Live Load Deflection Limits For Australian Road Bridges: Vic Nechvoglod Mieaust Cpeng and Frank
Live Load Deflection Limits For Australian Road Bridges: Vic Nechvoglod Mieaust Cpeng and Frank
Live Load Deflection Limits For Australian Road Bridges: Vic Nechvoglod Mieaust Cpeng and Frank
SYNOPSIS
Calculated live load deflection limits have been specified in Australian national bridge design
codes from the first 1953 code to the current Australian Bridge Design Code (1996).
This paper gives the results of a literature search carried out to find the origin and the
reasons for these calculated live load deflection limits. It appears that these were first specified
in U.S.A. codes and adopted in the first and subsequent Australian codes.
It was found that the calculated live load deflection limits were specified essentially for
steel bridges in the 1930's in an attempt to avoid undesirable psychological reaction by
pedestrians to bridge vibrations under road traffic and that the limits were not imposed for
safety or structural serviceability reasons.
It is suggested that the calculated live load deflection limits for both steel and concrete
road bridges should be removed from the Australian Bridge Design Code since their
application does not ensure avoidance of the undesirable psychological reaction by pedestrians
to bridge vibrations and leads to significant cost increase. It is also suggested that the
Australian code provisions for the dynamic response of road bridges carrying substantial
pedestrian traffic and for pedestrian bridges be reviewed.
1.1 This paper has involved an extensive search of references and review of codes undertaken
to find the origin of the calculated design live load deflection limits for road bridges. Some of
the references go back to the early part of the last century, most are out of print and others
have been difficult to obtain. Accordingly, extracts of relevant references are included in this
paper, however, not all references examined are included.
1.2 The paper is specifically directed to calculated design live load deflection limits but
includes brief comment on closely related issues such as design loads, impact (now called
dynamic load allowance) and vibration for road bridges with pedestrian footpaths. Each of
these areas warrant separate papers, but are not discussed here except to the extent considered
necessary to make clear their relation to the calculated design live load deflection limits for
road bridges.
1.3 The 1996 Australian Bridge Design Code (Ref. 1) live load deflection limits are given in
clause 6.3.3.2 for steel and composite construction bridges, and clause 5.2.4.2 for concrete
bridges, see extracts in Appendix Al which includes the related commentary clauses. Note
that for both steel and concrete the limit is span/800 for the calculated live load deflection and
that clause 5.8.5.1 gives "should normally comply" span to depth ratios for prestressed
105
concrete bridges, but these are not given for steel and composite construction bridges. The
commentary clauses provide little help in explaining the origin and reason for the calculated
live load deflection limit, or of the span/depth limits. These live load deflection limits existed in
previous Australian codes. A brief history is given below.
2.1 A succinct history of the development of past Australian codes was given by Mr. G. Marsh
in his introduction to the National Association of State Road Authorities (NAASRA) Bridge
Design Seminar held in September 1977(Ref.2). This is provided as an extract in the attached
Appendix A2. From this it is apparent that the NAASRA codes were based on the American
Association of State Highway Officials (AASHO) Standard Specifications for Highway
Bridges ( codes), starting with the 1949 AASHO code, which was used as the basis for the
first 1953 Australian national code. Subsequent Australian codes in 1958, 1965 and 1970
were also based on AASHO codes. The 1976 NAASRA code was the first Australian code to
differ significantly from the AASHO codes of the day, including an increase of some 33% in
the design loads over the AASHO code. It is worth noting that the 1976 NAASRA
code(Ref.3) is the allowable stress predecessor to the first limit states 1996 Australian Bridge
Design Code(Ref 1).
2.2 The live load deflection clauses for the AASHO 1949, 1957 and 1961 codes were selected
for particular examination because:-
• the 1949 AASHO code was used as the basis for the first 1953 Australian national code
• the 1957 AASHO marks the introduction of the design of steel composite bridges.
• the 1961 AASHO marks the introduction of the design of prestressed concrete bridges.
