You are on page 1of 57


Part 10 : 1980

Steel, concrete and

composite bridges -
Part 10: Code of practice for fatigue
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI

ICS 93.040


Issue 1, March 1999 BS 5400 : Part 10 : 1980

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).

Issue Page Issue

Front cover 2 6 Original
h i d e front cover blank 7 to 42 original
a 1 43 2
b blank 44 original
C blank 45 to da Original
d 1 49 2
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI

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

BS 5400 : Part 10 : 1 980 Issue 1, March 1999

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

4.1 Design life 9.2 Assessment without damage calculation 18
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI

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
5 E. Derivation of standard railway load
6.4 Stresses in welds attaching shear spectra
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

I '' Method of indicating minimum dass

requirementson drawings 4a

Effective width for wide lap
Type 3 failure modes
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI

I ;;: Reference stress in parent metal 5

- Reference stress in weld throat

27. Type 3.6 joint 46
28. Use of continuity plating to reduce
3. Axle arrangement of standard fatigue
stress concentrations in type 3.7 and
vehicle 7 3.8 joints 47
4. Plan of standard axle 7
29. Cruciform junction between flange
5. Designation of lanes for fatigue 47
purposes 9
30. Example of a 'third' member slotted
6. Transverse location of vehicles 10 through a main member 47
7. Impact allowance at discontinuities 10 Example of type 3.9 or 3.10 joint
31. 48
8. Values of oH for different road
32. Tee junction of two flange plates 48
categories 13 33. Alternative method of joining two
9. Derivation of ov and &. for damage flange plates 48
calculation 15 34. Single fillet corner weld in bending 48
10. Damage chart for highway bridges
(values of d, 20) 16
11. Miner's summation adjustment factor
KF for highway bridges 17

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
Part 6 specification for materials and workmanship,
Part 7 Specification for materials and workmanship,
concrete, reinforcement and prestressing

0 BSI 03-1999 1
BS5400:Part10:1980 Issue 2,March 1999

1. s c o p e 3. Definitions and symbols

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

(b) methods using first principles that can be applied in

all circumstances. NOTE. A convenient mode of expressinga load spectrum is to
denote each load intensityas a proportion ( K w ) of a standard load
1.4 Other sources o f fatigue damage. The following and the number of occurrences of each load as a proportion (Kn)
topics are not specifically covered by this Part of this of the total number of loading events.
British Standard but their effects on the fatigue life of a
structure may need to be considered :
3.1.4 standardlosdspectrum. The load spectrum that
has been adopted in this Part of this British Standard,
(a) aerodynamically induced oscillations; derived from the analysis of actual traffic on typical roads or
(b) fluctuations of stress in parts of a structure immersed rail routes.
in water, which are due to wave action and/or eddy 3.1.5 stress history. A record showing how the stress at a
induced vibrations ; point varies during a loading event.
(c) reduction of fatigue life in a corrosive atmosphere
3.1.6 combinedstress history. A stress history resulting
(corrosion fatigue).
from two consecutive loading events, i.e. a single loading
1.5 Limitations event in one lane followed by a single loading event in
1 . 5 . 1 Steeldecks. Highway loading is included in this another lane.
Part and is applicable to the fatiguedesign of welded 3.1.7 stress cycle (or cycle ofstress). A pattern of
orthotropic steel decks. However, the stress analysis and variation of stress at a point which is in the form of two
classification of details in such a deck is very complex and is opposing half-waves, or, if this does not exist, a single
beyond the scope of this Part of this British Standard. half-wave.
1.5.2 Reinforcement. The fatigue assessment of certain 3.1.8stressrange (orrange ofstress) (U,). Either
details associated with reinforcing bars is included in this
Part but interim criteria for unwelded bars are given in (a) in a plate or element, the greatest algebraic difference
Part 4. between the principal stresses occurring on principal
NOTE. These criteria are at present under review and revised
planes not more than 45" apart in any one stress cycle : or
criteria may be issued later as an amendment. (b) in a weld, the algebraic or vector difference between
1.5.3 Shearconnectors. The fatigue assessment of shear the greatest and least vector sum of stresses in any one
connectors between concrete slabs and steel girders acting stress cycle.
compositely in flexure is covered in this Part, but the 3.1.9 stress spectrum. A tabulation of the numbers of
assessment of the effects of local wheel loads on shear occurrences of a l l the stress ranges of different magnitudes
connectors between concrete slabs and steel plates is during a loading event.
beyond the scope of this Part of this British Standard.
This effect may, however, be ignored if the concrete slab 3.1.10 design spectrum. A tabulation of the numbers of
alone is designed for the entire local loading. occurrences of all the stress ranges caused by all the loading
events in the load spectrum, which is to be used in fatigue
assessment of the structural part.
2. R e f e r e n c e s
3.1.1 1 detai/c/ass. A rating given to a detail which
The titles of the standards publications referred to in this indicates its level of fatigue resistance. It is denoted by the
standard are listed on the inside back cover. following :A, B, C, D, E, F. F2,G, S, or W.
NOTE. The maximum permitted dass is the highest recommended
class, that can be achieved with the highest workmanship specified in
Part 6 (seetable 17). The minimum required class to be specified for
fabrication purposes relates to the lowest q N curve in figure 14,
which results in a l i e exceeding the design lie.

