical arrangement and proportions of the upon particular manufacturing or inspection requirements,
detail ; which are not generally specified in Part 6 of this
I (d) the methods and standards of manufacture and British Standard, the necessary standards of workmanship and
inspection. inspection should be indicated on the relevant drawings.
5.1 .I .3 In welded details there are several locations at which All areas of the structure where welded details classified as
potential fatigue cracks may initiate; these are as follows : dass F or higher are necessary should be shown on the
(a) in the parent metal of either part joined adjacent to; drawings together with the minimum required class and an
(1) theendof theweld, arrow indicating the direction of stress fluctuation (see figure 1).
(2) a weld toe, For inspection purposes this information should be
incorporatedonto the fabricator's shop instructions.
(3) a change of direction of the weld,
(b) in the throat of the weld. Note that a joint may have more than one class requirement if it
experiences significant stress fluctuations in two or more
In the case of members or elements connected at their ends
directions.
by fillet welds or partial penetration butt welds and flanges
with shear connectors, the crack initiation may occur either NOTE. The level of manufacturingquality can affect the fatigue life of all
in the parent metals or in the weld throat : both possibilities stuctural details. The manufacturingquality determines the degree to
should be checked by taking into account the appropriate which discontinuities, that may act as stress raisers, may be introduced
during the fabrication process. Such discontinuities can act as fatigue
classification and stress range. For other details, the points, which may reduce the fatigue life to an unacceptablelevel for the
classifications given in table 17 cover crack initiation at any detail under consideration. Details with a high permitted dass are more
possible location in the detail. Notes on the potential modes seriously affected by such discontinuities because of the restrictions
of failure for each detail are given in appendix H. already placed by table 17 on stress raisers inherent in the form of the
detail itself.
5.1.2 Classification of detailsin table 17
In order to dertermine which level of quality and inspection is required in
5.1 -2.1Table 17 isdivided into three parts which accordance with Part 6, the minimum required dass has to be derived.
correspond to the three basic types into which details may If a class higher than F2 is requiredthis has to be speafied on the
be classified. These are as follows : drawings, otherwise the required fatigue life may not be achieved. If a
(a) type 1, non-welded details, table 17 (a) ; dass higher than F2 is specified, but not required, an uneconomical
fabrication would result.
(b) type 2, welded details on surface, table 17 (b) :
(c) type 3, welded details at end connections of
members, table 17 (c).
5.1.2.2 Each classified detail is illustrated and given a type
number. Table 17 also gives variousassociated criteria and
E Fat
-
6. Stress calculations
6.1 General
6.1.1 Stress range for welded details. The stress range
in a plate or element to be used for fatigue assessment is the
greatest algebraic difference between principal stresses
occurring on principal planes not more than 45" apart in
any one stress cycle.
6.1.2 Stress range for welds. The stress range in a weld
is the algebraic or vector difference between the greatest
and least vector sum of stresses in any one stress cycle.
6.1.3 Effective stress range for non- welded details
For non-welded details, where the stress range is entirely
in the compression zone, the effects of fatigue loading may
be ignored.
For non-welded details subject to stress reversals, the stress
range should be determined as in 6.1 .l. The effective stress
range to be used in the fatigue assessment should be
obtained by adding 60 % of the range from zero stress to
maximum compressive stress to that part of the range from
zero stress to maximum tensile stress.
6.1.4 Calculation of stresses
6.1.4.1 Stresses should be calculated in accordance with
I Part 1 of this British Standard using elastic theory and
taking account of a l l axial, bending and shearing stresses
occurring under the design loadingsgiven in clause 7. N O
redistribution of loads or stresses, such a s isallowed for
checking static strength at ultimate limit state or for plastic
design procedures, should be made. For stresses in
0 BSI 03-1999 4a
Issue 2,March 1999 BS 5400: Part 10: 1980
n composite beams the modulus of elasticity of the concrete table 17, the stress should be based on the net section. Where
should be derived from the short term stress/strain indicated in table 17. stress concentrations should be taken into
relationship (see Part 4). The stressesso calculated should account either by special analysis or by the fadors given in
be used with a material factor Y m = 1. figure 22 (see also H.1.2).
6.1.4.2 The bending stresses in various parts of a Steel
orthotropic bridge deck may be significantly reduced as the 6.2.2 Shear stress may be neglected where it is
result of composite action with the road surfacing. numerically less than 15 %of a coexistent direct stress.
However, thiseffect should only be taken intoaccount On 6.2.3 The peak and trough values of principal stress
the evidence of special tests or specialist advice. should be those on principal planes which are not more
6.1.5 Effects t o be included. Where appropriate, the than 45" apart. This will be achieved if either
effects of the following should be included in stress (a) oX-oyis at least double the corresponding shear
calculations : stress r a t both peak and trough, or
(a) shear lag, restrained torsion and distortion, transverse (b) the signs of oX-oy and r both reverse or both remain
stresses and flange curvature (see Parts 3 and 5 ) ; the sameat the peak and the trough,
(b) effective width of steel plates (see Part3) ; where
(c) cracking of concrete in compositeelements (see UX, byand r are the coexistent values with
applications away from joints, member eccentricities at In either (a) or (b),provided that ox22 cq2at both peak and
- joints and rigidity of joints (see Part 3). trough, the required stress range will be the algebraic
difference between the numerically greater peak principal
6.1.6 Effects to be ignored. The effects of the following stress and the numerically greater trough principal stress.
need not be included in stress calculations :
(a) residual stresses; 6.3 Stress in weld throats other than those attaching shear
(b) eccentricities necessarilyarising in a standard detail ; connecton. The reference stress for fatigue of a weld throat
(c) stress concentrations, except as required by table 17 ; should be the vector sum of the shear stresses in the weld
(d) plate buckling. metal based on an effective throat dimension as defined in
Part 3, and on the assumption that none of the load is carried in
6.2 Stress in parent m e t a l bearing b e h n parent metals.This is illustrated in figure 2b.
When calculating the stress range, the vector difference of the
I
6.2.1 The reference stress for fatigue assessment should be
the principal stress in the parent metal adjacent to the potential greatest and the least vector sum stress may be used instead
I aack location, as shown in figure 2a. Unless otherwise noted in of the algebraic difference
cc-
I 4
-">
I i ,Welded attachment
Design s t r e s s P M/z)
=(/A+
M'
- L M-M' P o t e n t i a l crack
Location
/ \
Stress
distribution
I Figure 2a. Referencestress in parent metal
[ P, e +M1
t = combined size o f
Vector sum stress e f f e c t i v e weld throats
(from Part 3)
0 BSI 03-1999 5
BS 5400:Part 10: 1980
effective throat area of weld, transverse to the shear flow, 7.2.2 Standard loading
when the concrete is of normal density and from 0.85 x
throat area when lightweight concrete is used. For the 7.2.2.1 Standardloadspectrum. The standard load spectrum
purposes of this clause the throat area should be based on a should beasshown intable 11 whichgivesthe weight
weld leg length which is the least of the dimensions intensities and relative frequencies of commercial traffic on
tabulated below. typical trunk roads in the UK. The minimum weight taken
for a commercial vehicle is 30 kN. All vehicles less than
Channel connector Bar connector 30 k N are ignored when considering fatigue.
7.2.2.2 Standardfatigue vehicle. The standard fatigue
vehicle is a device used to represent the effects of the
of bar) standard load spectrum ;for highway bridges this is a single
Half the thickness of half the thickness of vehicle with a weight of 320 kN. It consists of four standard
beam flange beam flange axles with the dimensions as shown in figures 3 and 4.
NOTE. See appendix C for the derivationof the standard fatigue
vehicle.
6.4.3.2 It may assist calculation to note that in normal
density concrete, where the thickness of the beam flange is 7.2.2.3 Number of vehicles. The numbers of commarcial
at least twice the actual weld leg length and the weld vehicles that are assumed to travel along eachlaneofa bridge
dimensions comply with Part 5, the effective weld areas are : per year should be taken from table 1. If for any reason
vehicle numbers other than these are adopted, suitable
50 x 40 bar connectors i : 200 mm long, 1697 mm
adjustments may be made to the fatigue analysis in
25 x 25 bar connectors :, 200 mrn long, 1018 rnm
accordance with 8.2.3 or 8.3.2.1 (e).
127 and 102 channel connectors 150 mm long.
1212 mm2 7.2.3 Application of loading
76 channel connectors ;< 150 mm long, 1081 mm 7.2.3.1 Demarcation oflanes. For the purposes of this Part
6.5 Axial stress in bolts. The design stress for fatigue in of this British Standard the lanes should be the actual traffic
bolts complying with the requirements of 6s 4395 and lanes marked on the carriageway. They should be
bolts to dimensional tolerances complying with the designated in accordance with figure 5 and the loading
requirements of 6 s 3692 should be calculated from the should be applied to the slow and the adjacent lanes only.
following expression : Where a crawler lane is provided i t should be treated as an
F additional slow lane.
stress in bolt =-= -/ uB 7.2.3.2 Path of vehicles. The mean centre line of travel of
uu
where all vehicles in any lane should be along a path parallel to,
and within 300 mm of, the centre line of the lane as shown
F = 1,7kN/mm 2forthreadsof nominaldiarneterupto25mm in figure 6. The transverse position of the centre line of the
or vehicle should be selected so as to cause the maximum
F = 2.1 kN/mm2forthreadsof nominaldiameterover25 mm stress range in the detail being considered. In some
b g is the stress range on the core area of the bolt
instances it may be found that the use of multiple paths
determined on the basis of the minor diameter results in significantly less calculated damage and guidance
ou is the nominal ultimate tensile strength of the bolt on this is given i n C.1.4.
material in kN/mm2
7.2.3.3 Standardloading. The passage of one standard
When subjected to fluctuating stresses, black bolts fatigue vehicle along the entire length of one lane should be
complying with the requirements ot 6s 41 90 may only be taken as one loading event.
used i f they are faced under the head and turned on shank in
accordance with the requirements of B S 41 90. 7.2.3.4 Non-standardloadspectrum. If a load spectrum is
used, which differs i n any way from the standard load
spectrum, the passage of each vehicle forming the load
spectrum should be considered to provide a separate
loading event.