The AASHO 1949, 1957 and 1961 codes are identically worded for steel non-composite
beams with only the clause numbers changed. The 1957 and 1961 AASHO include a specific
requirement for the design of steel composite beams given as clause 1.9.6 in both codes. The
1949 AASHO 'Deflection clause 3.6.10" and 'Depth Ratios clause 3.6.11" as well as the
" Deflection clause 1.9.6" for composite beams in the 1949,1957 and the 1961 AASHO codes
are attached as extracts in Appendix A3.
It is apparent that the deflection limits which existed in the 1949 AASHO code could not have
been provided for the design of steel composite or prestressed concrete bridges. i.e. the limits
were for steel bridges only.
An interim AASHO specification issued in 1961 deleted from clause 1.6.11(this being
identical with clause 3.6.11 of the 1949 AASHO), the last sentence "If depths less than these
are used, the sections shall be so increased that the maximum deflection will be not greater
than if the ratios had not been exceeded'. This wording however was still in use in the 1970
NAASRA code (see Appendix A4). This wording was finally removed in 1972 by an
amendment (see Appendix A5 ) which also reworded the deflection requirement for the
structural steel design section to include a span/1000 calculated live load deflection limit for
"bridges carrying footpaths in urban areas". This limit is in stark contrast to the generous
span/300 limit for prestressed concrete for that period, see 2.3 below.
2.3 For prestressed concrete construction, the 1965 NAASRA code had a mandatory live load
deflection limit given in clause 8.5(i) as span/300, as well as a span-depth ratio limit of 30 (see
Appendix A6 ). The span/300 live load deflection limit was changed to span/800 in the 1970
NAASRA code (see Appendix A7).
106
It would thus appear that the span/800 calculated live load deflection limit originally
introduced for steel bridges in the first Australian national 1953 NAASRA code was not made
a mandatory requirement for prestressed concrete bridges until the 1970 NAASRA code.
In contrast, the position was reversed by the 1976 NAASRA code when the calculated live
load deflection limits were deleted from the structural steel design section of the code(see
Appendix A8), but retained in the prestressed concrete design section (see Appendix A9).
2.4 Examination of the 1936 Country Roads Board Specification ( Ref 4,see extract in
Appendix A10) for the design of Road Bridges for the state of Victoria reveals that the
span/800 limit is not present in that specification, which only includes limits on the span/depth
ratios. Similar provisions for span/depth ratios existed in USA Railway bridge design
specifications by the end of the nineteenth century (Ref.5) It appears that the span/800
deflection limit was introduced into Australian codes between about 1935 and 1949. This was
the case, see below.
3.1 The span/800 live load deflection provision first appeared in the 1941 AASHO and the
span/depth limit in the "Conference Specifications for Steel Highway Bridges", dated 1929,
compiled by a joint committee composed of representatives of the AASHO and AREA" (See
Ref. 6, page 1320-9.)
3.2 The origin of the deflection limits is given in the 1958 'Deflection Limitations of Bridges"
report by a Committee on Deflection Limitations of Bridges of the Structural Division of the
American Society of Civil Engineers. ( Ref 7, including Discussion by Fritz Leonhardt and
Octave W. Imer, part extracts of which are given below.)
According to the 'Deflection Limitations of Bridges" report at least one railway specification,
the 1871 Phoenix Bridge Company, limited deflection from the "passage of a train and
locomotive at 30 miles per hour to 1/1200 of the span". Other railway specifications of that
time contained depth/span ratios, typically 1/10 to 1/12 for trusses and plate girders(Ref.8).
Early road bridge specifications followed, with minor modifications, the lead of railways in
regards to span/depth limitations. There is some indication that these depth/span ratios were
deemed necessary to produce "economical depths", perhaps based on long practical experience
of the in service performance and deflection behaviour of steel railway bridges. Books of that
time give such ratios as design guides for "economical depth"(Ref.8, Ref.9 and Ref. 10).