2 0 BSI 03-1999
Issue 2, March 1999

- 3.1.1 2 Qr -N relationship or Or -Ncurve. The

quantitative relationship between Or and N f o r a detail
which is derived from test data on a probability basis.
N Miner's summation
Stress on the core area of a bolt, determined on
the basis of the minor diameter
UH Limiting stress range under loading from the
3.1 -13 design ITr-Ncurve. The Or-Nrelationship
standard fatigue vehicle on a highway bridge
adopted in this Part of this British,Standard for design on
UN Stress on net section
the basis of 2.3 % probability of failure.
Qo Constant amplitude non-propagating stress
3.1.14 design life. The period in which a bridge isrequired range(oratN= 10')
to perform safely with an acceptable probability that it will QP Algebraic value of stress in a stress history
not require repair. u p max l, Maximum and minimum values of oPfrom all
3.1.1 5 standard design life. 120 years, adopted in this Qp min j stress histories produced by standard loading
Part of this British Standard. J'r Range of stress (stress range) in any one cycle
3.1.1 6 Miner's summation. A cumulative damage Q r i , Or2 Individual stress ranges (ar) in a design
summation based on the rule devised by Palmgren and Miner. , . . etc spectrum
'JR max ((ip max - GP min) for a railway bridge
3.2 Symbols. Thesymbols in this Part of this British *R 1,'JR2 Stress ranges (in descending order of
Standard are as follows. . . . etc magnitude) in a stress history of a railway bridge
under unit uniformlydistributed loading
A Net area of cross section Limiting stress range under standard railway
A1 Effective weld throat area for the particular type
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI

of connector Nominal ultimate tensile strength, to be taken as
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
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
*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
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 .
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 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 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 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
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). Each classified detail is illustrated and given a type
number. Table 17 also gives variousassociated criteria and
E Fat

the diagrams illustrate the geometrical features and

potential crack locations which determine the class of each
detail and are intended to assist with initial selection of the
appropriate type number. (For important features that
change significantly from one type to another see the c ) f )

footnote to table 17.) Fat Fat E detail should only be designated a particular Figure 1. Method of indicating minimum class requirements on
classification if i t complies in every respect with the
tabulated criteria appropriate to its type number. drawings
4 0 BSI 03-1999
5.3.2 Detrimenta/ effects. The following occurrences
can result in a detail exhibiting a lower performance than its
classification would indicate :
(a) weld spatter;
(b) accidental arc strikes;
(c) unauthorized attachments;
(d) corrosion pitting.
5.4 Steel decks. The classifications given in table 17
should not be applied to welded joints in orthotropic steel
decks of highway bridges ; complex stress patterns usually
occur in such situations and specialist advice should be
sought for identifying the stress range and joint
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI

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 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). 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

Part 5 ) ; appropriate signs of the two orthogonal direct stresses

(d) stresses in triangulated skeletal structures due to load and the shear stresses at the point under consideration.
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI

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
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

I 4
I i ,Welded attachment

Design s t r e s s P M/z)

- L M-M' P o t e n t i a l crack
/ \
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)

I Figure Zb.Reference stress in weld throat

0 BSI 03-1999 5
BS 5400:Part 10: 1980

6.4 Stresses in w e l d s attaching shear connectors 7. Loadings for fatigue assessment

6.4.1 General. For shear connectors in accordance with 7.1 Design loadings. Highway and railway design
the dimensional recommendations of Part 5, the design loadingsappropriatefor bridges in the UK are given in 7.2
stresses for fatigue in the weld metal should be calculated and 7.3 respectively.
in accordance with 6.4.2 and 6.4.3. Where thedimensions The load factors YfL and Y f 3should be taken as equalling 1.O
of the shear connectors and/or the concrete haunches are (see Part2).
not in accordance with Part 5, the fatigue strength should be
7.2 H i g h w a y loading
determined in accordance with appendix G of this Part.
7.2.1 General. In determining the maximum range of
6.4.2 Studconnectors. The stresses in the weld metal
fluctuating stress, generally, only the vertical effects of
attaching stud shear connectors should be calculated from
vehicular live load as given in clause 7 should be
the following expression :
considered, modified where appropriate to allow for impact
stress in weld = as given in 7.2.4. In welded members the dead load stress
need not be considered. In unwelded members the dead
longitudinal shear load on stud
x 425Nlmm load stress will have to be considered in determining the
appropriate nominalstatic strength (from Part 5) effective stress range when compression stresses occur
6.4.3 Channelandbar connectors (see6.1.3). The stresses in the weld metal attaching channel Centrifugal effects need only be considered for
and bar shear connectors should be calculated from the substructures (see 7.2.5).
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI

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 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. 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. It may assist calculation to note that in normal
density concrete, where the thickness of the beam flange is 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 (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 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 Path of vehicles. The mean centre line of travel of
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 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. 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.

BS 5400 : Part 10 : 1980

80kN 80 kN 80, kN 80. kN (standard axles 1

- 1.8 m
t -
6.0 m
1.8 m

Figure 3. Axle arrangement of standard fatigue vehicle

twin tyre
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI

Alternatively a 225 mn
dia. circle may be used

Tyre contact area

load 20 k N per tyre

+tyre +tyre

Figure 4. Plan of standard axle


Table 1. Annual flow of commercial vehicles ( n , x 10 6 )

Category of road

I Number of millions of
vehicles per lane,
per year ( n e )

I Carriageway

I Number of
lanes per



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 (10 m * )
1 I 1.0 Not
All purpose Single (7.3 m * ) 2 0.5 Not
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI 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
detail under consideration.
the centrifugal force per axle = -

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)

SI. Adj. F F Adj. SI.

Two and three lane dual carriageways

Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI

SI. ' Adj.

n Two lane slip road

SI. ' F SI.

Three lane single carriageway

-L -- --I-

Two lane single carriageway

Lane marking

Hard shoulder or Slow l a n e Adjacent lane Fast lane
hard strip

NOTE. For two lane dual omit fast lane.

