You are on page 1of 6

Highwaybridgeloadings

Charles F. G a l a m b o s
Federal Highway Administration, Washington, D.C. 20590, USA

The paper presents a comparison of the A A S H T O design live Ioadings for


bridges with various other loading situations. Comparisons are made with
some European design live loads, with native and foreign legal loads, normal
permit overloads, and abnormal permit loads. The results of a bridge load
rating exercise are presented. Some actual bridge load histograms are given,
as well as a comprehensive histogram based on the national Ioadometer survey
for 1970. Fatigue Ioadings and damage are discussed in the light of actual
and design Ioadings.
It is concluded that it may be timely to increase the A A S H T 0 HS design
Ioadings. To improve the bridge load rating process, it is suggested that some
standard load rating vehicle and test method be employed. Further refinemenl
of the fatigue design provisions for steel bridges do not seem warranted in
light of the great variation of actual Ioadings on bridges.

Introduction The design live loads prescribed in the above specificationa


In the 1978 annual meetings of the American Association of are of two forms: (i), a uniform load per linear foot of load
State Highway and Transportation Officials (AASHTO) lane; and (ii), variations of axle loads in two standard trucks.
Bridge Subcommittee (there are four regional meetings), The uniform load is supplemented by one concentrated
there was considerable discussion again about the adequacy load (or two for continuous spans), which varies for moment
of the present design loadings, especially in view of con- or shear. Some of the axle spacings of the design truck can
tinued pressure by the trucking industry to ask for higher be varied (within limits) to produce maximum stress effects.
allowable loads, and also because of the recent increased The heaviest loadings are illustrated in Figure 1, For inter-
awareness of the state of deterioration of many of the state highway bridges, there is an alternate loading of two
bridges. It is estimated that over I00 000 of the 600 000 axles four feet apart, with each axle weighing 24 000 pounds.
highway bridges on all US road systems are structurally Whichever loading, lane load, truck load, or alternate loading
inadequate or obsolete for various reasons, and should be produces the maximum effect is the condition which should
replaced. Some of these bridges are already being replaced be used in any specific case.
through a special bridge replacement program and many These loadings, presently designated HS 20-44, have not
more will be replaced as a result of the funds made available been substantially changed since 1944.
by the recently passed 1978 Surface Transportation Act. It is of interest to compare the AASHTO design live
This act, for the next four years, makes available to the loads with those used in several other countries. This is
States 0.9, 1.1, 1.3 and 0.9 billion dollars expressly for shown graphically in Figure 2 for West Germany, the United
bridge replacement. It is therefore a most appropriate time Kingdom, Belgium, the Netherlands and the HS 20 AASHTO
to examine bridge load design practice and philosophies to loadings. The method of comparison is by means of bending
point out where changes for the better may be in order.
The views expressed here are, of necessity, those of a
researcher and not of a bridge design engineer, and they
express private opinions and not official Federal Highway
Administration policy. External loads from traffic only are
considered.

Design live loads


HS20-44 E~ 32k 32k
} 14' l 14' to 30' [
The great majority of highway bridges in the United States Stonderd truck toGding
have been, and still are designed according to the prevailing . d8k for moment
provisions of the 'Standard Specifications for Highway Concentroted toga- 26k for sheor
Bridges', l as adopted by the American Association of State l/Uniform LoQd64Oibper tineorfoot of toed tone
Highway and Transportation Officials. The latest published
full edition is the twelfth edition (1977). Selected interim
sections are issued as needed. Generally, the specifications Uniform loading
are republished every four years. Figure I AASHTObridge designloads