107
extended to limit the deflection for cantilever arms to 1/300 of the arm "except at
locations of dense traffic where the length of cantilever arm devided by 400 shall be
used "
These deflection limits were discussed and adopted in meetings of the AASHO Bridge
Committee in 1938. The present requirements were first used in the 1941 AASHO
specification. However, certain state highway departments limit the live-load-plus impact
deflection to 1/1100 of the span."
3.4 The "Deflection Limitations of Bridges" report discusses a number of issues. The
following are considered relevant to any current review of live load deflection limits.
• It appears that the live load deflection limits were developed as an attempt to avoid
objectionable psychological reactions by pedestrians to bridge vibrations, and by
passengers in vehicles when the vehicle is at rest on the span.
• The limits were determined based on a limited statistical study made to correlate the
reported objectionable vibrations of a bridge with the a property of the bridge involved.
The property chosen for this correlation was the "computed' live load deflection, this
being selected as a property that could be conveniently determined at the design stage.
• The live load deflection limits correlation with vibration was for steel bridges. It did not
include composite steel bridges and prestressed concrete bridges, these forms of
construction being developed at a later date. To quote from the 'Deflection Limitations of
Bridges" report, "when the live load limitations where first proposed, the standard floor
was plank, the supporting members were either simple beams, pony trusses or pin-
connected through trusses ...".
• It was known in the 1930's, and is the case today, that actual deflections for bridges are
generally less than calculated values for a number of reasons. This is markedly so for non
composite steel girder bridges, as indicated by many field measurement including those by
Foster And Oehler (Ref.11 ),who state:- "This rolled beam bridge is much stiffer than
assumed in design. When averaged for the six spans, the observed deflections is only 16
percent of the calculated deflection for a given load " The figure is a little higher for steel
composite girders, being reported to be about 40 percent. Furthermore, in the same paper
it is reported that there is little difference in the field measured vibration behaviour
between adjacent spans of the same bridge, despite one having a calculated design live load
deflection ratio of 1/790 and the other of 1/1880. This lack of accuracy in predicting levels
of bridge vibration has been observed in field tests by others. Wright and Walker(Ref 12)
state:- "No systematic variation in dynamic increment was apparent for span to depth
ratios of between 5 and 20, and for span to design live load ratios of between 1/200 and
1/4000." It would thus appear that the span/800 design live load deflection limit has very
poor accuracy in predicting acceptable bridge vibration behaviour. This not surprising. One
should bear in mind that live load deflection "computed" in the early 1930's could very
likely differ from the same deflection calculated now, and that the statistical study and
calibration was for bridges, design loads, traffic, other conditions and factors existing in the
1930's which bear little resemblance to those of today. Many of these have changed
drastically as indicated below.
• The design loading of the 1930's was the AASHO H-20-35 loading exemplified by a train
of 15 tonne trucks. This loading, depending on span and type of bridge, is less than half of
the 44 tonne truck loading of the 1976 NAASRA eode(Ref 3), and between about one
third and one fifth of the 160 tonne SM1600 design loading of the current Australian
Bridge Design Code(Ref 1).
108
3.5 The "Deflection Limitations of Bridges" report states that "it was frequently observed that
the maximum apparent oscillation occurred with the passage of medium weight rather than
heavier vehicles." This observation indicates that undesirable vibrations can and are set up by
individual vehicles, rather than by the full design "truck train" loading, such as the H-20-35
design loading used in the 1930's. This vibration behaviour under less than the full design load
has been was observed in field tests by others including Foster and Oehler (Ref 11) who state:-
"During normal load tests, the total load of trucks varied from 5.3 to 75.3 kips, but these
vehicles also varied in axle spacing and speed. Thus the effect of total load appears to be
masked by the influence of other variables. The two trucks causing the most significant
vibrations on Spans 1, 2 and 3 had total loads of only 19.5 and 18.5 kips while the largest
total load of any truck tested on these spans was 75.3 kips." It would thus appear that
correlation of the vibration behaviour with the full design live load deflection, calculated or
otherwise, does not necessarily provide an adequate prediction of vibration behaviour in
service. The complexity and factors involved in the bridge/vehicle interaction were recognised
in the 1930's.