6
BS 5400 : Part 10 : 1980
- 1.8 m
--
t -
6.0 m
-4
1.8 m
\
twin tyre
vehicle
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
Alternatively a 225 mn
dia. circle may be used
+tyre +tyre
7
BS5400:Part10:1980
Category of road
I Number of millions of
vehicles per lane,
per year ( n e )
Motorway
I Carriageway
layout
Dual
I Number of
lanes per
carriaaewav
3
Each
2.0
Each
1.5
Motorway Dual 2
All purpose Dual 3 1.5 1.o
All purpose Dual 2
Sfio road Sinale 2
All purpose
All purpose
Slip road
Single
Single (10 m * )
Single
1 I 1.0 Not
applicable
All purpose Single (7.3 m * ) 2 0.5 Not
applicable
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
7.2.3.5 Methodof loading. Only one vehicle should be The force assumed for any vehicle should not exceed
assumed to be on the structure at any one time and each 30 000 kN
lane should be traversed separately. The effects of r-150
combinations of vehicles are allowed for in clause 8.
7.3 Railway loading
7.2.4 Allowance for impact. Where a discontinuity 7.3.1 General. The loads to be considered should be the
occurs in the road surface, e.g. at an expansion joint, the appropriate combination of the nominal live load, impact,
static stress at every point affected by a wheel, at or within lurching and centrifugal force, as specified in Part 2 of this
5 m of the discontinuity, should be increased by British Standard.
magnifying the relevant influence line, as shown in figure 7.
In welded members the dead load stress need not be
7.2.5 Centrifugal forces. The effects of any centrifugal considered. In unwelded membersthedead load stress will
force associated with the fatigue loading defined in 7.2.2 have to be considered in determining the effective stress
need only be considered for substructures; the force should range when compression stresses occur (see 6.1.3).
be taken as acting at and parallel to the road surface. The
magnitude of the force should be calculated at the 7.3.2 Application of loading. The loads should be
appropriate design speed of the particular road, for the epplied to the appropriate lengths of the point load
individual vehicles of the standard load spectrum shown in influence lines of not more than two tracks, so as to produce
table 11 as follows : the algebraic maximum and minimum values of stress at the
WY2
detail under consideration.
the centrifugal force per axle = -
127,
(kN)
where
Wis theaxle load of thevehicle (kN)
vis the design speed of the road (km/h)
r is the radius of curvature at the particular lane on which
the vehicles are assumed to travel (m)
8
SI. Adj. F F Adj. SI.
1
SI. ' F SI.
-L -- --I-
SI. SI.
Lane marking
SI.
I
1
Adj.
1
1
F
1
Hard shoulder or Slow l a n e Adjacent lane Fast lane
c
hard strip
9
BS 5400 : Part 10 : 1980
1
t r a f f i c lane
Permitted l o c a t i o n ~A:*L3
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
4 wheels 4I
vehicle
c wheels
-Mean p o t hs
of wheels
900 900
I
I All dimensions are in millimetres
I Figure& Transverse location o f vehicles
Di scontinuit y
4 /
Adjusted stress
7.3.3 Standard/oad spectra. The load spectrum for a length. However, reference to tables 2 and 3 is not necessary
permanent railway bridge subjected to standard loading when theassessment procedure given in 9.2 is used. Where
should be taken from either table 2 for RU loading or table 3 the volumes of traftic differ from the 27 Y 1O6 tonnes per
for RL loading. These tables relate proportions of the annum, which are assumed in tables 2 and 3, or where a
standard loading Kw to the total number of applied cycles design life other than 120years is specified, the appropriate
nR x 1 0 6 occurring in a design life of 120 years and for a values of n~ may be obtained by direct proportion.
traffic volume of 27 x 106 tonnes per annum. They also NOTE. For the derivationof load spectra see appendix E.
allow for variations in the loading events with influence line
I 10
I
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
n
P.-
U
n
5
m
11
8s 5400: Part 10: 1980
Table 3. Standard load spectra for RL loading (c) determine the maximum range of stress gVmax equal
to the numerical value of opmax - OP min. For non-welded
Qroup number 1 2 3 4 5 6 details the stress range should be modified asgiven
in 6.1.3;
Lord proportion, Kw 3.55 0.45 0.35 0.25 0.15 0.05 (d) obtain the appropriate limiting stress range OH from
0.5 0.4 0.3 0.2 0.1 0 figure 8.
to to to to to to NOTE. The sign convention usedfor b pis immaterial provided it is
Range 3.6 0.5 0.4 0.3 0.2 0.1 applied consistently.Where stress reversaldoes not occur, the
rotalnumberof liveloadcycles
value of either bp or 0, mln should be taken as zero.
Length, 1 (m)
nR Y 10') forvariousloadinggroupsand 8.2.2.2 For class S details the values of OH may be adjusted
yp- by the factors given in 8.2.3, when appropriate.
2 9 120 189 42 0 0 8.2.2.3Where ov max does not exceed OH the detail may be
3 1 112 68 10 170 0 considered to have a fatigue life in excess of the specified
4 0 29 75 3 74 180 design life.
5 0 6 110 0 2 75
7 0 38 65 0 0 77 8.2.2.4 Where oVmax exceeds OH either of the following
10 1 10 56 37 0 77 options may be adopted.
15 1 13 0 49 30 15 (a) The detail may be assessed by the alternative
20 1 13 0 0 50 80 procedure given in 8.3if it is not a class S detail, or by the
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
60
50
40
30
20
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
10
1 2 5 10 20 50 100 200
80
70
.=Ov)
ul
z2
N
E
60
50
r
B
\
-
z
bs 40 C
d,
CI,
C
30
U)
U)
D
cE E
ul S
0,
.E 20
F
.-
c
F2
.-E
-I
G
W
L lm)
(c) three lane all purpose, two lane all purpose (1 0 m),
two lane slip road
13
BS 5400 Part 10: 1980
assessed .aeach peak and each trough in the stress 8.3.2.2 Where the predicted fatigue life of the detail is less
history of each lane in turn (see figure 9). than the specified design life, the detail should either be
NOTE. It issufficientlyaccuratetocalculateeachpeakor strengthened to reduce the value of uv max or redesigned t o
trough value of the direct stress and to obtain the principal stress a higher class and then re-checked as in 8.3.2.1.
by combining these with the coincidentshear stress, or vice- As a guide, an approximate stress range for the same class
versa where this is more severe.
of detail can be obtained by multiplying the original value
(c) When the maximum and the minimum algebraic by :
values of stress op max, op min. result from vehicle
(predicted life) "(m+O
positionsin thesamelane (referred toascase 1 in figure
9) the damage should be calculated for the stress design life
historiesfor each lane separately. where
When c ~ marp and opmin result from vehicle positions in m is the inverse slope of the appropriate log ar/log N
different lanes (referred to as case 2 in figure 9) an curve given in table 8.
additional combined stress history should be derived, If the detail is to be redesigned to a higher class the
which allows for the increased maximum stress range procedure given in 11.5( b) may be used as a guide to
produced by a proportion of the vehicles travelling in assess the adequacy of the proposed detail.
alternating sequence in the two lanes. Inthis case the 8.4 Damage calculation, vehicle spectrum m e t h o d
damage should be calculated for the combined stress
8.4.1 General. This method involves an explicit calculation
history as well as for the separate lane stress histories.
of Miner's summation and may be used for any detail for
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
(d) Derive the stress spectrum ovl, uv2 etc.. from each which the a,-Nrelationship is known and for any known
stress history determined from (c). load or stress spectrum.
Where a stress history contains only one peak and/or only 8.4.2 Design spectrum
one trough, only one cycle results, as shown in figure 9
8.4.2.1 The individual stress spectra for the detail being
for lanes C and D, and the stress range can be determined
assessed should be derived by traversing each vehicle i n the
directly.
load spectrum along the various lanes. Account should also
Where a stress history contains two or more peaks and/or be taken of the possibility of higher stressrangesdue to
two or more troughs, as shown in figure 9 for lanes A and some of the vehicles occurring simultaneously in one or
8 , more than one cycle results and the individual stress more lanes and/or in alternating sequence in two lanes.
ranges should be determined by the reservoir method For non-welded details the stress range should be modified
given in appendix B. as given in 6.1.3.
(e) Determine the effective annual flow of commercial 8.4.2.2 In the absence of other evidence, allowance for
vehicles, Tic million, appropriate t o each stress spectrum impact should be made in accordance with 7.2.4. The
as follows : design spectrum should then be determined by combining
(1) where case 1 of figure 9 applies, & = nc and may be the stress spectra with the specified numbers of vehicles in
derived directly from table 1 unless different vehicle the respective lanes.
flows are adopted ; 8.4.2.3 In assessing an existing structure, a design spectrum
(2) where case 2 of figure 9 applies the effective annual may be compiled from strain readings or traffic records
flow fiC should be obtained as indicated in figure 9 for obtained from continuous monitoring.
case 2. 8.4.3 Simplification of design spectrum. The design
(f) For each stress range ovof each stress spectrum, spectrum may be divided into any convenient number of
determine the appropriate lifetime damage factor d, 2o intervals, as shown in figure 13, with all the stress ranges in
from the damage chart of figure 10 and multiply each of any one interval being treated as the maximum range in that
these values by the appropriate value of ifc. For non- interval but low stress ranges should be treated in
welded details the stress range should be modified as accordance with 11.3. It should be noted that the use of
in 6.1.3. small intervals will reduce the conservatism in fatigue
(g) Determine the value of the adjustment factor K F from assessment.
figure 11 according to the base length L of the point load 8.4.4 Calculation of d8mege. Using the design spectrum,
influence line (see figure 12) and the stress range ratio
KB defined in figure 11. the value of Miner'ssummation - I
= should be calculated in
For a combined stress history from two lanes (see (c) accordance with clause 11. This value should not exceed
above and case 2 in figure 9) KB should be taken as zero 1.Ofor the fatigue life of the detail to be acceptable.
for determining K,.
NOTE. For the derivation of KF see appendix C.
(h) Determine the predicted fatigue life of the detail from
the following. expression
. :
120
fatigue life (in years) =
ZKFEcd, 20
where the summation includes all the separate lane stress
histories as well as the combined stress history, where
appropriate.