0141-0296/79/050230-06/$02.00
230 Eng. Struct., 1979, Vol. 1, October @ 1979 IPC Business Press
Highway bridge Ioadings: C. F. Galambos
5000 combination is shown in Table 1. The values shown do not
include the slight increase in loads allowed in some jurisdic-
4500 tions due to statutory enforcement tolerances.
The table shows that except for a few countries (Belgium,
4000 I, L ,I France, Italy and Spain), the allowable single axle and
tandem axle loads are not too much different from the
AASHTO allowables: 22 kips vs 20 kips, 35 kips vs 34 kips.
3500
The 5-axle vehicle allowable gross loads are somewhat
Z greater in Europe, an average of 88 vs 76 kips, or about a
3000
No,h,.oo0, 15% increase over the AASHTO allowable values.
If Table I is now compared with Figure 2, it is clear that
25OO in the USA bridges are designed to carry a slightly smaller
load than they are allowed to carry: 72 kips vs 76 kips,
2OOO whereas in most European countries the design loads are
much higher than the allowable loads. Although the above
15OO statement is true, it is of course quite a simple statement.
Further discussion follows.
IOO0o l
,o 3'0 40
' sb
I
70
I
80 q~O 100 A l l o w a b l e o v e r l o a d s or p e r m i t loads
L,(m)
Figure 2 Comparison of design loads
Occasionally, it happens that it is desirable, or necessary in
the case of emergencies, to move a load that is heavier than
the legal loads discussed above. The question then arises as
moment calculations, taking account of impact factors and to how heavy such a load can be and still not damage a
the respective allowable stress used in each country. A simply specific bridge. This then brings us to the whole complicated
supported bridge was used, and the design vehicles were question of the actual load rating of bridges. At this point,
approximated by a conversion to a uniform distributed it is interesting to note that bridge engineers all over the
load spread out over a distance of 10 metres. The calculated world like to distinguish between 'normal overloads' and
values are only of theoretical interest; they cannot be con- 'exceptional overloads'. In the USA there is the variation of
sidered as practical allowable values for the spans considered, the above in terms of stresses that translate into 'inventory'
but they are a good indication of the relative ultimate rating, and 'operating' rating. The reasoning behind the two
capacity of the structure. No lateral load reduction for more ratings is the same - because of a certain conservatism in
than one loaded lane was allowed for the comparison. design, construction, material properties, and the fact that
What is immediately obvious from Figure 2 is that there it is highly unlikely that two or more heavy vehicles should
is considerable variation between countries, and that by far precisely meet at a critical section of a specific bridge, it is
the lowest design loading is the AASHTO loading. Another said that a certain amount of overloading will not harm the
way of saying this is that these other countries build more bridges. (We are for the moment leaving out the possibility
conservatism into their bridges, since the design loads are of increased fatigue damage from the discussion.) In the US,
very heavy compared with the AASHTO loads. In Europe, this is based on a stress below 0.55 o yield. In several of the
the design philosophies differ somewhat from those in the European countries, a specific weight has been assigned as
USA, in that the Europeans tend to design for longer life the upper limit of such 'normal', unrestricted overloads.
(120 years in the UK), and they worry more about being Some of these weights are tabulated in Table 2. Total loads,
able to transfer heavy military loads. They also wish to be or gross vehicle loads are shown, but most of the countries
able to occasionally transfer extremely heavy peace-time also have axle load limitations, and one must have a permit
loads (power plant generators, etc.) without unduly to operate such vehicles; some of the permits are for single
endangering the structural integrity of their bridges. passages only, but many of them (such as for construction
equipment) are for a season or a year at a time. Usually there
is no restriction placed on speeds of travel and the vehicles
Legal l o a d s
mix in with the usual traffic stream, and no damage is
Closely related to the above discussion on design loads are thought to accrue from an unlimited number of such
the legally allowed loads. These vary somewhat from State passages. Note that several of the values shown in Table 2
to State in the United States, and also differ a great deal in are also the legal lihaits for a 5-axle vehicle.
the European countries. A tabulation of axle loads (single The 'exceptional' overloads are only allowed occasionally,
and tandem), and gross loads for a 5-axle tractor-trailer and are more carefully controlled. Usually, they have to have

Table 1 Maximum legal weight of freight vehicles (in kips)

>=
r-

E ~ ~ i,. - := = -c I- ~
" -- ¢ (3 *-, t~tJ3 ~"

-- ~ E)<C I --J

Single axle 29 22 22 22 29 22 26 22 22 29 22 22 22 20 24 18
Tandem axle 44 44 35 35 46 35 42 35 35 46 35 31 45 34 40 29
Gross 5-axle 84 110 97 79 84 84 97 97 86 84 - - 72 76 80 71