Figure S. Designation of lanes for fatigue purposes

BS 5400 : Part 10 : 1980

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

of mean c e n t r e l i n e Mean centre line of t r a v e l

of vehicles (selected t o cause maximum
stress range)

4 wheels 4I
c wheels

-Mean p o t hs
of wheels

900 900
I All dimensions are in millimetres
I Figure& Transverse location o f vehicles

Di scontinuit y
4 /
Adjusted stress

Figure 7. Impact allowance at discontinuities

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
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI


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 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 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 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

30 0 8 6 0 0 2 6 5 procedure given in 8.4 if it is a class S detail. However, If

250 1 1 3 0 0 0 80 ov may > 1.30OH for class S details or > 1.55 OH for the
NOTE 1. L is the base length of the point load influence line (see other classesthis option will not satisfy the recommen-
figure 12). For intermediatevalues of L. permissible stress ranges dations of 8.3 and 8.4.
may be derived from the spectra for the two adjacent lengths (b) The detail may be strengthened in order to reduce the
shown in the table and the values interpolated. n~ values apply to value of ov max or it may be redesigned t o a higher class.
one track.
NOTE 2. The values are based on a traffic volume of 27 x 1O6 8.2.3 Adjustment factors foraH. class S details only.
tonnes per annum. The values of OH obtained from figure 8 may be adjusted by
multiplying successively by the following factors where
8. Fatigue assessment of highway bridges
(a) Non-standard design life :
8.1 M e t h o d s of assessment
8.1.1 General. Three procedures for the fatigue assessment factor = ( 120 ..-)
design life in years

of details in highway bridges are given in 8.2,8.3 and 8.4.

The selection of the appropriate procedure depends upon (b) Non-standard annual flows :
the detail classification, the design life, the load spectrum nc (from table 1 ','*'
and the assumed annual flow of commercial vehicles.
8.1.2 Simplified procedures. As an alternative to the
factor =
( nc (assumed) )
more rigorous procedure of 8.4, the simplified procedures where
nc is the annual flow in the lane loaded to produce
of 8.2 and 8.3 may be used provided the conditions stated
are satisfied. Ov max - Op max - b p min
NOTE. In the case where opmsx and op," are produced by
NOTE. Appendix C gives the derivationof standard highway
bridge fatigue loading. loading in two lanes, n, should be taken as the sum of the flows
in those two lanes.
8.2 Assessment w i t h o u t damage calculation
(c) Reduced values of abnormal load capacity
8.2.1 General. This method determines the limiting value (see C.4.4.2) :
of the maximum range of stress for a 120 year design life
factor = 1.3 for bridges designed for 37.5 units H B
and is generally simpler but more conservative than the
factor = 1.7 for bridges designed for 25 units HB
more exact methods of 8.3 and 8.4. It should only be used
where all the following conditions are satisfied : 8.3 Damage calculation, single vehicle m e t h o d
(a) the detail class is in accordance with table 17 ; 8.3.1 General. This method determines the fatigue life of
(b) the design life is 1 20 years ; the detail in question and may be used where a more precise
(c) the fatigue loading is the standard load spectrum assessmentthan that provided by the method of 8.2 is
(see ; required or where the standard design life and/or the annual
flows given in table 1 are not applicable. It should only be
(d) the annual tlows of commercial vehicles are in
used where the following conditions are satisfied :
accordance with table 1 .
(a) the detail class is in accordance with table 1 7 but is
NOTE. For class S detail only, 8.2.3providesfactorsby which
non-standarddesign life, different traffic flow and design HB not class S,
loading of less than 45 units may be taken into account. (b) the fatigue loading is the standard load spectrum
8.2.2 Procedure (see ) . The following procedure should be used (see 8.3.2 Procedure
appendix D) : The following procedure should be used (see
(a) apply the standard fatigue vehicle to each slow and appendix D).
each adjacent lane in turn, in accordance with 7.2.3;
(a) Apply the standard fatigue vehicle to each slow lane
(b) apply the impact allowance of 7.2.4, i f appro- and each adjacent lane in turn, in accordance with 7.2.3.
priate, anddeterminethe maximum and minimumvaluesof
principal stress or vector sum stress for weld throat, (b) Apply the impact allowance of 7.2.4,if appropriate,
o p max'and o p min occurring at the detail being assessed, and determine the algebraic value of principal stress, or
whether resulting from the fatigue vehicle in the same lane for weld throat, the vector sum stress at the detail being
or not ;





Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI

1 2 5 10 20 50 100 200

(b) Uual two lane motorway. dual three lane all

purpose, dual two lane all purpose




bs 40 C
cE E
ul S
.E 20

L lm)
(c) three lane all purpose, two lane all purpose (1 0 m),
two lane slip road

Figure 8. Values of OH for different road c.ategories

BS 5400 Part 10: 1980

assessed .aeach peak and each trough in the stress 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
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 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
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 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 ; 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
. :
fatigue life (in years) =
ZKFEcd, 20
where the summation includes all the separate lane stress
histories as well as the combined stress history, where

BS 5400 : Part 10: 1980

c Lane reference* I

Lane a t r o u hinow Lane streas

Number of cyclea
per loedlng *V*nt
ffoctlve lane

I - 1 A .
Lane A

Lane B

Lane C
I -p
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI

Lane 0
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


Combined stress history A/B

5 min
Case 2. Highest peak and lowest trough withvehicles in different lanes

0 Peak stress Stress range Q,
0 Trough stress for cycle shown
X Datum stress

*The lane reference letters A, B etc. should be allocated i n

descendingorder of magnitude of stress ranges O V l A , CJvi 6 etc.
t Values of b, should be obtained by the method given in
appendix B.

Figure 9. Derivation o f gvand is, f o r damage calculation

Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI

BS 5400 Part 10: 1980


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
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.

Figure 11. Miner's summation adjustment factor KF f o r highway bridges

- -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).