14
BS 5400 : Part 10: 1980
c Lane reference* I
I
I - 1 A .
Lane A
Lane B
Lane C
I -p
1c
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
Lane 0
1
Case 1. Highest peak and lowest trough with vehicles in same lane
Lane reference' one n r e u hlnory Lane straso dumber of cyclea ffoctlv. Imne
apmctrat )er loading event low. iic
Lane A
Lane B
Lane C 1c k c
Lane D
I
Pv*rre
Key
0 Peak stress Stress range Q,
0 Trough stress for cycle shown
X Datum stress
16
in
U
.-VI
m
al
5
I
em
BS 5400 : Part 10 :.1980
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
NOTE 1. L is the base length of the point load influence line (see figure 12).
K, is the ratio -
oVl B
UVlA
where
QVlA is the largest stress range produced by loading in any one lane
ov1B is the nexl largest stress range produced by loading in any other lane ( ~ V BC < oV1A).
For non-welded details the stress range should be modified as given in 6.1.3.
( K e may be taken as zero for the combined history of figure 9, case 2).
NOTE 2. This figure is applicable only to detail classes B to G, F2 and W.
c
- -Largest ordinate ( t v e or -vel
NOTE. L is the base length of loop containing the largest ordinate measured in direction of travel. For an element of a highway
bridge loaded by more than one lane, L should be determined from the influence line for the lane producing the largest value of
b v i ( = O V l A ) (seefigure9).
17
BS 5400 : Part 0 :1980
Stress
ranges
10;1
__---
1
I ,Simplified design spectrum
I
Spectrum as calculated
or recorded
_--__---
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
Number of repetitions ( n )
Figure 13. Slmpliflcstlon of e spectrum
18
BS 5400: Part 10: 1980
< 3.4 1.00 1.00 1.01 1.14 1.09 1.09 1.13 1.28 1.37 1.60 1.60 1.71
3.4 to 4.0 1.09 1.09 1.13 1.28 1.23 1.22 1.30 1.46 1.53 1.79 1.80 1.71
4.0 to 4.6 1.23 1.22 1.30 .46 1.37 1.36 1.46 1.46 1.71 1.79 1.80 1.71
4.6 to 7.0 1.37 1.36 1.46 .65 1.53 1.56 1.62 1.65 1.92 2.05 2.00 1.95
7.0 to 10.0 1.53 1.56 1.62 -65 1.71 1.75 1.81 1.83 2.19 2.31 2.24 2.20
10.0 to 14.0 1.71 1.75 1.62 .65 1.92 1.95 2.03 1.83 2.46 2.31 2.50 2.20
14.0 to 28.0 1.92 1.95 2.03 .83 2.19 2.18 2.03 1.83 2.74 2.56 2.50 2.20
> 28.0 2.19 1.95 2.03 .83 2.46 2.18 2.03 1.83 3.06 2.87 2.50 2.20
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
1 I
- NOTE. I! is the base length of the point load influence line (see figure 12).
0. 5 to 0.6 0.6 to 0.75 0.75 to 0.9 0.9 to 1.0 0.0 to 0.7 0.7 to 1.0
Length, L (m) V ~ I J OO
S f ha
9.2.4 Multiple cycles. Where the loading event produces respective stress spectra. These should then be combined
more than one cycle of stress the value of k2 should be with the appropriate total occurrences in the design life of
taken as : the bridge to compile the overall design spectrum. For
m non-welded details the stress range should be modified as
( OR2
I+(--) +(-)
OR1
OR3
OR I
+.... given in 6.13.
9.3.3.2 In assessing an existing structure, a design spectrum
where may be compiled from strain readings or traff ic records
rn is defined in 9.2.3 obtained from continuous monitoring.
UR I , ORZ. OR 3 etc. are the stress ranges, in descending 9.3.4 Simplification of spectrum. Where a non-
order of magnitude, at the individual cycles produced by standard loading is used in accordance with 7.1, or the
the approach, passage and departure of a unit uniformly stress ranges are obtained from strain gauge readings, the
distributed load. design spectrum should be divided into at least 10 equal
NOTE. Such cycles should be counted and the individual strew intervals of stress. All the stress ranges in any one interval
rangesdetermined by the reservoir method given in appendix 8. should be treated as the mean range in that interval and low
An illustrationof the multiple cycle stress history is given in stress ranges should be treated in accordance with 11.3.
example 4 of appendix F.
9.3.5 Calculation of damage. Using the design spectrum,
9.3 Damage calculation
the value ot Miner's s u r n m a t i o n Z i should be calculated
9.3.1 General, This method involves a calculation of
Miner's summation and may be used for any detail for which in accordance with clause 11 and should not exceed 1.O
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
the or -N relationship is known and for any known load or for the fatigue life of the detail to be acceptable.
stress spectra. It may also be used as a more precise
alternative to the simplified procedure of 9.2.
9.3.2 Design spectrum for standard loading 10. Fatigue assessment of b r i d g e s carrying
9.3.2.1 Applying the standard railway loading as given highway and r a i l w a y loading
in 7.3.1 and 7.3.2 the value of OR max should be derived in In the case of bridges carrying both highway and railway
accordance with the procedure set out in 9.2.2.1 (a) to (c). loadings, the total damage (i.e. 120years divided by the
The design spectrum should then be determined by the use predicted life) should bedetermined for each loading
of either table 2 for RU loading or table 3 for RL loading condition separately, in accordance with 8.3 or 8.4 and 9, I.
(amended where appropriate in accordance with 7.3.3). To obtain the total damage, the sum of the two damage
These tables indicate, for simply supported members, the values should be multiplied by a further adjustment factor
equivalent frequency of occurrence of stress ranges of which takes into account the probability of coexistence of
varying magnitudes resulting from the passage of the the t w o types of loading. This factor should be determined
individual trains forming various standard traffic types, for a given member after consideration of the fact that
where the stress ranges are expressed as proportions of the coincidence of highway traffic on multiple lanes and of
maximum stress range. railway traffic on multiple tracks has already been taken
9.3.2.2 In the case of loading from more than one track, into account in assessing the separate damage values.
account should be taken of the possibility of stress Except at very busy.railway stations, where the probability
fluctuations arising from the passage of trains on not more of coincidence of rail and road traffic is higher than on the
than two tracks, both separately and in combination. As an open track, the adjustment factor is not expected to exceed
approximation, the effects of two track loading may be 1.2 where the stressesfrom highway and railway loading
obtained by dividing OR max (see 9.3.2.1) by the are of the same sign.
coefficientkS which can beobtained from table 6.
9.3.2.3 Where the approach, passage and departure of a
unit uniformly distributed load produces more than one
cycle of stress, as for instance in multi-span longitudinal or 11. The P a l m g r e n - M i n e r r u l e
cross members or in continuous deck slabs, all the cycles
should be taken into account. The appropriate standard
trains of figure 19 or figure 20 should be traversed across
11.1 General. Thevalue of Miner'ssummation xi for use
the relevant point load influence lines and the resulting in 8.4.4 and 9.3.5 should be determined from the following
stress histories should be analysed by the reservoir method, expression :
given in appendix B, to derive the respective stress spectra.
These should then be combined with the appropriate
fl2 . . . . . . 2)
Nn
annual occurrences obtained from table 15 or table 16
proportioned for the required traffic volume and multiplied where
by the specified design life to produce the overall design n , , n 2 . ..n, are the specified numbers of repetitions of the
spectrum. As an approximation, the effect of the additional various stress ranges in the design spectrum, which occur
cycles may be obtained by dividing either UR max in the design life of the structure.
(see 9.3.2.1 ) or OR max/ks(see 9.3.2.2) by the coefficientkZ NOTE. The number of repetitionsmay be modified in
which should be obtained from 9.2.4. accordancewith 11.3, and for non-weldeddetails the stress
rangeshould be modified as given in 6.1.3.
9.3.3 Design spectrum for non-standard loading
N I . N2.. .N,, are the corresponding numbers of
9.3.3.1 Where the loading does not comply with 7.3.1
repetitions to failure for the same stress ranges, obtained
the appropriate train should be traversed across the relevant
from 11.2.
point load influence lines and the resulting stress histories
should be analysed by the rainflow method to derive the
*The rainflow method is described in ORE D128 Report No. 5 'Bending moment spectra and predictedlives of railway bridges'.
publishedby the Office for Researchand Experimentsof the InternationalUnion of Railways. The reservoir method of cyclecounting,
described in appendix 6 for highway bridges. may be appliedtostress histories for railway bridges (see example 4 of appendix F) and
will produce the same results as the rainflow methodfor many repetitionsof the loading event.
20
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
21
BS 5400 : Part 10 : 1980
B S 5 4 0 0 : P a r t l O : 1980
11.2 Design ur -N relationship. The number of repetitions 11.S Miner'ssummation greater than unity. If the
to failure Nof any one stress range orshould be obtained ,conditions in 8.4.4 and 9.3.5 for highwav and railway
from either of the following equations, which have been
plotted in figure 16 :
bridges respectively are not met, i.e. if 12
N
> 1.O, the
N x U': = K, following alternative actions should be considered.
Either
-
Log,, N = Log10 Kn m Logqour
(a) strengthen the detail to reduce the values of Or. The
where
strengtheneddetail should be satisfactory if the reduced
K2 and m have the values given in table 8 for the different values of stresses lie between the limits obtained by
detail classes. dividing the original values by the following factors :
NOTE. The valuesof K2 correspond to a probablityof failure of
2.3%withinthe design life. The basic equations and a mean-line
plot i.e. for 50 %probabilityof failure, are given in appendix A.
( E;) 'lmand
It may assist in calculations to note that : upgrading to any class up to 0, the value of uofor the new
class should be between
22
BS 5400: Part 10: 1980
- Appendix A
Basis of a,-N relationship
Table 10. Probability factors
Id
::: ;i
Probmbllity of
trlluro
A.l General. The or-N relationships have been
established from statistical analyses of available
experimental data (using linear regression analysis of log or
and log N) with minor empirical adjustments to ensure
compatability of results between the various classes.
The equation given in 11.2 may be written in basic form as : 2.3 % 2.0t
N x U?' = K O x Ad 0.14%
where *Mean-line curve.
tThe standard design curve of 11.2.