Eng. Struct., 1979, Vol. 1, October 231


Highway, bridge Ioadings: C. F. Galambos

Table 2 Allowable 'normal' overloads (in kips) is seen that the Canadian Standards Association is willing to
use almost 100% of the fully plastic moment of the bridge,
provided that the vehicle load path and speed are strictly
controlled, and that the vehicle is the only one on the span,
and that such a load happens very infrequently, and that
supposedly the bridge was built according to plan, that the
material properties are equal or stronger than was assumed
Total load 88 110 97 92 84 84 110 92 113 72
in the design, and that the bridge has been maintained to
stay brand new. Other jurisdictions proceed with somewhat
more conservatism.
Because as was just shown above, there can be such a
an escort, travel very slowly along prescribed paths across a
variation in even as simple and common a structure as a
bridge, and be the only vehicle on the span. Again, in the
70 ft beam and slab bridge, many engineers feel that the
US, the load allowed is derived from the allowable stress not
only way to really rate a bridge is to actually test it with a
beyond 0.75 a yield. This is called the operating stress. In
moving test load. Properly conducted load tests with realistic
other countries, careful stress calculations for 'exceptional'
vehicles can be of great help in assessing the load carrying
overloads are also made, although in Belgium and France, an
capacity of a structure. However, such tests must be well
actual upper gross load limit of 792 kips and 880 kips,
respectively, exist, along with certain axle load limitations. planned and executed, with strain and deflection gauges
It is recognized that monstrous loads of 800 kips take very placed at major points of interest. If this is done, the tests
special vehicles for safe transport across a structure. help to determine the actual lateral load distribution, the
effect of bracing members, deck behaviour, joint behaviour
(pinned, frozen, or intermittently free), the degree of
L o a d rating composite action, the amount of help received from curbs,
sidewalks, and railings, the impact factors, and the actual
It is desirable for a number of reasons to have a good vibration characteristics and other actual bridge behaviour.
inventory and the best estimate ef the present live load A word of caution must be inserted here to say that as
capacity of our highway bridges, whether we are talking useful as such load tests are, they do not tell everything
about a smaller local jurisdiction or even the entire about a structure. The extent of fatigue cracking, for
combined highway system of a nation. Rules and regulations 2 example, and the material properties such as crack growth
for compiling an inventory and for the procedure to be used rates and notch toughness, cannot be determined by load
for the load rating of bridges have been issued in the United testing. Nor can the usual load tests tell what the ultimate
States by the Federal Highway Administration for the carrying capacity of the structure might be.
Federal Aid Highway Systems, and it is not the purpose
here to dwell on details, except to point out that much of
the load rating process still comes down to something called
'engineering judgment'. Substantial differences in load-
carrying capacity can result for the same structure depending
on who looks at it, which calculation methods he uses, how
he treats items like lateral load distribution, impact, effect
of corrosion, settlement, scour, material strengths, and any
number of engineering and non-engineering parameters. I- 1;, -22,
To illustrate what variations in load ratings are possible, Figure 3 3S2 rating example vehicle. Axle weights: 1,8 kips;
an interesting experiment was performed recently by one of 2, 16 kips; 3, 16 kips; 4, 16 kips; 5, 16 kips
the committees of the Office of Economic Cooperation and
Development (OECD) a headquarters in Paris, France. The
assignment was as follows: having been given the design Table 3 Rating results for 70-ft span

drawings of a specific bridge with the material properties Canada


(yield strength, ultimate strength, etc.) and a specific CSA S6 Ontario UK Norway USA
vehicle, a 5-axle tractor-trailer combination shown in
Figure 3, each of the cooperating agencies was to assess the Unsupervised and
mixing 166 . . . .
live load that this vehicle could carry. Canada, Norway, the
Supervised and
province of Ontario, the United Kingdom, and the United only vehicle 350 . . . .
States, all submitted answers as to what they thought Readily available
prudent and reasonable. The bridge was a 70-foot simple permit - 198 - - -
Controlled special
span, non-composite steel beam and concrete slab structure,
permit - 304 - -- -
taken from 'Standard Plans for Highway Bridges - Steel Inventory working
Superstructures'. 4 The answers are presented in Table 3, stress -- -- 95 - 103
and they vary from a low gross load for the vehicle of Inventory load
95 kips to a high of 350 kips. factor . . . . 111
Operating working
Differences come about because of four general areas of stress -- -- 136 -- 168
variations: inventory ratings vs operating ratings, and working Operating load
stress vs load factor calculations. Other differences arise factor . . . . 184
from philosophical differences about impact factors, load W o r k i n g stress -- - - 270 --
Load factor - - - 306 -
distribution, live load-dead load ratios, and what fraction
G r o s s l o a d in
of the yield or fully plastic moment it seems reasonable to kips
use. All these variations are tabulated in Table 4. There it