Figure 12. Typical p o i n t load influence line

BS 5400 : Part 0 :1980

I ,Simplified design spectrum

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

(d) Obtain the appropriate limiting stress range b~

9. Fatigue assessment of railway bridges
from the following expressions:
UT = k , x k; x k 3 x k , x k s x d ofor RU loading
9.1 Methods of assessment
QT = k l x k 2 x k 4 x k s x ke x dofor RLloading
9.1.1 Gener81. Two methods for the fatigue assessment of
details in railway bridges are given in 9.2 and 9.3. The where
choice of the appropriate method depends upon the detail k , = 1.O if the design life is 120 years, otherwise it is
classification and the nature of the loading. obtained from 9.2.3
9.1.2 Simplifiedprocedure. 'As an alternativeto the k 2 = 1.O if the loading event produces only one cycle of
more rigorous method of 9.3 the simplified procedure stress, otherwise it is obtained from 9.2.4
of 9.2 may be used provided the conditions stated therein kr is obtained from table 4
are satisfied. k 4 is obtained from table 5
k B is obtained from table 6
9.2 Assessment w i t h o u t damage calculation k s is obtained from table 7
9.2.1 Gener81. This method determinesthe limiting value go is the constant amplitude non-propagating stress
of the maximum range of stress for the specified design life. range for the appropriate class of detail and is
It should only be used where the following conditions are obtained from table 8
satisfied : NOTE. The sign conventionused for bpis immaterial providing
(a) the detail class is in accordance with table 17 ; it is consistentlyapplied. Where stress reversaldoes not occur
under the loading described, either bpmsx or bp,,,in should be
(b) the loading is the standard railway bridge loading in be taken as zero.
accordance with 7.3. Where U R max does not exceed UT the detail may be
The simplified procedure producesthe same results as the considered to have a fatigue life in excess of the specified
method given in 9.3 when the coefficientsk,, k,, k 4 and design life.
k 5 (see 9.2.2) are equal to unity. In other cases the method
may be more conservative than the method given in 9.3. Where CYR max is found to exceed GT either of the
following options may be adopted :
9.2.2 Procedure (a) the detail may be assessed by the more precise The following procedure should be used. procedure given in 9.3;
(a) Apply the standard railway loading in accordance (b) the detail may be strengthenedin order to reduce the
with 7.3.1 and 7.3.2. value of OR max or it may be redesignedto a higher class.
(b) Determinethe maximum and minimum values of 9.2.3 Non-st8nd8rd designlife. Where the specified
principal stress or vector sum stress for weld throat design life is other than 120 years, the value of k , should be
u p max and CJ p min, occurring at the detail being assessed, taken as the lesser of either :
by loading the appropriate loops of the point load
influence line, asshown in example 4of appendix F.
whether resulting from railway loading on the same track (a) (design life
120in years )
or not. or
(c) Determinethe maximum range of stress OR max equal
to the numericalvalue of crp max-bp min. For non-welded
(b) (design life in years
details the stress range should be modified as given
in 6.1.3. where
m is the inverse slope of the br -N curve appropriate to
the detail class and is obtained from table 8.

BS 5400: Part 10: 1980

Heavy trafflc Mullurntrafflc Llght trafflc

Detail clem D C 8 S 0 C B S O C 8 S
F2 F2 F2
0 0 0

Length. L (m) Valume of &3

< 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).

Table 5. Values of k4 for railway bridges

42 to 27 27to18 18to12 12to7 7to5 <5

0.89 1.o 1.13 1.27 1.42 1.6

Table 6. Values of k5 for railway bridges

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

1.42 1.27 1.125 1.o 1.o 0.89

NOTE. P , is the numerical value of stress due to track 1.

P, is the numerical value of stress due to track 2 (where track 1 is the track
causing the greater stress at the detail under consideration),
Trains on two tracks should be considered in the same longitudinalposition.

Table 7. Values of ks for RL loading of

railway bridges
Detail class 0 C 8 S

Length, L (m) V ~ I J OO
S f ha

< 3.0 1.23 1.28 1.35 1.65

3.0 to 3.4 1.34 1.37 1.45 1.71
3.4 to 4.0 1.43 1.49 1.55 1.80
4.0 to 10.0 1.57 1.62 1.68 1.91
10.0 to 15.0 1.77 1.79 1.90 2.10
15.0 to 20.0 1.98 1.99 2.00 2.10
> 20.0 2.08 2.05 2.09 2.10
B S 5 4 0 0 : P a r t l O : 1980

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+(--) +(-)
+.... given in 6.13. 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 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 (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 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 by the are of the same sign.
coefficientkS which can beobtained from table 6. 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)
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 ) or OR max/ks(see 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 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.
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI

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
> 1.O, the
N x U': = K, following alternative actions should be considered.
Log,, N = Log10 Kn m Logqour
(a) strengthen the detail to reduce the values of Or. The
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

11.3 Treatment of low stress cycles. The number of

repetitions of each stress range Or less than U , should be
reduced in the proportion (Ur/ao) 2
where m is obtained from table 8
U , is the stress range given by the equation in 11.2 for or
N= 10' and tabulated in table 8. (b) redesign the detail to a higher class. As a guide for
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI

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

times the value of uoof the original class of the detail.

11.4 Procedure. The following procedure should be used
in applying the Palmgren-Miner rule. Table 8. ur -N relationships and c o n s t a n t
a m p l i t u d e n o n - p r o p a g a t i n g stress r a n g e values
(a) Determinethe class of each detail in accordance with
table 17.
(b) Calculatethe stresses and hence the stress ranges at W 13.0 0.16 x 1012 25
each detail in accordance with clause 6 and determine G 13.0 0.25 x 10" 29
the design spectrum in accordance with 8.4 and 9.3. The F2 13.0 10.43 x 10'2 I 35
number of low stress cycles should be modified in
accordance with 11.3. F 13.0 10.63 x 10'2 I 40
(c) Determinethe number of repetitions to failure Nof E 13.0 1 1 . 0 4 ~ 1 0 1 2 I 47
each of the stress ranges in the design spectrum in D (3.0 1 1 . 5 2 ~ 1 0 1 2 I 53
accordance with 11.2. C 13.5 14.23 x 1013 I 78
(d) Evaluate Miner's summation in accordance with 11.leB 14.0 11.01 x l O l s 1100
S 18.0 12.08 x 1022 I 82
NOTE. Values applicable to non-standardcriteria may be obtained
from appendix A.