N is the predicted number of cycles to failure of a stress
range ur A.2Treatment of low stress cycles. Under fluctuating
KO is the constant term relating to the mean-line of the stress of constant amplitude, there is a certain stress range
statistical analysis results below which an indefinitely large number of cycles can be
sustained. The value of this'non-propagating stress range'
m is the inverse slope of the mean-line log Or -log N curve varies both with the environment and with the size of any
A is the reciprocal of the anti-log of the standard initial defect in the stressed material. In clean air, a steel
deviation of log N detail which complies with the requirements of Parts 6,7 or
d i s the number of standard deviations below the 8 is considered to have a constant amplitude non-
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
: ;1
Detril clard KO Id m
the defect. The final result is an earlier fatigue failure than
W 10.37 x 1012 10.654 13.0 could be predicted by assuming that all stress ranges below
10.57 x 10" 10.862 uoare ineffective.
1.23 x 10l2 0.592 This phenomenon has been studied on principles derived
from fracture mechanics. It is found that an adequate
;2 1.73 x 101 2 0.605 3.0 approximation to the fatigue performance so predicted can
E 13.29 x 10' I
0.561 13.0 be obtained by assuming that a certain fraction (flr/Oo) of
D 13.99 x 1012 10.617 13.0 stress ranges Q, less than oocause damage in accordance
withtheformula inA.1.
C Il.08 x 1014 10.625 13.5
B 12.34 x 10' I 0.657 14.0
S 12.13 x lOZ3 10.313 18.0
Endurance N (cycles)
NOTE. The use of these curvesfor calculation purposes is not recommended.
Figurel5. Summary of mean-linea,-Ncurves
23
BS 5400:Part 10: 1980
standard deviations below the mean-line with a probability 1.6 1.0 0.6 0.0
of failure of 2.3 %. In certain cases, a higher probability of W 1.07 1.15 1.24 1.32
failure could be acceptable, for example, where fatigue
G 1.07 1.15 1.23 1.32
cracking would not have serious consequences, or where a
crack could be easily locatedand repaired. F2 1.09 1.19 1.30 1.42
The probabilitiesof failure associated with various numbers F 1.09 1.18 1.29 1.40
of standard deviations below the mean-line are given in E 1.10 1.21 1.34 1.47
table 10. The Qr -Ncurves appropriateto other numbers of D 1.08 1.17 1.27 1.38
standard deviations below the mean-line can be derived
from theformula given in A.l. C 1.07 1.14 1.22 1.31
Where the methodsof clauses 8.9 or 11, which are based B 1.05 1.11 1.17 1.23
S
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
on two standard deviations below the mean-line, predict 1.07 1.16 1.24 1.34
the fatigue life (or damage), the life (or damage) appropriate
to other numbers of standard deviations below the mean-
line can be obtained by multiplying the calculated life (or
dividing the calculateddamage) by the following factors :
Number of rtandard dOvlatlon8
- - --
bSlow1 a moan ne
-1.5
-1.o
-0.1
-
0.0
itatic limitations
----- ----_-_
Effective curve obtained under
I variable amplitude loading,
I \
I \
\
I \
\
I
I \
\
I
I
\
\
\
m+2 -.\
\
I \
1o7
Endurance N (cycles) - log scale
NOTE. Only that portion of this figure shown as a full line is based on experimental evidence.
Figure 16. Typical U,-N relationship
24
BS 5400: Part 10: 1980
Appendix B Appendix C
Cycle counting by the reservoir method Derivation of standard highway bridge fatigue;
loading and methods of use
6.1 General. The purpose of cycle counting is to reduce an
irregular series of stress fluctuations to a simple list of stress C.l Standard loading
ranges. The method given in this appendix, and shown in the C.l.l ~tenderdloedspectrum.(See7.2.2.1.) Table 11
figure below, is suitable when dealing with short stress shows a 25 band spectrum of commercial vehicle weights,
histories, such as those produced by individual loading axle arrangements and frequencies of occurrence, which is
events. It consists of imagining a plot of the graph of each typical of the full range of commercial traffic on a trunk road
individual stress history as a crosssection of a reservoir, in the UK. Other relatively uncommon vehicle types have
which is successively drained from each l o w point, counting been included in the types which are nearest to them on the
one cycle for each draining operation. The result, after many basis of equivalent damage. Private cars and light vans
repetitions of the loading event, will be the same as that below 15 kN unladen weight are not included as their
obtainable by the rainflow method (see the footnote contribution to fatigue damage is negligible. Table 11
to9.3.3). includes vehicles operating under both the Motor Vehicles
6.2 M e t h o d (Construction and Use) Regulations and the Motor
8.2.1 Derive the peak and trough values of the stress Vehicle (Authorization of Special Types) General Order.
history, due to one loading event, in accordance The proportions of the various types of vehicles of the
with 8.3.2.1 (c). Sketch the history due to twosuccessive spectrum have been taken from sample traffic counts. To
occurrences of this loading event. The calculated values of allow for variations in the loads being carried by similar
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
peak and trough stresses may be joined with straight lines vehicles, the various types have been divided into heavy,
if desired. Mark the highest peak of stress in each medium and light loading groups (H, M and L). The axle
occurrence. If there are two or more equal highest peaks in loads have been averaged from weighbridge records ot
one history, mark only tho first such peak in each moving traffic taken between 1971 and 1974.
occurrence.
c.1.2. Standerdfetigue vehicle. (See 7.2.2.2.) The
8.2.2 Join the two marked points and consider only that proportions of the damage caused by individual vehicle
part of the plot which falls belowthis line, like the section types, compared with the total damage by all vehicles,
of a full reservoir. varies between the limits shown in table 12. The standard
8.2.3 Drain the reservoir from the lowest point leaving the fatigue vehicle has been devised to represent the most
water that cannot escape. If there are two or more equal damaging group which, for the majority of detail classes and
lowest points the drainage may be from any one of them. influence line lengths, isgroup 4A-H.
List one cycle having a stress range uvrequal to the vertical The axle spacings of the standard fatigue vehicle are the
height of water drained. same as thoseforthe short HB vehicle (see Part 2) and the
8.2.4 Repeat 8.2.3successively with each remaining body 80 kN axle weight is equivalent to the standard 18 000 Ib
of water until the whole reservoir is emptied, listing one axle, which has been used for some years as the datum axle
cycle at each draining operation. in the fatigue design of road pavements. As the damage
8.2.5 Compile the final list which containsall the done by the single tyred wheels on vehicles listed in table 11
individual stress ranges in descending order of magnitude is normally less than 4 % of the damage done by the double
G,,, ovt etc. Where two or more cycles of equal stress range tyred wheels, the standard fatigue vehicle with double tyred
are recorded, list them separately. wheels gives an adequate representation of the wheel
8.2.6 For non-welded detailsonly, a horizontal line damage for all types of vehicles.
representing zero stress should be plotted and those parts of
thestressrangesin thecompressionzonemodified asin 6.1.3
25
BS 5400 : Part 10 : 1980
C.1.3Stenderdlene flows. (See 7.2.2.3.)Theannual C.3.2 Demege fectofs d , 20. (See figure 1 0 and 8.3.) The
flows of commercial vehicles nc x 108 given in table 1 are damage chart ot figure 10is based on the cumulative fatigue
based on the design capacity of the particular road type, as damage caused by the design spectrum which is obtained
specified by the Department of Transport. The proportion of from the passage of the vehicles represented by the gross
commercial traffic above 15 kN unladen weight has been weight spectrum of table 13 over an influence line base
taken to be 20 %of all traffic for all-purpose roads and 25 % length of 25 m. It assumes 106 cycles of stress per year for
for motorways. the 120year design life and the damage, assessed by
C.1.4 Multipleperhs. (See 7.2.3.2.)A significant Miner's summation, is given in relation to the stress range
reduction in assessed damage can be achieved by the caused by the passage of a standard fatigue vehicle.
consideration of multiple paths when the transverse The adjustment factor K F should be applied to the results
influence line profile departs rapidly from thevalue of the from figure 10in order to allow for influence line base
mean path ordinate. Figure 17 shows a histogram of lengths of less than 25 m and for the effects of multiple
occurrence of proportion factors for 100 mm wide intervals vehicles. The effects of numbers of vehicles other than the
of carriageway width, derived from observations of traftic 120 million assumed are allowed for by multiplying the
patterns. lifetime damage from figure 1 0 by nc.
The standard fatigue vehicle centre line should be traversed C.4 A d j u s t m e n t f a c t o r s
along the centre line of each 100 mm strip and a stress C.4.1 Genere/. Because the assessment charts of figures 8
spectrum obtained for each strip in accordance with 8.3.2, and 10are based on the passage of single vehicles and also
with the annual flow of vehicles in any strip being derived on an assumed influence line base length of 25 m,
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
from the appropriate proportion (from figure 17) xnc x 106 adjustment factors are necessary to allow for shorter base
(taken from table 1). In cases where the transverse influence lengths and for the effects of combinations of vehicles.
line changessign across the 1300 mm histogram width, These factors, which are labelled X, Y and 2 in figure 18,
an alternating succession of vehicles along paths of make up the Miner's summation adjustment factor KF
opposing sign would produce a reduced number of cycles of given in figure 1 1. Factor X is described in C.4.2 and
enhanced stress range. However, trial calculations, which factors Y and 2 in C.4.3. It should be noted that the values
take into account the probability of the occurrence of of these factors cannot be determined precisely but that the
alternating sequences, have shown that the increase in values of KF given are sufficiently accurate for design
damage is not significant and may be neglected. purposes.