232 Eng. Struct., 1979, Vol. 1, October


Highway bridge Ioadings: C. F. Galambos
Tab/e 4 Variations in load rating parameters

Working stress Load factor

Load multiple
Fraction of lane Capacity of Capacity of
per stringer Impact structure Live Dead structure

USA Inventory 0.712 0.256 0.55 fyS 5/3 1 0.8 fyS


Operating 0.712 0.256 0.75 fyS 1 1 0.8 fyS

Norway 0.560 0.400 0.636 fyS 1.3 1.2 0.870 fyS

Canada CSA S6 Design situation 0.586 0.256 - 2.35 1.400 0.703 fyZ
Unsupervised and
mixing 0.586 0.256 - 1.933 1.286 0.831 fyZ
Supervised and
only vehicle 0.686 0.256/3 -- 1.192 1.166 0.972 fyZ
Ontario Readily available
permit 0.630 0.45 - - 1.35 1.169 0.92 fyZ
Controlled special
permit 0.607 0.15 - 1.15 1.169 0.92 fyZ

fy, yield strength of steel; Z, plastic section modulus of stringer; S, elastic section modulus of stringer; CSA, Canadian Standards Association

On older bridges, where no design or construction plans the figures. The double peak of the gross weight histograms
exist, load tests can serve the most useful purpose in deter- seems to be typical of many American tests, showing in
mining whether the normal truck traffic can continue to use general empty tractor-trailer freight vehicles (25-30 000
the bridge, or whether some load restriction has to be pounds) and loaded ones (67-75 000 pounds). A composite
imposed. The bridge rating crew of the Province of Ontario, loading histogram made up of a number of studies nation-
Canada, does a great deal of load testing of the kind just wide for the year 1970 is shown in Figure 10. Note again
mentioned and they generally find that most bridges are the double peak.
stronger than one thinks they are. s A very legitimate observation is often made in connec-
tion with bridge loading history studies. And that is that the
Actual loads 9'0
In recent years, a great deal of actual weighing of vehicles
has taken place in the USA, much of it for route planning 72
purposes or for legal load enforcement, yet for the most
part, such truck weighing is not directly applicable to bridge
Oo
loading problems. We have therefore encouraged and partici-
pated in specific bridge loading history studies in which care x 5"4 r ~ ~
is taken to weigh every vehicle (above a certain relatively ag 3"6
minimum weight) crossing the structure in some representa-
tive time period. Usually, such studies also include the 18
collection of strains (stress ranges) in selected members of n
the structure at the same time. fl . . . . ITI'~ ~'{] 17
Several typical samples of bridge loading histograms are O 10 20 30 40 50 6'0 70 80
shown in Figures 4-9. Both axle load distributions and gross Gross vehicle weight, (kips)
vehicle weights are shown. The mean load, standard devia- Figure 5 Histogram for gross vehicle weight, all trucks, Shaffer
tion and number of cases weighed is also shown on some of Creek Bridge, 1968. N = 249; # = 40.5; a = 20.6; c.o.v. = 0.51

15 140
,2 rl
O
112
n
o ,I I
II 84
)
Z"
5t-
6 56
C)
£3

3f f
0 25 50 75 1(iO 12.5
Axle weight s (kips)
15O 17'-5 2 0 0
2.8

0 2:5 5 ) 7:5
rl
Axte weight, (kips)
•O 12'5 15-O 17'5 20-0

Figure 4 Histogram for axle weight, all axles, all trucks, Shaffer Figure 6 Histogram for axle weight, all axles, all trucks, C.B. & Q.
Creek Bridge, 1969. N = 3540; # = 8.4; o = 4.1 ; c.o.v. = 0.49 Bridge, 1969. N = 6751 ; # = 10.1 ; a = 3.9; c.o.v. = 0.39