BS 5400: Part 10: 1980

- Appendix A
Basis of a,-N relationship
Table 10. Probability factors

::: ;i
Probmbllity of
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

mean-line. propagating range ooequal to the value of urobtained from

the formula in A.l when N = 1 07.
NOTE. This corresponds to a certain probability of failure. as
shown in table 10. When the applied fluctuating stress has varying amplitude,
The relevant values of these termsare given in tables 9 so that some of the stress ranges are greater and some less
and 10 and the mean-line relationships are plotted in than uo,the larger stress ranges will cause enlargement of
figure 15. the initial defect. This gradual enlargement reduces the value
of the non-propagating stress range below u0.Thus, as
Table 9. Mean-line or-N relationships time goes on, an increasing number of stress ranges below
uocan themselves contribute to the further enlargement of

: ;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
BS 5400:Part 10: 1980

The same result can be obtained by using a notional

log ar/log Ncurve, which has the inverse slope m 2 +
where Nisgreaterthanlo.
These points are illustratedin figure 16, which shows a
typical log Or/lOg Ncurve.
A.3 Fatigue l i f e for various failure probabilities. The Datal1 Numbmrofatmndard davlatlona
standarddesign Qr-Ncurves in figure 14are based on two clar balowthamun-llna

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
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


W 1.24 1.53 1.89 2.34

G 1.23 1.51 1.86 2.28
F2 1.30 1.69 2.20 2.85
F 1.29 1.65 2.1 2 2.73
E 1.34 1.78 2.38 3.18,
D 1.27 1.62 2.06 2.63
C I.27 1.60 2.02 2.56
B 1.23 1.52 1.88 2.32
S I.79 3.20 5.71 10.21

itatic limitations

Constant amp1i tude loading

in clean a i r

----- ----_-_
Effective curve obtained under
I variable amplitude loading,
I \
I \
I \
I \
m+2 -.\
I \
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

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 (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.
c.1.2. Standerdfetigue vehicle. (See 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

First occurrence I Second occurrence

I -
- -.

H i ghes t peak Imaginary reservoir

\ /

Shaded oreos a r e thoser

p a r t s of t h e reservoir
which successively
become e m p t y V

BS 5400 : Part 10 : 1980

C.1.3Stenderdlene flows. (See 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 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.

BS 5400 : Part 10 : 1980

? 5% 5K %K
- 5544 m m m

0 0 0 3 0 0 0 0 3 0 0
3 88
D O 0
0 0 0 3 0
0 0 0 3 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 -

0 0 0 0 0 0 c)mm
0 0 -03 dQ)
m m w m m
(Do N c r
r c
0 0 oal 0 0
0 0 -03 d
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
0 0 0 0
7 0 3 N
0 0
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 0
dco 0 0 0 0
0 0
2: 2:
zz 0
0 0
coo 0 0 0 0
F - (zm (zco

0 0

- 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

.E n

1 1 1 1 1 I x z - I rz-l


- Pm p
0 ID

c" .-0

d m (v

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

Table 12. Proportional damage from individual groups of typical commerclal vehicles


-- -
D to a, F i s n d W C % i
Dstril clrw
'm 3.0) m
! 1.5

- -
1.5 L - 25
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 %


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
600 mm

Figure 17. Multiple paths

BS 5400 : Part 10 : 1980


1 0
Figure 18. Typical Miner's summation adjustmont C U ~ V O
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI

- Table 13. Typical commercial vehicle gross

weight spectrum
Table 14. Typical commercial vehicle axle
weight spectrum
Vehicle Proportion of Proportion Total axle Totalnumber of exlerfor
of tote1 weight 106 vohicloa
264 240
0.000 01 231 120
18GT-M 2.38 0.00003 176 160
9lT-H 5.03 0.000 02 165 560
9TT-M 2.34 0.000 04 154 100
7GT-H 14.09 I 0.00003 143 780
7GT-M 2.13 0.000 07 121 80
110 90 040
7A-H 2.47 0.000 02 99 240 280
5A-H 1.97 0.000 28 93 320 000
5A-M 1.13 0.01 4 50 88 59 320
0.015 77 59 350
71 180 000
66 59 930
4A- M 0.090
61 165 000
55 290 040
4R-H 0.88 0.015 49 150 000
4R-M 0.75 0.01 5 44 120 000
4R-L 0.38 0.01 5 39 320 000
3A-H 0.67 0.030 33 380 000
3A-M 0.44 0.030 22 60 000
3A-L 0.28 0.030 17 360 000
3R-H 0.75 0.01 5 Total 2 856 000 axles for 10'
3R-M 0.61 0.01 5 vehicles
3R-L 10.38 I 0.015
2R-H 10.42 I 0.170 NOTE 1. This table is based on L < 1.5 m.
2R-M 0.20 0.1 70
2R-L 10.09 10.180 NOTE 2. Thesevaluesinclude the 1 0 %increase referred to in C.2.

Total 1.0

NOTE. This table is based on L = 25 m (rectangular loop).

Appendix D a vertical stiffener, given that analysis, in accordance
with (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

Alternatively, 8.3 provides a more precise method for detail -

Henceuv max = 17.8 ( -8.0) = 25.8 N/mm2
classes B to G and F2 and W leading to a life prediction (to (see8.2.2.1 (c)).
97.7% probability of survival) and indicating the extent of Fora dual two lane motorway with L = 75 m and a class F
required changes in thedetail when the predictedlife is too detail,
short. U H = 20.5 N/mm(seefigure8(b)).
The basic assessment procedures for a steelwork detail are Hence UV max exceeds OH and adequate fatigue life is not
illustrated in 0.2, while the application of a combined stress demonstrated (see By reference to
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI

history is illustrated in D.3 (see8.3.2.1 (c)). Atypical either

procedure for shear connectors is given in D.4.
(a) the procedure of 8.3may be used, since
D.2 Example using t h e basic assessment procedures uv max < 1 . 5 5 ~ H
f o r a steelwork detail or
0.2.1 Type of bridge. A three span (50 m/75 m/45 m) (b) the detail should be strengthened to reduce UV max
twin girder highway bridge that carries a dual two lane or improved to a higher class.
motorway and is designed to carry standard UK loading.
D.2.5Assessment using8.3. Given thatthe stress
D.2.2 Details of theproblem. These assess the fatigue histories for each loading event (passage of the standard
resistanceof the main girder bottom flanges at the mid-span fatigue vehicle in each lane) are as follows:
of the main span, which is adjacent to the transverse weld of