C.2 Derivation of load spectra based o n t h e standard The adjustment factorKF has been included in the
f a t i g u e vehicle. The principle behind the assessment derivation of figure 8 but should be applied explicitly when
procedures of 8.2 and 8.3 is that each commercial vehicle figure 10 i s used. For the assessment method of 8.2 there is
in table 11 is represented by one vehicle of the same gross an additional adjustment, which may be made in the case
weight but with axle configurations identical to those of the of class S details, where the design is based on reduced H B
standard fatigue vehicle. The resulting load spectrum shown loading and this is described in Ci4.4.2.
in table 13, where the various vehicle group weights are C.4.2Influence line bese length less then 25 m. The X
expressed as a proportion of the standard fatigue vehicle
component of the KF adjustment factor has been obtained
gross weight, has been used in the derivation of the
by comparing the stress histories for a selection of the most
procedures of 8.2 and 8.3 instead of the complete vehicle damaging vehicles in table 11 with those derived by
spectrum of table 11. Table 13 has been derived using a
representing the vehicles by standard fatigue vehicles with
datum influence line which is 25 m long and rectangular in
the same gross weights. The two groups of vehicles were
shape. This has the effect of limiting the gross weight of the traversed across 11 different shapes of influence line, each
18 GT group so that static design stresses are not exceeded.
with a range of loop lengths, and the Miner's summation for
For values of L less than 25 m the design spectrum becomes damage calculated using the Or -Nrelationships for all the
increasingly influenced by bogie and axle spacings and the detail classes except S.The X factor is the average value
weights. The weighbridge records, from which table 11 is of theratio of damage due to the table 11 vehicles to that
derived (see C.1.1).show a wider variation in axle loads due to the equivalent standard fatigue vehicles. The scatter
than in gross vehicle loads, and an increase of 10 % in results between different influence line shapes was found
(indicated by trial calculations) over the individual values to be acceptable for design purposes.
of table 11 has been allowed in deriving the spectrum of
C.4.3 Multiple vehicle effects. During normal conditions
axle weights of table 14. When L is greater than 25 m, the
of traff ic flow, instances will occur when more than one
design spectrum for individual vehicles will be proportional
vehicle will contribute to the stress in a detail at any
to the gross vehicle weight spectrum. However, as L
particular time and the stress may be increased above that
increases, account should be taken of the contribution to
due to either vehicle alone. These multiple vehicle effects
damage done by two or more vehicles acting simultaneously.
may besub-divided according to the following :
All these effects have been catered for by means of a simple
adjustment factor KF (see figure 11) which is described in (a) more than one vehicle in the same lane
more detail in C.4. simultaneously ;
C.3 Assessment c h a r t s (b) vehicles in different lanes simultaneously causing
stress of the same sign ;
C.3.1 Limiting stress ranges OH. (See figure 8 and 8.2.)
The graphs of allowable stress ranges in figure 8, which are (c) vehicles in different lanes in alternating sequence
based on a 120 year design life and the appropriate traffic causing stresses of opposite signs and so increasing the
flowsfrom table 1, have been derived with theaid of the stress range.
damage chart of figure 1 0 (see C.3.2).The worst The chance that damaging vehicles will be sufficiently close
assumptions have been made about the stress to each other, either in the same or different lanes, has been
contributions from each lane and about the allocation of assessed on a probability basis. Correction factors have been
traffic flows between lanes so that the results are always on derived from the comparison between the damage from the
the safe side. The adjustment factor K F , from figure 11, has combinations of vehicles and the damage due to such
also been included in the derivation of figure 8. vehicles on their own.
26
BS 5400 : Part 10 : 1980
? 5% 5K %K
U
m
- 5544 m m m
K
0 0 0 3 0 0 0 0 3 0 0
0
03
N
3 88
D O 0
0 0 0 3 0
0 0 0 3 0
0
0
0
0 0 0 3 0 0
3 0 0
3 0 0 m m m 3 0 0 m m m ?X8
nmQ) 3 m m
-
e r - - l - c
r - l -
T
1-F
0 0 0 0 0 0 c)mm
0 0 -03 dQ)
N
m m w m m
(Do N c r
r c
0
0 0 oal 0 0
0 0 -03 d
N
m m
m m N e c
0 0 0
C I C -00 m
N l-
n m m
0 0
c c 0 0 0 0 0 0 w m m
- 0 3 m - d N o m o
E2 N r c o m e
n o m
b o ~
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
0 0 0 0 0 0 0
bco -03 # d N
N N
0 0 0 0
dQ)
7 0 3 N
N
0 0
bQ)
h C
N 0 0
0 0 0 m
c c
s- z- F c
- -
0 0
22 e m 0
r e 0
0
m
0 0
dco 0 0 0 0
N
0 0
2: 2:
zz 0
I-
0
I-
0 0
coo 0 0 0 0
F - (zm (zco
--
0 0
CQCO
i
- E
mw .- 0 0 0 0 n o o o m o n m o
3 w m
L me4 Q) m r e m bar4
c" i s corn
m -
l- co N - N r l - r
W
.E n
c
mo
1 1 1 1 1 I x z - I rz-l
rr!
*
9
g,
(v
- Pm p
.-P
(P
0 ID
0
A
c" .-0
a
d m (v
27
BS 5400 : Part 10 : 1980
The effect given in (a) above is allowed for by means of C.4.4 Class S details
factor Y (see figure 18) which shows that the effect is C.4.4.1 General. Trial calculations indicate that for class S
negligible for L less than 50 m but increases with details, significant variations occur in the values of X
increasingL due to the greater probability of having two (see C.4.2) both with variations in L and with variations
vehicles simultaneously on the same influence line base in influencelineshapes. Hence the adjustment factor X does
length. not apply to class S details and the assessment procedure
The effect given in (b) above is allowed for by means of of 8.3 cannot be applied.
factor 2 (see figure 18),which isgiven for severalvalues of C.4.4.2 ReducedHB designloeds. (See8.2.3 (c).) Table
factor K g (see figure 11), i.e. the ratio between the major 12 shows that the heaviest abnormal load vehicles of table
stress ranges produced by vehicles travelling separately in 11 contribute a very small percentageof the total damage for
the two lanes producing the most severe stress effect. all detail classes except class S. Hence no relaxation is
Combinations of vehicles in more than two lanesdo not provided in the assessment proceduresof 8.2 and 8.3for
generally increase the total damage significantly. these other classes when the bridge is subject to reduced
The effect given in (c) above is taken account of by the values of HB loading.
additional combined stress history, referredto as case 2 in However, these heavy vehicles do contribute the greater
figure 9. Inthis case the effect of vehiclestravelling proportion of the total damage for class S details and
simultaneously in two lanes may be neglectedand so factor hence 8.2.3 (c) provides reduction factors where the bridge
Z can be taken as zero by making K s equal to zero. is designed to carry less than 45 units of HB loading.
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
I'
Table 12. Proportional damage from individual groups of typical commerclal vehicles
-
~~
-
-- -
D to a, F i s n d W C % i
Dstril clrw
Group.
'm 3.0) m
! 1.5
3.5)
- -
1-25
m
i
4.0)
1.5 L - 25
m
-1.5
8.0)
I L -25
18 GT-H 1% 5%
9 lT-H 1% 4%
18 GT-M and
9lT-M to 5A- H 3% 6% 13%
5A-M and L 14% 4% 14%
4A-H, M and L 57 % 63 % 57% 54% 0.057 %
4 R - H to 2 R-L 34 % 8% 33% 6% 0.003 %
'Seetablell.
of mean p a t h of vehicles
(see figure 6 1
% frequency
distribution of
vehicles for each
100 mrn interval - 1
I I "I
- 600 rnm
ce
600 mm
c
c
c
1 0
L
Figure 18. Typical Miner's summation adjustmont C U ~ V O
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
I
Total 1.0
29
Appendix D a vertical stiffener, given that analysis, in accordance
~
with 8.2.2.1 (a) and (b), produces principal maximumand
Examples of fatigue assessment o f highway minimum stressvalues of 17.8 Nlrnmz tensile and
bridges by simplified methods 8.0 N/mm compressiverespectively,under the standard
D.l General. D.2 to D.4 give examples of typical fatigue vehicle, with the vehicle positioned in thesame lane.
calculations for fatigue assessment. D.2.3 Classification. The potential fatigue crack should
For all detail classes, 8.2 and figure 8 providea limiting be classified as F (from type 2.9 of table 17 (b)), provided
stress range QH which will always be safe where standard that the weld end is not within 10 mm of the flange toe.
loading conditions are applicable but which may be too 0.2.4 Assessment using 8.2
conservative in some cases. The U H values given in figure 8 From D.2.2, op may = 17.8 N/mm2
thus provide a simple check which is very suitable for initial
design purposes. UP min = -8.0 N/mm2
17.8
- 8.0
1st carriageway, slow lane 1st carriageway, adjacent lane
d 7 -2.8
17.8
0 0
- 8.0
The values of uv for the other lanes may be determined in a
similar way.
30
BS 5400: Part 10: 1980
,=:
I I
I
I 1 IlS5
2nd Adj. C 1.o
2 0 Neglect -
( < 0.001 )
2nd Slow
,"'" 1-1.5 7.8 0.001
t
:01
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
*See appendix B.
tSee figure 10.
#See table 1 and figure 9.
Adjustment factor K, (see figure 11) D.2.6 Comments. The method of 8.3 predicts a fatigue life
KB = 15.5/25.8 = 0.6 that is in excess of the design life and hence a reduction in
1=75m cross section could be tolerated. By reference to 8.3.2.2,
hence KF = 1.59 -
the stress range could be increased by
.
(142
\
\ o'2si.e. by 1.04.
ILV/
Total damage = Z K F ECd , 2o = 1.59 x 0.53 = 0.843
Estimatedlife= 2010.843 =i 4zyears(see 8.3.2.1 (h)) In contrast, the procedure of 8.2 produces a limiting value
which is greater than the specified design life of 120years Of uH equal to 20*5/25.8 = 0.79 times the applied
and so the detail may be regarded as satisfactory. maximum stress range, thus demonstrating the
conservatism of this method. Nevertheless, the simplicity of
determination of OH is such that the method should, in many
cases, prove a useful 'first stage'as an alternative to the
more precise method of 8.3, or the lengthier procedure
of 8.4.
31
0.3 Example of the application of a combined stress provide adequate fatigue resistance at the lap-welded
history bracinglgusset connection.
D.3.1 Typeofbridge. A highway bridge that carriesa two D.3.3 Cl8ssification. The potential fatigue crack should
lane, single carriageway, all-purpose road and is designed beclassfiedasG (fromtype2.11 oftable17(b)) sincethe
to carry standard UK loading, but is deemed to be subject to ,weld will beattheedgeofthemember (seealsofigurel).