These aspects of bridge design, namely vibration and impact, are not the specific subject of
this paper. It is however worth noting that vibration and dynamic loading, or impact as it was
called in the past, has been a difficult area in bridge design for over 100 years. Waddell (Ref. 9,
page 120) had written in a paper on "Some Disputed Points in Railway Bridge Designing"
published in the Transactions of the American Society of Civil Engineers 1892, in regards to
impact loads:- "In respect to this subject it appears that we are all at sea; and we are liable to
remain there until such time as much needed experiments on actual working
stresses... Meanwhile we shall have to jog along in the best way we can, letting each engineer
use his own judgment concerning the intensities to employ.... until we have some real
knowledge of the effects of dynamically applied loads.."
Compare this with the statement on the design of road bridges by Paultre, Chaallal and Proulx
(Ref.13, page 260) in 1992":- "The dynamic amplification factor (DAF) is an important
parameter in the design of highway bridges and yet no worldwide consensus has been
reached so far as to its value. Some disagreement exists between provisions of various
national codes. This is because the DAF depends, in addition to the maximum span or natural
frequency, on many other parameters that are difficult to take into account with reasonable
accuracy. Vehicle speed , weight, and dynamic characteristics, the state of the structure,
roadway roughness, expansion joints, the type of bridge supports, soil-structure interaction,
and influence of secondary elements are some of the aspects influencing the DAF".
3.6 A potential reason put forward by some for the need for deflection limits is to limit
deformations to prevent damage to the bridge or its elements. The "Deflection Limitations of
Bridges" report stated that:-
• "With the exception of a few cases of failure in stringer connections and cracking of
concrete floors at points of negative moment, there were no reports of structural
damage."
It would appear that any damage reported was the result of lack of detailing to accommodate
deformations, which is designer dependent. The discussion by F. Leonhardt (Ref.7(a),on page
1787-67) is relevant in this regard. He states in part:-
"In 1946, the writer designed the first extremely slender steel bridge across the Rhine from
Cologne to Deutz with a depth of only 3.3m over a large part of the main span of 184m i.e. a
depth/span ratio of 1/56. The writer made this design with the clear knowledge that neither
limitation of the span/depth ratio nor the deflection is a criterion for safety or the behaviour
of the bridges if the factor of safety against the loads is sufficient and if the influence of
109
possible stress oscillations due to vibrations on the ultimate strength is considered correctly.
Of course the design must be such that all deformations will be possible without damaging
any members, especially those of the floor system" and concludes "With these
experiences, in Germany the depth/span ratio is no longer a design criterion for highway
bridges".
3.7 As well as its effects on live load deflection, the adoption in Dec.1998 of the 160 tonne
SM1600 design loading for the Australian Bridge Design Code(Ref.1) involves an increase of
serviceability design loading of between 1.5 and 2.5 times the previous 44 tonne T44
truck/lane loading, as reported by Rapattoni and Nechvoglod (Ref. 14). The Authors consider
that this could have important ramifications in regards to serviceability issues involving
deterioration due to excessive serviceability stresses in steel and concrete elements. These
aspects have been recognised by others, including Sergeev and Pressley(Ref. 15) who have
concluded that the artificial design live load deflection limit should be removed, and that the
serviceability requirements of the Australian Bridge Design code(Ref, 1) be reviewed.
5 OTHER CODES
5.1 The 1994 AASHTO LRFD Bridge Design Specification(Ref 16) has removed the
span/800 deflection limit as a mandatory requirement.
5.2 The Eurocode 1 : Part 3 - Traffic Loads on Btidges(Ref.17 ) has no specific live load
deflection limit requirements for road bridges.