- 8.0
1st carriageway, slow lane 1st carriageway, adjacent lane

d 7 -2.8

2nd carriageway, adjacent lane

-1.9 -1.5

2nd carriageway, slow lane


From figure 9, in the first carriageway, the slow lane will be

designated lane A, since it has the greateststressrange (i.e.
17.8 - (-8.0) = 25.8).The reservoir method (seeappendix
6)may be used to determine the values of oVas shown
below :


0 0

- 8.0
The values of uv for the other lanes may be determined in a
similar way.

BS 5400: Part 10: 1980

This assessment procedure can be tabulated as follows :

*D. 4 10t &: Domigo
numkr Puk Cliu F &d,m
17.8 0.32 0.48
1st Slow 1.5
1 l2 0 Neglect -
( < 0.001 )
12.0 0.032 0.03
0 Neglect -
( c 0.001 )


1 10.2 0.01 5 0.02

I 1 IlS5
2nd Adj. C 1.o
2 0 Neglect -
( < 0.001 )

2nd Slow
,"'" 1-1.5 7.8 0.001

Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI

-0.8 0.8 Neglect

( < 0.001 )

*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,
hence KF = 1.59 -
the stress range could be increased by
\ o'2si.e. by 1.04.
Total damage = Z K F ECd , 2o = 1.59 x 0.53 = 0.843
Estimatedlife= 2010.843 =i 4zyears(see (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.

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

Derive the combined stress history (see (c) and

case 2 of figure 9).

12.9 12.9

- 12.9 -1 2.9
Lane A/B

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
Lane B
2 cycles per
loading event

Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI

Combined A/B

,,A,B= 25.8

The assessment procedure can be tabulated as follows

Lene CVCI. Range dttot "C! Damaga

number OV Claaa 0 fie d i z o

1 12.9 0.055 0.03

A }I;,:. = 0.6
2 12.9 14.2 8.7 0.01 0.01


-1 2.9



} ;. 1 2:1 .2
= 0.6


1 :zl .2
= 0.6
Z& dc ao 0.86

*See appendix E.
tSee figure 10.
$See figure 9.

0.3.6Comments. The estimated life is in excess of the 60

Adjustment factor K F (see figure 11) year design life and hence a reduction in area i s allowable.
for separate histories
KO = 12.9112.9 = 1.O
The reduction factor will be( g)
. .
i.e. 0.91 7 (see

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.

0.4 Exampleof a typical procedure for shear Appendix E
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
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 :
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
D.4.5 Comments. It is not possible to use 8.3 for the
assessment of shear connectors as the damage chart (see
type. 1 i!:;, 1 FcK I
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

+See figure 20.

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
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


2.6 11.5 2.6 2.6 11.5 2.6
i m
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI

2 Electric multiple unit total load = 372 t V = 145 km/h

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

2.7 11.5 2.6 2.6 11.5 2.6 2.6 11.5 2.7

+t----H 3L# - $4 t t k+ m
3.2 3.2
4 Southern Region suburban total load = 172 t V = 145 km/h

12 x L n10 t
e , i1 ++
1.1 2.2 6.9 2.2 3.62.6 11.5 2.6
2.2 22
1.0 m

5 Diesel hauled passenger train total load = 600 t V = 160 km/h

3.3 6.7 3.3 2.6 11.5 2.6

p m
6 Electric hauled passenger train total load = 572 t V = 160 km/h

Figure 19. Trains included in table 2 spectra

I BS 5400: Part 10: 1980

9. 6.9 2.2 2 11.0 2 m

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

2.2 6.9 2.2 5.5

+++-.- 2.2 2.2 3.5

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

1:: +%+:! !2n7:::; I Lx20
1:: &+ : : ~
1. 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
4 4 1
6 4
1 q m
2.1 2.1 4.6 3.8 3.7 3.7 4.6 21 2.1 4.5

9 Mixed freight total load = 852 t V = 120km/h

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.

Figure 19. (Concluded)

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
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

9? NO *rn rn- 90 In- O N

%.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
4 4 4 4
3.7 I31
4 4 4 4
4 4 4 4
1 /In

2 total load = 253 t

Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI

Lx8.65 Lx8.74 L x8.7L Lx8.74 Lx8.65

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

L x 9.08 L 9.08 Lx 10.13 L x 10.13 LxlO.13 L x 9.08

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

Figure 20. Trains included in table 3 spectra.

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.


4-i- - 4-
I --

1.84 I 6m J.8 m

Influence line diagram Standard fatigue vehicle

Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI

L < 1.8

1.8 < L < 3.6

/-\<'<n ,,/,r
I -
,- -w--y--j

... ... ... ... ... ... .......

. .. .. . .. .. .. .
, ... ... .. .
.. ..
. .. ..


6 < L < 15.6

/ 5 2


ot 1
Stress histories Cycle counting diagrams

.( /

N c
-b em 23

Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI


b 0)

Em 5
U n
c al
B- \
U \
Q \ n
U U)
m \ 3
-0C 0
m .-cC

3 E

z Q)

E 0
.-C $
r U
0 C E
C 9)
m n



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

Influence line diagram RU Loading to be multiplied by the dynamic factor

L = base length of loop containing largest ordinate (measured i n
direction of travel)
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI

Loading diagram for uDml,. Loading diagram for upmin

Loading diagrams for stress history


Stress history

r , - - ..