1.2 x 106 commercialvehicles per year, in each lane, for a D.3.4Asssssm8nt using8.2. The information given
design life of 60 years. in D.3.1 and D.3.2 does not comply with (b) and (d)
0.3.2 Details of theproblem. Given that the analysis of 8.2.1 and therefore 8.2 is not applicable.
predicts, as shown in D.3.5, the forces in a transverse D.3.5 Assessment using8.3. Assume a cross-sectional
bracing member due to the standard fatigue vehicle in area of 2200 mmz and determine the stress histories for
alternate lanes and given that the influence line base length passage of the standard fatigue vehicle in each lane.
is 8 m, determine an area of cross section for the member, to
Given that these are as follows :
12.9 12.9
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
Lane A
- 12.9 -12.9
Lane B
12.9 12.9
- 12.9 -1 2.9
Lane A/B
32
n The stress ranges uvmay be determined by the reservoir
method (see appendix B) and the resulting stress spectra
(see figure 9) for the individual and combined lane stress
histories will be as shown below :
Lane A 54
Z'cycles per
loading event
0
Lane B
"Pt
0
2 cycles per
loading event
4
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
Combined A/B
,,A,B= 25.8
B
1
2
0
-4.2
-1 2.9
-12.9
12.9
8.7
0.055
0.01
} ;. 1 2:1 .2
= 0.6
0.03
0.01
0.08
I I
;.
1 :zl .2
= 0.6
0.84
0.01
0.01
Z& dc ao 0.86
*See appendix E.
tSee figure 10.
$See figure 9.
L=8m and hence an acceptable area will be 201 7 mm2. Repeat the
henceKF = 1.81 assessment with an amended initial assumption of the area.
for a combined history NOTE. Iterationprovides a solution of 2027 mm2.
KB = 0
L =8m
hence KF = 1.47
Total damage = Z K d , 2o
LI = 1.81 x 0.08 -7 1.47 x 0.86
= 1.41
Estimated life = 12011.41 = 85 years (see 8.3.2.( h ) )
which is greater than the specified design life of 60 years
and so the detail may be regarded as satisfactory.
33
0.4 Exampleof a typical procedure for shear Appendix E
connectors
D e r i v a t i o n of standard railway load spectra
D.4.1 Type of bridge. A 30 m simply supported composite
highway bridgethat carries a dual carriageway all-purpose E.l RU loading. The load spectra given in table 2 have
road and is subject to standard UK loading, but which is been based on the typical trains shown in figure 19. The
limited to 37.5 units of HB loading. numbers of these trains, assumed for the three broad traffic
D.4.2 Detailsof theproblem. To investigatethe fatigue types, are shown in table 15 together with the make-up of
capacity of the attached shear connectors (which are in the total annual tonnages. These spectra will cover most
accordancewith Part 5) in a normal density reinforced traffic of this type running on lines in Europe.
concrete deck slab of flat soff it. 'For further information on the derivation of the spectra, the
D.4.3 Clessification. Since the connectors are in following reports published by the Office for Research and
accordancewith Part 5, they also comply with 6.4.1 and Experiments of the International Union of Railways should
the weld throat stresses may be calculated according be consulted :
to 6.4.2 or 6.4.3. Report ORE D128/RP5
The potential fatigue crack should be classified as S (from Report ORE D128/RP6
type3.12oftable17(c).
Report ORE D128/RP7
D.4.4 Assessment using 8.2. Assuming, for this example,
that the shapeof the shear force influence line of the girder E.2 RL loading. The load spectra given in table 3 have been
under consideration is similar to that of a single simply based on the typical trains shown in figure 20. The numbers
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
supported girder, the value of L will lie between 15 m, for of each type of train assumed for the standard spectra,
connectors at mid-span, and 30 m for connectors at the together with the make-up of the total annual tonnages, are
ends of the member. shown in table 16.
Hence, for a dual carriageway all-purpose road and class S Table 15. R U loading: annual t r a f f i c t o n n a g e
-
(seefigureB(b). withthe 1.3factorallowedfor37.5unitsof
HB loading (see8.2.3(c)), OH 46 x 1.3 = 59.8 N/mm2
for connectors at mid-span, and u ~ = 4 0x 1.3 = 52N/mmZ Traffic
for standard traffic types
1 Train 1 Trrln 1Numbor 1 Total
for connectors at the ends. WP. typo. wolght, of trains annual tonnage,
tonnrs par annum tonnes x 10'
The above values may be checked against uvmax as
described in 8.2.2 and illustrated in D.2 or, alternatively, for Horvy
preliminarydesign purposes in this particular type of
example it may be noted that:
-
at mid-span up max will equal OP min
at the ends bp min will equal zero.
Hence, the limiting value of b p max may be determined as : Medium 5 I 600 122500 I 13.50
29.9 N/mm'for connectorsat mid-span and 7 11120 I 2411 I 2.70
52 N/mmZforconnectorsat theends.
8 1120 6027 6.75
Thus for stud connectors (see 6.4.2) the maximum 1 1794 2 257 4.05
allowable shear load per stud under loading from the
standardfatigue vehicle, positioned in accordance
with 7.2.3 may be expressed as :
light
29.9 Pu/425 = 0.070 PUkN at mid-span and
52.0 Pu/425 = 0.1 22 PukN at the ends 14516 5.40
where 8.1 0
Puis the nominal strength of the stud from Part 5. 4 I 172 (47093 I 8.10
Similarly, for bar or channel connectors, the maximum 5 I 600 I 4 500 I 2.70
allowable shear load per connector under loading from the I6 I 572 I 2360 I 1.35
standard fatigue vehicle, positioned in accordance
with 7.2.3, may be expressed as: 1 Total 127.00
29.9 x A, x lC3kN a t mid-span and *See fiaure 19.
52.0 x A, x lC3kN attheends.
where Table 16. RL loading: annual t r a f f i c t o n n a g e
A, is the effective weld throat area in mm2for the a n d composition of standard t r a f f i c m i x
particulartype of connector, obtained from 6.4.3.1.
D.4.5 Comments. It is not possible to use 8.3 for the
assessment of shear connectors as the damage chart (see
Train
type. 1 i!:;, 1 FcK I
perannum
tonnage,
tonneax 10'
figure 10) does not include factors for class S details. 1 I246 111 545 I 2.84
2 I253 I54 032 I 13.67
3 I280 I 9786 I 2.74
4 I203 I 6453 I 1.31
26 986
34
BS 5400: Part 10: 1980
225 6.1 21 1.9 1.5 1.5 5.2 1.5 1.5 1.5 1.5 5.2 1.5 1.5
l0vv
l#u
n*
b U b
4 1 - 41.4 Y#**H---+++-
2.5 21 L .O 3.52
1 Steel train total load = 1794 t V = 80 km/h
r
L Q I
t
eF++F=%
2.6 11.5 2.6 2.6 11.5 2.6
i m
3.6
L-a-a-a-L-a-a-a
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
2.7 11.5 2.6 2.6 11.5 2.7 11.5 2.6 2.6 11.5 2.6 2.6 11.5 2.7
+.+#-+
11.5 2.6 2.6
---$-$-+.$- - k2.7+-3c7(L- --$+++--. -;1Ly-+$- --IoL m
3.2 3.2 3.2 3.2 3.2
3 Southern Region suburban total load = 344 t V = 145 km/h
12 x L n10 t
11,
e , i1 ++
1.1 2.2 6.9 2.2 3.62.6 11.5 2.6
++++++-
2.2 22
1.0 m
35
I BS 5400: Part 10: 1980
22 2.2 36
7 Heavy freight total load = 1120 t V = 72 km/h for medium traffic and 120 km/h tor heavy traffic
8 Heavy freight total load = 1120 t V = 72 km/h for medium traffic and 120 km/h for heavy traffic
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
~~
6x20
1:: +%+:! !2n7:::; I Lx20
1:: &+ : : ~
6x20
,,11
6x20
1. t(
t
2.1 4.L 2.1 6.5 1.8 12.8 1.8 6.5 2.1 L.L 26 1.8 12.8 1.8
* v
1 1
I
v
1
LIbb
11-7
VLI/
4 4 1
YL,
6 4
U P
1 q m
2.1 2.1 4.6 3.8 3.7 3.7 4.6 21 2.1 4.5
L-B-B-R-B-B-B-B-~~B-B-B-B-B-S-R~B-B-R~B-0-B-B-S
NOTE. In deriving the table 2 spectra, impact effects were calculated in accordance with the recommendationsof Leaflet 776-1 R,
published by the International Union of Railways (UIC), 14 Rue Jean-Ray F, 75015 Paris.
36
2.4 7.9 2.4 2.4 8.0 2.4 2.4 8.0 2.1v 2.4
v v
8.0 2.4 2.4 7.9 2.4 2.4V 8.0 2.4 2.LV 7.9 2.L
V
A I 4 4 4 4 A A 4 4 4 4 A 4 4 4 A A 4 A A 4 1/1 4 4 A
3.2 3.1 3.1 3.2 3.2 3.2
1 total load = 246 t
%.L 7.9 2.4 2.L 8.0 2.4, 2.L 8.0 2.L z . 4 8.0 2.4 2.4 7.9 2 . t b4 8.0 2.t 2.4 7.9 $4
4 4 4 4
3.2
4 4 4 4
3.7 I31
4 4 4 4
3.2
4 4 4 4
3.2
1
3.21
1 /In
2.3U 8.5 2.3 2.3 7.3 2.3 2.3 8.5 2 3 , 2.3 7.3 2.3 2.3 7.3 2.3 2.3 8.5 2.3
A 4 4 1 A 4 4 4 A A / r 4 A A A A 1 4 4 4 flfl 4 4
3.3 3.2 3.3 32 3.3
5 total load = 209 t
2.3 8.5 2.3 2.3 7.3 2.3 2.3 8.5 2.3 2.3 7.3 2.3 2.3 7.3 2.3 2.3 8.5 2.3
w v U v u U
4 4 4 4 4 4 A A A A 4 4 n 4 A 4 1 4
3.3 3.2 3.3 3.2 /I 3.3
6 total load = 231 t
NOTE. In deriving the table 3 spectra, an impact of 3 0 %was taken for all trains and all spans
37
Appendix F
Examples of stress histories and cycle counting procedure
Example 1. Highway bridge. This shows the midspan bending of simply supported spans loaded by the standard
fatigue vehicle and illustratesthe effect of variations in L.