Eng. Struct., 1979, V o l . 1, O c t o b e r 233


Highway bridge Ioadings: C. F. Galambos

frame into stiffer members. Fatigue cracks therefore grow


that are deflection related. The latter case seems to be a
design related problem.
~3 The fatigue damage caused on bridges, in a global sense,
is a function of the number of live load stress ranges applied
2 to the bridge. The stress ranges vary more or less linearly
Or"
with vehicle gross load for the main bending members of
bridges. For deck elements and floor beams, the wheel and
u_ 1 axle loads produce the stresses. Enough crack growth
information is available from laboratory fatigue tests and
from actual bridge damage observations on various cracked
0 10 30 20 40 50 60 7o 80
details, for a relationship between loading and damage
Gross vehicte weight, (kips)
usable for design purposes to be developed.
Figure 7 Histogram for gross vehicle weight, all trucks, C.B. & Q. One such relationship is shown in Figure 11. The
Bridge, 1969. N = 1482; ;~ = 46.0; a = 18.7; c.o.v. = 0.41
'damage factor' incorporates several relationships between
average daily truck traffic, desired design life, ratio of actual
22.~ vehicle to design vehicle weights, and Miner's hypothesis of
damage. A fuller description of the reasoning used is given
¢- 20 elsewhere, l° but the above was the basis for the latest
O TM fatigue provisions for the steel design portion of the
AASHTO Specifications.
15 In Figure 11, it is seen that by far the greatest amount of
E damage is done by the heavily loaded, but still legal vehicles
~ ~o of 60 to 80 kips gross loading. Figure 11 is based on the
composite truck weight histogram shown in Figure 10.
"6 5 It is not the purpose of this paper to dwell on fatigue
problems, except by way of coming around to the question
of what is the proper fatigue loading spectrum to be used
0 510 20 30 40 50 6 0 7 0 BO 90 100
Truck weights, (kips)
16
Figure 8 Truck-weight histogram constructed from east-bound
weighing data obtained from Westport (solid line) and adjusted values
for Bridgeport test site (broken line)
Y
weights obtained are not really representative of day-in,
day-out traffic, because the illegally overloaded truck
drivers soon learn about the tests and then take an alternate F --m
route to go around the test site. The only way to avoid this cr
is to monitor weights all the time, and some States are doing u_
this with built-in weighing stations. There is also considerable
research and testing going on worldwide on equipment and
schemes for weighing trucks in motion, most of them with
pavement platforms of one kind or another, but some of
them are also using the bridge as the weighing mechanism. 6-9 i
Many of the schemes are workable and in the next few 10 20 3O 40 50 60 70 80 90 1OO
years, much more attention will be paid to the monitoring Gross toed, (kips)
and enforcement of loads. Figure 9 Tractor-trailer gross weight distribution from locations in
Ohio. All MR locations 2328 semi-trailers

Fatigue
The interest in the actual loads on the highways is by no 50
means just academic. Loads produce stresses and stresses,
even small ones when repeated enough times, will cause a 4C
flaw to grow and possibly even cause the rupture of a
member, but at the least, will make it necessary to spend
=z 3.0
money for crack repairs. There is increasing evidence of
fatigue crack growth in American steel highway bridges.
Often, the problem originates in secondary members in
which a relatively large initial flaw exists, possibly due to
poor fabrication practice and careless shop inspection.
Several instances of the complete fracture of a major load
2.0

10
G
carrying welded plate girder have been experienced in
recent years, caused by initial flaws in gusset plates and
% 3'0 S'0
4 '0 60 70 80 90 100
horizontal and vertical stiffeners. An even greater number Gross vehicte weight (kips)
of fatigue problems come about when flexible members Figure 10 Composite gross vehicle weights, 1970

234 Eng. Struct., 1979, V o l . 1, O c t o b e r


Highway bridge Ioadings: C. F. Galambos

t.8 Based on these thoughts on highway loadings, the


following conclusions and recommendations are offered:
1.6 (I) It would seem that it is time to increase the AASHTO
HS design loadings. There is something disquieting about
1.4 having the lowest comparative design load in the industri-
i alized western world. However, such an upgrading of the
loading should not be done without a thorough economic
12
"t.
o study of the long-range consequences.
u (2) Although not treated in great detail in this paper, it is
o 1.O believed that considerable benefits in terms of longer
O
pavement and bridge life would result from a universal and
~ 0"8 strict enforcement of legal allowable loads. In this respect,
0 reference should again be made to the fatigue damage curve
06 o f Figure 11. This curve does not show the influence of
illegal loads.
(3) Considerable improvements are yet to be made in the