5.3 The Eurocode 4 : Part 2 - Composite Bridges(Ref.18 ) has no specific live load deflection
limit requirements for road bridges, it does however give serviceability stress limits
Eurocode 3: Design of steel structures - Part 2: Steel bridges(Ref.19) likewise has no
specific live load deflection limit requirements for road bridges.
5.4 The 1991 Ontario Highway Bridge Design Code(Ref 20) has no specific live load
deflection limit requirements. It has a separate vibration limit state to cater for bridges with
provision for pedestrians.
6 CONCLUSIONS
6.1 The review of the historical background of the current Australian code span/800
calculated live load deflection limit shows that its origin is based on a limited statistical study
carried out in the U.S.A. in the 1930's. The live load deflection limits were developed as an
attempt to avoid objectionable psychological reactions by pedestrians to bridge vibrations,
and by passengers in vehicles when the vehicle is at rest on the span.
6.2 The span/800 limit was "calibrated" for conditions and structures of the 1930's using
much lower design live loads than the 44 tonne truck or the new 160 tonne SM1600 design
110
loading of the Australian Bridge Design Code. The "calibration" was essentially for steel
bridges only of types and detail that are no longer constructed, moreover, it did not include
steel composite and prestressed bridges which did not exist in the 1930's but are commonly
constructed now.
6.3 The span/800 limit has not been a constant value in design application. Examination of
past Australian codes indicates that its value was for:-
6.4 There are no safety or structural serviceability grounds for the span/800 calculated live
load deflection limit for road bridges.
6.5 History shows that vibration and dynamic loading, or impact as it was called in the past,has
been a difficult area in bridge design for over 100 years and has defied consensus and
agreement between the provisions of various national codes in regards to this area. This is
considered unlikely to change in the near future, for reasons given below.
6.6 Experience and field tests indicate that the span/800 calculated design live load deflection
limit does not ensure that bridge vibrations objectionable to pedestrians are avoided, for a
variety of reasons. One aspect that appears to have been underestimated is that the pedestrian
response to bridge vibration involves an individual's psychological reaction. The authors
conclude that it is difficult, if not impossible, to combine this psychological aspect with the
multitude of inanimate physical variables inherent to the vibration of road bridges, in order to
produce an acceptable provision for practical design use in the general case. This is because the
psychological reaction of an individual and the acceptance of levels of deflection and vibration
of a bridge is probably strongly influenced and will vary depending on that individual's
perception of the "solidity" of that particular bridge. What is more, tolerance to bridge
vibrations is a subjective human response not directly measurable. As stated by Wright and
Walker (Ref.12):- "People report vibrations to be perceptible, unpleasant, or intolerable; the
same vibration may elicit widely varying reactions among different subjects". Having said this,
it appears that road bridges are perceived by many members of the public as solid structures
that should not vibrate. It is suggested that, since safety is not an issue, it may more cost
effective to "educate" the public and allay the concerns of individuals that may be particularly
sensitive to bridge vibrations, than to limit vibrations by design to values that would be
acceptable to all under all traffic conditions. This could be by appropriate signs such as "This
bridge will vibrate safely under some traffic conditions"
7 RECOMMENDATIONS
It is recommended that:-
7.1 The span/800 live load deflection limit be deleted from the 1996 Australian Bridge Design
Code (Ref.1) for road bridges. It is ineffective, and its application results in significant cost
increase as shown by the authors in another paper (Ref.14).
111
7.2 Pending a review of requirements specifically for pedestrian use bridges based on the
current state of knowledge and experience in this area, the provisions for the dynamic response
for road bridges carrying substantial pedestrian traffic and for pedestrian bridges be kept
temporarily as set out currently in the 1996 Australian Bridge Design Code(Ref.1). In
developing any new provision, consideration should be given to the psychological aspects
of the problem of bridge vibration as perceived by pedestrians and the need to "educate"
the public and to allay the concerns of individuals that may be particularly sensitive to
bridge vibration. This could be by appropriate signs such as "This bridge will vibrate safely
under some traffic conditions"
8 REFERENCES
112
13. PAULTRE, P, CHAALLAL, 0 and PROULX, J. "Bridge dynamics and dynamic
amplification factors - a review of analytical and experimental findings". Canadian Journal
of Civil Engineering, Vol. 19 ,1992.