4at80kN
4-i- - 4-
I --
1.84 I 6m J.8 m
L < 1.8
I
/-\<'<n ,,/,r
I -
,- -w--y--j
..'.'.'...'.
t.,.,.,'.'.'
..
e",
Qv2
otl
ot 1
Stress histories Cycle counting diagrams
38
0
x
.( /
\
\
N c
E
-b em 23
.-U
CO
m
.-cc
C
a
0
0
-Q
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
U
Y
-
b 0)
Em 5
>.
U n
C U
m
c al
U)
Q
5
>
\
\
B- \
n
U \
5
Q
I
Q \ n
U U)
m \ 3
-0C 0
3
m .-cC
::
U
C
8C
0)
zn
CI
m
n
e
3 E
U)
-
z Q)
e
.-
U)
X
FI
m
u-
E 0
r-
m
WJ
.-C $
c
U)
r U
0 C E
C 9)
m n
n
x?
E
c
m
L
m
Q
c
U)
Q
5
:
c
U)
.-c
U)
+
ai
0
0
L
n
39
BS 5400: Part 10: 1980
Exemple4. Railway bridge. This shows the mid-span bending of a three-span continuous beam loaded with
standard RU loading.
4 at 250 k N
80 kNlm 80 k N l m
2
3
Stress history
,
r , - - ..
t
/
-.
. /
/
pR 3
NOTE. In examples 1 to 4 given above, the cycle counting diagrams, for comparative purposes, follow the same profile as the stress
history. Since the analogy dependssolely on the depth of water retained in each section, it is immaterial whether the profiles are as
illustrated or with successive peaks and troughs joined by straight lines (as proposed in appendix 6 ) .
40
- Appendix G
Testing of shear connectors
Appendix H
Explanatory notes on detail classification
G.l General. This appendix outlines the procedures which
should be followed if the fatigue strength of shear
H.l General
connectors is to be determinedby testing, as required H.l.l Scope. This appendix gives backgroundinformation
by 6.4.1. on thedetail classificationsgiven in tables 17(a), 17(b) and
17(c). This includes notes on the potential modes of failure,
6.2 Procedure.Test the specimens under constant
important factors influencing the class of each detail type
amplitude loading at frequencies not exceeding 250
cycles/min. Ensure that the frequency of the applied and some guidance on selectionfor design.
loading isthe same for each specimen within a particular H .1.2 Geometricalstress concentration factors.
series of tests. Ensurealso that the maximum load on any Unlessotherwise indicated in table 17, thestress
connector does not exceed 0.5 times the nominal static concentrations inherent in the make-up of a welded joint
strength of the connector (determined in accordancewith have been taken into account in the classification of the
Part 5) with the appropriateconcrete strength, a t the time of detail. However, where there is a geometricaldiscontinuity,
testing, being determined in accordancewith the such as a change of crosssection or an aperture (see figure
requirementsof BS 1881. 21 ) and/or where indicated in table 17, the resulting stress
Stresses may either be determinedfrom the applied test load, concentrations should be determined either by special
in accordancewith clause 6, or derived from strain gauge analysis or by the use of the stress concentration factors
given in figure 22.
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
readings.
P o t e n t i a l crack locations
nl /
Welded attachment
Manhole or r e - e n t r a n t corner
/
41
BS 5400: Part 10: 1980
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
Length of
st raight22r
K H W
RC
Stress
fluctuation
0;
%v
(b) Fatigue stress concentration factor for re-entrant corners KRc
(based on net stress at X)
42
Issue 2,March 1999 BS 5400: Part 10: 1980
H.2.2 Generelcomments. In welded construction, material which has previously been fully heat-treated, but
fatigue failure will rarely occur in a region of unwelded such componentsshould be subject to special test and
material since the fatigue strength of the welded joints will inspection procedures.
usually be much lower. Forthe useof black boltscomplying with the
requirements of BS 41 90 and subjected to fluctuating
tensile loads, see 6.5.
Where bolts or screwed rods are pre-tensioned to a value in
excess of an applied external load, stress fluctuations will be
H.2.3 C o m m e n t s onparticuler d e t a i l types governed by the elasticity of the pre-compressed elements.
The increase in tension will rarely exceed 10 %of an
Type 7.3.All visible signs of drag lines should be removed external load applied concentrically with the bolt axis, but
from the flame cut edge by grinding or machining. where the load is eccentric, a further increase will result
Types 1.3end 7.4. The presence of an aperture or re-entrant from prying action.
corner implies the existence of a stress concentration and the H . 3 Type 2 classifications, w e l d e d details on surface
design stress should be the stress on the net section o f member. Seetable 17(b).
multiplied by the relevant stress concentration factor (see
figure 22). H .3.1 Notes O n p O t 8 n t i 8 l m O d 8 S o f fei/ure.Seefigure23.
Type 1.4.The controlled flame cutting procedure should When the weld is essentially parallel to the direction of
ensure that the resulting surface hardness is not suff icient to stressing, fatigue cracks will normally initiate at the weld
ends, but when the weld is transverse, cracking will initiate
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
cause cracking.
at the weld toes. In either case the cracks will then
Type 7.5. This type may be deemed to include bolt holes for propagate into the stressed element. For attachments
attaching light bracing members where there is negligible connected by single welds, cracks in parent metal may also
transference of stress from the main member in the direction initiate from the weld root. Cracks in stressed weld metal
of Or. willinitiatefrom the weldroot (seetype3.1 l).Awayfrom
Type 7.6. This covers connectionsdesigned in accordance weld ends, fatigue cracks normally initiate at stop-start
with Part 3 for slip resistance a t the ultimate limit state and positions, or if these are not present, a t weld surface ripples.
where secondary out-of-plane bending of the joint is With the weld reinforcement dressed flush, failure tends to
restrained or does not occur (i.e. double-covered symmetric be associated with weld defects.
joints). Failure initiates by fretting in front of the hole.
H.3.2 Genera/comments.
Type 7.72. This classification applies to failure at the root of H.3.2.1 Edge disfence. (See figure 24.) No edge distance
the thread in normal commercial quality threaded criterion exists for continuous or regularly intermittent welds
components. Attention should be paid to the details of head away from the ends of an attachment (see types 2.1 to 2.5).
fillets, waisted shanksand thread run-out in components, However, a criterion exists (types 2.6 to 2.1 0) to limit the
not covered by an appropriate British Standard, to ensure possibility of local stress concentrations occurring at
that they have satisfactory fatigue resistance. A higher unwelded corners as a result of, for example, undercut, weld
fatigue resistance can be obtained with a rolled thread on
Long a t t a c h m e n t Short a t t a c h m e n t s
A
f >
w Y
Crack types Crack t y p e s 2.9 or 2.10
2.6. 2.7 or 2 8 ( o r 2.11 i f o n e d g e )
(or 2.11 a t edge)
NOTE. For classificationpurposes, an 'attachment'shouldbe
taken as the adjacent structural element connected by welding to
the stressed element under consideration. Apart from the
particular dimensional requirementsgiven for each type in table
17(b). the relativesize of the 'stressed element' and the
'attachment' is not a criterion.
Figure 23. Failure modes at weld ends
Avoid or g r i n d out to a s m o o t h profile,
any undercut t o these
0 BSI 03-1999 43
BS5400:PartlO:1980
spatter and excessive leg size at stop-start positions or cause local buckling (see Part 3). If intermediategaps
accidental overweave in manual welding. Although this longer than 2.5 hare required the class should be reduced to
criterion can be specified only for the'width' direction of an F. This type also includes tack welds to the edges of
element, it is equally important t o ensurethat no accidental longitudinal backing strips irrespective of spacing, provided
undercutting occurs on the unwelded corners of, for that the welds comply in all respects with the workmanship
example, cover plates or box members (see figure 24 (b) requirements for permanent weldsand that any undercut on
and (c)). Whereit does occur, it shouldsubsequently be the backing strip is ground smooth. The effects of tack welds
ground out to a smooth profile. which are subsequently fully ground out or incorporated
into the butt weld by fusion, need not be considered.
Part 5 recommends the provision of a minimum edge
distance of 25 mm forshear connectors, hence the Type 2.7.The classification may be deemed to include stress
criterion given in this part will automatically be met. concentrations arising trom normal eccentricities in the
thickness direction.
.H.3.2.2Attachment ofpermanent backing strips. If a
permanent backing strip is used in making longitudinal butt This type includes parent metal adjacent to the ends of
weldedjointsitshould becontinuousor made continuous flange cover plates regardless of the shape of the ends.
by welding. These welds and those attaching the backing Types2.7and2.8.Where a narrow attachment is
strip should also comply with the relevant class transferring the entire load out of a wide member, as in the
requirements. The classification will reduce to E or F (type case of a welded lap type connection between, for example,
3.3 or 3.4) at any butt welds in the backing strip or class E J cross brace and a gusset, the stress i n the gusset at the end
at any permanent tack weld (see H.3.3. type 2.4). It should of the cross brace will vary substantially across the section.
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
be noted that transverse butt welds on backing strips may be For assessing the stress in the gusset the effective width
downgraded by tack welds close to their ends (see H.4.3, should be taken as shown in figure 25.
type 3.4).
H.3.2.3 Stress concentrations. These are increased, and
hence the fatigue strength is reduced, where :
(a) the weld ends or toes are on, or near, an unwelded
corner of theelement (see H.3.2.1) ;
(b) the attachment is 'long' in the direction of stressing,
and as a result, transfer of a part of the load in the element
to and from the attachment will occur through welds
adjacent to its ends ;
(c) such load transfer is through joints which are not
symmetrical about both axes of cross section of the
stressed element.