. /

pR 3

Cycle counting diagram

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 ) .

- 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


- 6.3 Class S criteria. To enable the weld metal attaching

shear connectors to be classified as a class S detail, the
, welds should be shown by tests carried out in accordance
withG.2to havea97.7%probabilityofsurviving 106and
H.2Type 1classifications, non-welded details.
See table 17 (a).
H.2.1 Notes onpotentialmodes of failure. In
unwelded steel, fatigue cracks normally initiate either at
107 repetitions of stress ranges of 146 Nlrnm and surface irregularities,at corners of the cross sections, at
82 N/mm2 respectively,where the stress ranges are holes and re-entrant corners or at the root of the thread for
computed in accordance with either 6.4.2 for stud bolts or screwed rods. In steel, which is holed and connected
connectors or 6.4.3for bar and channel connectors. with rivets or bolts, failure generally initiates at the edgeof
Where these conditions are not satisfied, the design a,-N the hole and propagates across the net section, but in double
relationship for the detail (i.e. 2.3% probability of failure) covered joints made with H.S.F.G. bolts this is eliminated by
should be derived in accordance with appendix A and the the pretensioning,providing joint slip is avoided, and failure
method given in 8.4 should be used to assess the fatigue initiates on the surface near the boundary of the
life. compressionring due to fretting under repeatedstrain.

P o t e n t i a l crack locations

nl /
Welded attachment

ypical stress distribution

The design stress is

At the attachment thedesign
applied to the appropriate
stress is applied t o the
plain material classification
appropriate joint classification

Manhole or r e - e n t r a n t corner

I The design stress for either location should be

taken as the stress on the net section
multiplied by the stress concentration factor

Figure 21. Typical example of stress concentrations due t o geometrical discontinuity

BS 5400: Part 10: 1980
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI

(a) Fatigue stress concentration factor for unreinforced apertures KUA

(based on net stress at X)

Length of
st raight22r



(b) Fatigue stress concentration factor for re-entrant corners KRc
(based on net stress at X)

Figure 22. Stress concentration factors

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
f >

Weld failure cracksltype.3.11)


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

(a) (to toe of weld)

Figure 24. Edge distance

0 BSI 03-1999 43

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.

BS 5400 : Part 10 : 1980

- 3.2
or 3.3

3.11 3.11

ir U
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI

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

H.4.2 Generalcomments dependent upon the welding procedures adopted.

H.4.2.1 Misalignments.The classifications may be deemed Accordingly, their use is not recommended unless subject
to include for the effects of any accidental misalignments up to special tests and strict procedural control.
to the maximum value specified in Part 6, provided that the H .4.2.6 Partialpenetration butt welds. All butt welds
root sides of joints with single sided preparations (i.e. single transmitting stress between ends of plates, sections or
bevel -J, -U or -V forms) are back-gouged to a total width built-up members in bridgeworks should be full penetration,
at least equal to half the thickness of the thinner element. except,where permitted in types 3.8 and 3.1 0 (junctions
H.4.2.2 Design stresses. For elements where out-of-plane with transverse members). In these laner cases, even
bending is resisted by contiguous construction (e.g. beam though the joint may be required to carry wholly
flanges supported by webs, wide plates supported by compressive stressesand the non-penetrated surfaces may
effectively continuous stiffeners, etc.) eccentricities due to be machined to fit, for fatigue purposes, the total stress
axial misalignments in the thickness direction may be fluctuation should be considered to be transmitted through
neglected. Wheresuch support is not provided (e.g. tension the welds (e.g. column caps and bearing stiffeners).
links) the design stress should include an allowance for the H -4.2.7 Welding of reinforcing bars for concrete. Welding
bending effects of any intentional misalignment, i.e. the of reinforcement should comply with Parts 7 and 8. Lap
nominal distance between the centres of thicknessof the welding of bars is not classified since adequate control
two abutting components. For components tapered in cannot be exercised over the profile of the root beads and its
thickness, the centre of the untapered section should be use is not recommended under fatigue conditions.
used. H .4.3 Comments on particular detail types
H.4.2.3Elernentedges. In allcases,failures tend to be Type 3.7. The significance of defects should be determined
associated with plate edgesand careshould be taken to with the aid of specialist advice and/or by the use of a
avoid undercut at the weld toes on the corners of the cross fracture mechanics analysis. The N.D.T. technique should
section of the stressed element (or on the edge at the toes be selected with a view to ensuring the detection of such
of any return welds). Should it occur, any undercut should significant defects. This class should not normally be used
be ground out to a smooth profile. in bridgeworks (see
H.4.2.4 Part width welds. Bun type welds may also occur Type 3.2. Shop welds made entirely in the downhand
within the length of a member or individual plate as, for position, either manually or by an automatic process other
example, in the case of : than submerged arc, tend to have a better reinforcement
(a) a plug weld to fill a small hole; shape from the point of view of fatigue than positional. site
(b) a weld closing a temporaryaccess hole with an infill or submerged arc welds (i.e. larger re-entrant angles at the
plate ; toes and more uniform profiles). Accordingly, to cater for
exceptional circumstances, joints made in this manner may
(c) a hole or slot for a transverse member to slot through be up-graded to class D.
a wider member.
Types3.2and3.3.Thickness variations and surface
Although such geometries have not been given specific
misalignments up to the maximum valuesspecified in Part 6
categories in table 17(c), types 3.3 and 3.4 may be deemed
may be deemed to be included (see H.4.2.1).
to cover plug and infill plate welds and types 3.7 and 3.8
may be deemed to cover sloned members. See also H.4.3. These types do not normally include joints between rolled
or built-up sections. See the note to type 3.6.
H.4.2.5 Joints weldedfrorn one side only. Unless made on
a permanent backing (type 3.4) welds made entirely from Type 3.3.Grinding smooth the reinforcement of butt welds
one side are not classified since the root profile will be until flush with the plate surface on both sides is generally

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).