14 RAPATTONI. F. and NECHVOGLOD. V. "Future "Super Heavy" Trucks? Bridging
Uncertainty with Upgradeable Steel Bridges". AUSTROADS 4th Bridge Engineering
Conference, Adelaide, Australia, December 2000.
15. SERGEEV. S. N. and PRESSLEY. J. S. "Towards Rational Serviceability Criteria in the
Australian Bridge Design Code". Current and Future Trends in Design, Construction and
Maintenance page 167-178, Thomas Telford. 1999.
16. AASHTO LRFD BRIDGE DESIGN SPECIFICATIONS, SI Units, First Edition, 1994.
American Association of State Highway and Transportation Officials, Washington, USA.
17. EUROCODE 1. ENV 1991-3, Basis of Design and Actions on Structures - Part 3: Traffic
Loads on Bridges, European Committee for Standardization (CEN), English Version
March 1995.
18. EUROCODE 4. ENV 1994-2, Design of composite steel and concrete Structures - Part 2:
Composite bridges. European Committee for Standardization (CEN), English Version
December 1997.
19. EUROCODE 3: ENV 1993-2, Design of steel Structures - Part 2: Steel bridges. European
Committee for Standardization (CEN), English Version, October 1997.
20. ONTARIO HIGHWAY BRIDGE DESIGN CODE, 3"I. Edition,1991,Ontario, Canada.
21. COSRA, Conference of State Road Authorities of Australia, Highway Bridge Design
Specification 1953, First Edition, clause 3.6.10 and 3.6.11, page 67-68.
22. COSRA, Conference of State Road Authorities of Australia, Highway Bridge Design
Specification 1953, Second Edition, clause 3.6.10 and 3.6.11, page 54-55.
113
APPENDIX
Unless further investigation is undertaken, the deflection of road bridge girders under
live load plus dynamic load allowance shall not exceed 1/800 for spans and 1/400 for
cantilevers. This requirement need not be applied to HLP loadings. The deflection of
railway underbridge girders under live load plus dynamic load allowance shall not
exceed 1/640 for spans and 1/320 for cantilevers.
The Commentary Clause C6.3.3.2 is:-
C6.3.3.2 Deflection Limits for Beams.
Deflection limit calculations should use gross section properties, and ignore the effect
of bolt and other similar holes, which need to be taken into consideration in strength
calculations.
114
5.2.4.2 Deflection limits for beams and slabs.
The deflection of beams and slabs under service conditions shall be controlled as
follows:-
(a) To ensure that deflections do not detract from the appearance of the structure, the
deflections under the permanent effects given in Section 2 shall be such that the hog
does not exceed 1/300 of the span and sag should not occur.
(b) A limit for the calculated deflection of the member shall be chosen appropriate to
the structure and its intended use. Unless further investigation is undertaken the
deflection under live load plus impact shall not exceed 1/800 of the span or 1/400
of the cantilever projection as applicable.
(c) The member shall be designed so that, under the design load for serviceability given
in Section 2, the deflections calculated in accordance with Article 5.8.5 or 5.9.3 as
appropriate, do not exceed 1/250 of the span.
The Commentary Clause C4.2.4.2 is:-
C5.2.4.2 Deflection limits for beams and slabs
The total deflection is measured from the as-cast position but does not provide specific
guidance on the treatment of camber which could be used to eliminate the effect of part
or all of the total deflection and possibly permit slender members, particularly for
longer spans. If camber is used to significantly reduce the stiffness of the member then
care should be taken to check incremental deflection, support rotations and the
possibility of excessive vibration. Details of the methods of calculating deflections are
set out in Article 5.8.5, Deflection of Beams, and Article 5.9.3, Deflection of Slabs.