H.3.2.4 Weldforms. Full or partial penetration butt welded
jointsof Tform (such as would connect attachmentsto the NOTE. For failure in the cross brace at X the cross brace is the
surface of a stressed element) should be completed by fillet 'member' and the gusset is the'attachment'.
welds of leg length at least equal to 25 %of the thickness of
Figure 25. Effective width for wide lap connections
the attachment. The fillets exclude the possibility of an
increase in stress concentration arising at an acute re-entrant Type 2.70.This applies where any applied shear stress range
angle between the element surface and the toe of the weld, is (numerically) greater than 50 % of a co-existent applied
and thus, in considering the effects on the stressed element, direct stress range.
it is immaterial whether the attachment is fillet or butt
Experimental evidence indicates that where significant
welded to the surface, since a similar toe profile results in
shear stress co-exists with direct stress, the use of
both cases.
principal stress values may be conservative and
H.3.2.5 Tack welds. Tack welds, unless carefully ground out accordingly the classification is upgraded.
or buried in a subsequent run, will provide potential crack
Type 2.7 1 . This type applies regardless of the shape of the
locationssimilar to any other weld end. Their use in the
end of the attachment. In all cases, care should be taken to
fabrication process should be strictly controlled.
avoid undercut on element corners or to grind it out to a
NOTE. Apart from the width transverse to U,,neither the shape of smooth profile should it occur. In particular, weld returns
the end of an attachment nor the orientationor continuity of the across a corner should be avoided and the use of cover
weld at itsend affects the class.
plates wider than the flange, to which they are attached, is
H .3.3 Comments onparticular detail types not recommended.
Type 2. I Finish machining should be in the direction of H.4 Type 3 classifications, w e l d e d details a t end
Or. The significance of defects should be determined with connections of member. Seetable 17(c).
the aid of specialist advice and/or by the use of a fracture H.4.1 N o t e s o n p o t e n t i a l m o d e s o f failure. (See figure
mechanicsanalysis. The N.D.T. technique should be 26.)With the ends of butt welds machined flush with the
selected with a view to ensuring the detection of such plate edges, or as otherwise given below, fatigue cracks in
significant defects. This type is only recommended for use in the as-welded condition normally initiate at the weld toe
bridgeworks in exceptional circumstances. and propagate into the parent metal, so that the fatigue
Type 2.2. Accidental stop-starts are not uncommon in strength depends largely upon the toe profile of the weld. If
automatic processes. Repair to the standard of a C the reinforcement of a butt weld is dressed flush, failure can
classification should be the subject of specialist advice and occur in the weld material if minor weld defects are exposed,
inspection and should not be undertaken in bridgeworks. e.g. surface porosity in the dressing area (see H.4.3, type
Type 2.4.The limiting gap ratio m / h applies even though 3.3).
adjacent welds may be on opposite sides of a narrow In the case of butt welds made on a permanent backing,
attachment (as in the case of a longitudinal stiffener with fatigue cracks initiate at the weld metal-strip junction and
staggered fillet welds). Long gaps between intermittent then propagate into the weld metal.
fillet welds are not recommended as they increase the risk In fillet or partial penetration butt welds, fatigue cracks in
of corrosion and, in the caseof compression members, may weld metal will normally initiate from the weld root.
44
BS 5400 : Part 10 : 1980
- 3.2
or 3.3
3.2
3.11 3.11
3.3
3.9
ir U
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
e--+
NOTE. Fatiguecracks in reinforcingbars will normally initiate in
similar locationsto those for structuraljoints. given similar stress
conditionsand joint geometry.
Figure 26. Type 3 failure modes
45
BS 5400 : Part 10 : 1980
beneficial. Provided that N.D.T. is done after grinding, this weld is made on a permanent backing (type 3.4, see figure
treatment can be assumed to raise the class to D. 27). Dressing of the weld reinforcement isadvised to
overcome poor reinforcement shape resulting from the
Types3.3and3.4. These types may be used for holes which
greater misalignments which may occur in the jointing of
are either filled with plugs of weld metal or welded infill
sect ions.
plates. Such holes may also be required for stitching
laminations or repairing lamellar tears. The welds should be NOTE. This joint is frequently made using a semi-circular cope
full penetration and should be considered to be equivalent hole.Thisgives improvedaccessto the flange butt welds when
webs or longitudinalstiffenershave already been attached. The
to type 3.3 or, if welded onto permanent backing material, end of the web butt weld at the cope hole can be consideredto be
type 3.4.The slot or hole dimensions should be in equivalentto class D with a stress concentration factor of 2.4
accordance with appendix A of BS 51 35 : 1974. providedthat the end of the butt weld and the reinforcement
Plug welds should not be used in bridgeworks for within a distanceequal to the radius ( I )are ground flush. Cope
transmitting tensile force across two lapping plates. Their holes of 45O mitre are not recommended.
use for transmitting shear force is not recommended for Types3.7and3.8. Weld metal failure will not govern with
major structural connections, but where they have to be full penetration welds.
used, failure through the weld throat should be considered Where the third member is a plate it may be assumed that
to be class W, based on the minimum throat area projected plane sections remain plane in the main members and that
in the or direction. axial and bending stress distribution in the or direction are
Type 3.4.If the backing strip is fillet or tack welded to the unaffected. Where the third member is an open shape, for
plate (type 2.9)the detail class will not be reduced below example, an I section or a hollow tube, particularly i f
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
class F unless permanently tacked within 10 mm of the different in width, a discontinuity in the main memberstress
member edge, in which case it will be class G (type 2.11). pattern will occur. In this case the stress parameter should
Type 3.5. The effect of the stress concentration at thecorner be the peak stress Concentration at the joint. In theabsence
of the joint between two individual plates of different of published data on a particular joint configuration, the
widths in line may be included inthe classification. Where stress concentration factor may have to be determined by
the end of one plate is butt welded to the side of another, finite element or model analysis.
refer to type 3.9. Plane sections may be assumed to remain plane where the
Stress concentrations due to abrupt changes of width can main member stress can be continued through the transverse
often be avoided by tapering the wider plate (see types 3.2, member by additional continuity plating of comparable
3.3and 3.4). cross-sectional area, which is in line with the main member
components (see figure 28).In thistypeof connection it is
Type 3.6.Butt welds between rolled sections or between important that the joint regions of the third member are
built-up sections are prone to weld defects, which are checked before welding for lameller rolling defects and after
difficult to detect, in the region of the web/flange junction welding for lamellar tears.
(see f igure 27).Special preparations, procedures and
inspection may be undertaken in exceptional Where t w o flat plates intersect in the same plane, as in the
circumstances and type 3.3may then be applied unless the case of flanges at the junction of two girders, the stress
concentration factor due to the abrupt change of width
should be used (see figure 29).If the weld is a full
penetration butt carried out in accordance with all the
recommendations for type 3.5 the detail may be classed as
F2 without applying a stress concentration factor.
These types may be deemed to cover the case where a
narrow third member is slotted through a single main
member away from an end connection (see figure 30).In
thiscase, the third member should be assumed to transmit
the stress which the parent material would have carried
before the slot was cut. If the length of the slot is longer
than 150 mm in the ar direction. type 3.7 (full penetration
buttjoints) should be reclassed from F to F2.Note thatthis
detail should generally be avoided, when possible, as slots
are difficult to cut accurately and fit-up for welding is often
poor. Where member B iscalled upon to carry high tensile
stress, a slot in A avoids any risk from lamellar tears.
However, with respect to stress fluctuation in member B the
detail shown in figure 30 is type 2.11 (class G) at point Y.
If B is critical and A is not, circular cut-outs at the corners of
Figure 27. Type 3.8 joint B will improve the class to F (type 2.9).
46
Typical stress patterns
A
thickness
direct ion
Detail X (cruciform joint)
\pia t i n g
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
Figure28. Useof continuity plating to reducestress concentrationsin type3.7 and 3.8 joints
0;
Figure29. Cruciform junction between flange plates
cut out
Class F type 3.7
47
BS 5400:Part 10: 1980
I
type 3.1 1) will normally govern, unless the total weld Stress distribution 4--b
leg length is about twice the element thickness. It wilt also \ if
I
40
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
\
\
,
,,
P
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
n Standards publications referred to
BS 1881 Methods of testing concrete
BS 3643 IS0 metric screw threads
BS 3692 I S 0 metric precision hexagon bolts, screws and nuts
BS 419O I S 0 metric black hexagon bolts, screws and nuu
0s 4395 High strength friction grip bolts and associated nuts and washers for structural engineering
BS 4604 The use of high strength friction grip bolts in structural steelwork. Metric series
BS 5135 Metal-arc welding of carbon and carbon manganese steels
BS 5400 Stwl concrete and composite bridges
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
n
issue 2, March 1999 BS 5400 : Part 10 : 1980
This British Standard, having been prepared under the direction of In response to orders far international standards, it is BSI policy to
the Civil Engineering and Building Structures Standards Committee, supply the BSI implementationof those that have been published
was published under the authority of the Executive Board and as British Standards, unless othecwise requested.
comes i n t o effect o n 3 1 January 1980
8 BSI 03-1999 Information on standards
ISBN 0 580 10567 9
BSI provides a wide range of information on national, European
The following BSI references relate t o rhe work on this standard : and international standards through ita Libmy and its Technical
Committee reference CSB/30 Draft for comment 74/13197 DC Help to Exporters Service. Contact the Information Centre.
Tel: 0181 996 7111: Fax: 0181 996 7048.
BSI b the independantnationalbody mponsible for preparing
Bdtbh Standards.It presents the UK review on standards in Subscribing members of BSI are kept up to date with standards
E u r o ~ eand at the htematiial level. It is hcorporeted by Royal developments and receive substantial discounts on the purchase
Charter. price of standards. For details of these and other benefits contact
MembershipAdministration. Tel: 0181 996 7002;
Rwloirlonr Fax: 0181 996 7001.
British Standards are updated by amendment or revision. Users of
8ritbh Standards should make sure that they possess the latest Copyrlght
amendments or editions. Copyright subsists in all BSI publications. BSI also holds the
copyright. in the UK. of the publications of the international
It is the anstant aim of BSI to improve the quality of our products standardizationbodles. Except as permitted under the Copyright,
and secv(ces.We would be grateful if anyone finding an inaccuracy Designs and Patents Act 1988 no extract may be reproduced,
or ambiguity while using this British Standard would inform the stored in a retrieval system or transmitted In any form or by any
Seaetary d the responsible technical committee, the identity of - -
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI
-
British Standards InstiMion a 3 8 9 Chiswick High Road London W4 4AL Telephone 0181 996 9OOO Fax 0181 996 7400