Typical stress patterns

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

Figure29. Cruciform junction between flange plates

cut out
Class F type 3.7

Alternative detail if member '8' is critical

NOTE. Main rnernber'A issloned, 'third' rnernber'8' is
continuous and is welded all round the slot to'A.

Figure30. Example of a'third' member slotted through a main member

BS 5400:Part 10: 1980

Type3.8. In afillet welded joint, weld metal failure (see

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

govern in a partial penetration butt welded joint except

where reinforced with fillet welds of adequate size.
Types3.9end3.tO.Thesedetail types are distinguished
from types 3.7 and 3.8 by the absence of a similar member in
line on the far side of the joint. In this case an axial
component in the first member will induce bending moment
and hence curvature in the transverse member. Unless the
latter is very stiff in bending an uneven stress distribution
will result. Members with bolted end connections via
transversely welded end plates are particularly susceptible
to localincreaseofstress (seefigure31). Notethat if the
transverse member is an open or hollow section, local
bending will increase the peakstress further (as in the case
of types 3.7 and 3.8).
As far as fatigue failure of the transverse member is Figure 31. Example o f type3.9 or 3.10 joint
concerned, the first member is treated as a type 2
attachment and thestress parameter is the stress in the
Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI

transverse member without the application of a stress Full p e n e t r a t i o n

concentration factor. In hollow or open transverse members
this stress is often magnified by local bending of the walls.
b u t t weld
Often the load is transmitted from a member to a transverse
member primarily via flange plates in the same plane. This
can occur in the case of a junction between cross girders and
main girders, diagonals and truss chords, or in vierendeel
frames (see figure 32). If the transverse member is relatively
stiff (i.e. its width isat least 1.5 times the width of the first
member) and a full penetration butt weld is used in
accordance with the recommendations ot type 3.5, the
classification in this particular case may be considered to be
eftectively F2 with a stress concentration factor of unity.
Otherwise the class shall be F with the appropriate stress
concentration factor (see comment on detail types 3.7 and
3.8). Note that in the case of trusses, secondary stresses due
to joint fixity should be taken into account. The fatigue concentration f a c t o r ?
strength of both flange plates may be improved by the
insertion of a smoothly radiused gusset plate in the Flgure32. Tee junction o f t w o flange plates
transverse member so that all butt welds are well away from
re-entrant corners (see figure 33).
Type3.11. Class W is primarily intended to apply to all
fillet or partial penetration butt weld joints where bending
action across the throat does not occur. Where lapped joints
are welded on t w o or more sides, or tee or cruciform joints
are welded from both sides (asshown in table 17(c)), such
bending action is normally prevented. In certain cases
difficulty of access may only allow welding to be done on
one side of the joint. This applies particularly to small
hollow members with welded corners, which if subject to
loading that distorts the cross section, may cause failure of
the corner weld in bending (see figure 34). Whereaxial
stress i s also present, the stress range at the face of the weld
may be different from that at the root. Failure from ripples or
stop-start positions on the face may give a higher strength
than class W, but expert advice should be sought if a higher
strength is required. In most cases failure from stress
fluctuation in this root will be critical and thisshould always Figure 33. Alternative method o f joining t w o flange plates
be classified as W.
Type3.12.Thestressratioand effective weld sizecriteria of
this clause are intended to aid in the exclusion of premature
failures by local crushing of the concrete, by tearing of the
attached flange or in the body of the connector.
This type covers embedded shear connectors a t any
position along a girder. The reference to 'end connections' in
the title of table 17(c) refers to the end of the member in
which failure occurs; in thiscase the welded end of the
shear connector.
I - . .. . ..
Type3.73. single Sided manual metal arc procedures, with
or without the use of backing material, are not recommended Corner detail
unlessspecialist advice is sought. Weld metal failure need
not be considered. Figure=. Single fillet corner weld in bending

Licensed copy:UNIVERSITY OF NOTTINGHAM, 17/11/2008, Uncontrolled Copy, BSI



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

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

means electronic, photocopying, recording or otherwise without

which can be found on the inside front cover.Tel: 0181 996 9OOO; prior written permlssion from BSI.
Fax: 0181 996 7400.
This does not preclude the free use. in the course of implementing
BSI offers members an individual updating service called PLUS the standard,of necessary details such as symbols, and size, type
which ensures that wbsaibers automaticallyreceive the latest 01 grade designations. Ifthese detaik are to be used for any other
ediths of standards. purpose than implementationthen the prior written permissionof
Buylng standards BSI must be obtained.
Orders for all BSI. international and foreign standards publications If permission is granted, the terms may include royalty payments or
should be addressed to Customer Services. Tel: 0181 996 7000: a licensing agreement. Details and advice can be obtained from
Fax: 0181 996 7001. the Copyright Manager. Tel: 0181 996 7070.

Committees responsible for this British Standard

The Civil Engineering and Building Structures Standards Committee, Department o f the Environment (Transport and Road Research
under whose direction this British Standard war prepared, consists Laboratory1
o f representatives from the following Government departments and Department of Transport
scientific and industrial organizations: Federation of Civil Engineering Contractors
Association of Consulting Engineers Greater London Council
Association of County Councils Institution of Civil Engineers
British Constructional Steelwork Association Institution of Highway Engineers
British Precart Concrete Federation Ltd. Institution of Municipal Engineers
British Railways Board Institution of Structural Engineers
British Steel Industry London Transport Executive
Cement and Concrete Association Ministry of Defence
Concrete Society Limited Sand and Gravel Association Lid.
Constructional Steel Research and Development Organization Scottish Development Department
Department of the Environment (Building Research Establishment) Welding Institute

Amendments issued since publication

Amd. No. I Dateof issue I Text affected

9352 I March 1999 I Indicated by a sideline

British Standards InstiMion a 3 8 9 Chiswick High Road London W4 4AL Telephone 0181 996 9OOO Fax 0181 996 7400