o4f
O2
process of the load rating of bridges. It is believed that some
standard load testing vehicle and scheme can be a very good
tool in the rating process, especially when used in con-
junction with bridge replacement priority settings.
0- (4) In light of the considerable variation in actual live loads
2O 30 40 50 60 70 80 90 10(D and the lack of enforcement of legal loads in some jurisdic-
Gross vchicNt weight, (kips) tions, it would seem counter-productive to further refine the
Figure 11 Probable damage caused by various truck weights fatigue live load design provisions for steel bridges. It may,
in fact, be more prudent in the long run to avoid fatigue
problems as much as possible and imitate the concrete
for the design of highway bridges. The AASHTO Bridge
provisions of the specifications.
Specifications have a definite difference in fatigue design
philosophy between concrete and steel bridges. A final note is added to the policy makers charged with
In concrete structures, the specification writers seem not the establishment of highway bridge design and allowable
to want to tolerate any hint of fatigue problems in the loads. The time is past when each mode of transportation
reinforcement in that the allowable live load design stress such as highways, railways and canals can independently set
range has been set high above (near 21 ksi, depending its own limits on sizes and weights. Transportation of goods
somewhat on minimum stress and reinforcement deforma- and people has become highly interchangeable between modes
tions) and damaging fatigue loading likely to be experienced and any changes in weight limits must be made with the
in the expected life of the structure. (This lbhilosophy seems coordination and cooperation of all the affected sectors of
to work, in that to the best of the author's knowledge, there transportation.
have been no instances of fatigue ruptured reinforcing bars
on highway bridges.) References
In the steel design portion of the specifications, the 1 'Standard Specification for Highway Bridges' (Twelfth edn),
elaborate table presented relates geometric details to 1977, American Association of State Highway and Transporta-
allowable stresses and to the repetitive number of these tion Officials, Washington, D.C.20004
allowable stresses, which were derived as shown above. 2 Manualfor Maintenance Inspection of Bridges (Second edn),
It is clearly evident that the steel fatigue design specifi- June 1974, American Association of State Highway and
Transportation Officials, Washington, D.C.
cations concerning allowable stresses are much closer to 3 'Evaluation of Load Carrying Capacity of Existing Road
what actual bridge members experience daily, when com- Bridges, Report of Committee CM-2', OECD, Paris, France.
pared to the reinforcement in concrete structures. For (To be published in 1979)
instance, the allowable fatigue stresses in concrete reinforce- 4 'Standard Plans for Highway Bridges Volume II, Structural
Steel Superstructures', Federal Highway Administration,
ment, when compared to those for steel members, would say Washington, D.C.20590, April 1968
that for over 2 000 000 cycles of loading, one should not 5 Bakht, B. and Csagoly, P. F. 'Testing of Perley Bridge, Ontario',
allow any detail lower than type B. It appears that some Ministry of Transportation and Communication, Research and
re-thinking of the AASHTO fatigue specifications may be in Development Division, Rep. RR 207, January 1977
order. 6 'Weighin Motion Instrumentation', Rep. No. FHWA-RD-78-81,
Moses, F. and Kriss, M., Federal Highway Administration,
Washington, D.C. 20590, June 1978
Summary and conclusions 7 FothergiU, J. W. et al. 'Feasibility of Utilizing Highway Bridges
to Weigh Vehicles in Motion', Volume I, Rep. No. FHWA-RD-
For the most part, this paper presents a comparison of the 75-33, Federal Highway Administration, Washington, D.C.,
AASHTO design live loadings with various other loading November 1974
situations. Comparisons are made with some European 8 Moses,F. and Goble, G. 'Feasibility of Utilizing Highway
Bridges to Weigh Vehicles in Motion', Volume lI, Report No.
design live loads with native and foreign legal loads, normal FHWA RD 75-34, Federal Highway Administration,
permit overloads, and abnormal permit loads. The results of Washington, D.C., October 1974
a bridge load rating exercise are presented. Some actual 9 Siegel,H. J. 'Feasibility of Utilizing Highway Bridges to Weigh
bridge load histograms are given, as well as a more compre- Vehicles in Motion', Volume III, Rep. No. FHWA RD 75-35,
Federal Highway Administration, Washington, D.C., November
hensive composite histogram from 1970. Fatigue loadings 1974
and damage are discussed in the light of actual and design 10 Fisher,J. W. 'Bridge Fatigue Guide - Design Details', American
loadings. Institute of Steel Construction, New York, N.Y. 10020

Eng. Struct., 1979, Vol. 1, October 235

You might also like