115
parameters can lead to a overly conservative design. To design effectively for
serviceability, the designer must have an understanding of the non-linear behaviour of
concrete structures.
C5.8.5.1 General
Span to depth ratios are given as a guide to reduce the likelihood of excessive
deflections of beams. However, the use of these ratios without a critical assessment of
the variables used may not eliminate serviceability problems.
After the formation of the various State Road Authorities, generally in the late 1920's, there
was no comprehensive bridge design standard in use in Australia. However, the States were in
fairly close contact through the Conference of State Road Authorities of Australia (COSRA)
established in 1934 and before that for a few years through meetings of State Ministers
supported by technical staff.
Each State was thus fairly well aware of what the other was doing in regard to all important
questions such as design loads. The major States had compiled various design criteria and
design aids which were published in various forms, and all States tended to used at least some
of these publications in addition to SAA design codes and standards. Well known standards
from other countries such as BS 153 for steel bridges were widely used.
In 1944 AASHO, the American Association of State Highway Officials revised its load design
standard and shortly afterwards most Australian States upgraded their own design loads and
adopted the heaviest of the AASHO design loads, viz the HS20-44 load. This obviously was
thought to be more appropriate to Australian conditions than for instance, the heavier BS
bridge design load.
At the same time it gradually dawned on the State Road Authorities that it was desirable than
uniform design standards be adopted nationally, and as a consequence COSRA arranged for
the State Bridge Engineers to meet in Melbourne in 1950 to prepare such a standard. This was
the first meeting of the committee, now known as the NAASRA Bridge Engineering
Committee, which has held meetings at least annually ever since.
This committee produced the first NAASRA (then COSRA) Highway Bridge Design
Specification in 1953. This was based on the 1949 AASHO Highway Bridge Design
Specification. It is understood that this was selected because it was the only comprehensive
design specification written in English that was readily available.Revised editions of the
NAASRA Highway Bridge Design Specification appeared in 1958, 1965, 1970 and of the
retitled NAASRA Bridge Design Specification in 1976.( The 1958 edition was in fact
published as the second COSRA edition- authors note).
Until 1965 the revisions were mainly prompted by revisions in the AASHO Highway Bridge
Design Specification, but after this there was increasing disenchantment with many features of
the NAASRA Specification. The 1970 edition was little more than a stop gap edition
necessary to correct some anomalies pending a comprehensive re-examination of the whole
specification which culminated in the publication of the 1976 edition.
116
A3 1949 AASHO - Structural Steel Design ( 1953,1958 COSRA identical, Ref 21,22)
3.6.10 Deflection.
The term "deflection" as used herein shall be the deflection computed in accordance with the
assumption madc for loading when computing the stress in the member.
Steel beams or girders having simple or continuous spans shall be designed so that the
deflection due to live load plus impact shall not exceed 1/800 of the span, the span length being
considered the distance center to center of bearings.
The deflection of cantilever arms due to live load plus impact shall be limited to 1/300 of the
cantilever arm.
When bridges have cross bracing or diaphragms sufficient in depth and strength to insure
lateral distribution of loads, the deflection may be computed for the standard loading,
considering all beams or girders as acting together and having equal deflection.
Sidewalk live load may be neglected in computing deflection.
The moment of inertia of the gross cross-sectional area shall be used for computing the
deflections of beams and girders.
The gross area of each member shall be used in computing the deflections of trusses. When
perforated cover plates are used in truss members, the effective area shall be the net volume
divided by the length from center to center of perforations.
If depths less than these are used, the sections shall be so increased that the maximum
deflection will be not greater than if these ratios had not been exceeded.
If depths less than these are used, the sections shall be so increased that the maximum
deflection will be not greater than if these ratios had not been exceeded.
117
A4 1970 NAASRA - Structural Steel
In cantilever bridges the corresponding ratios for cantilever spans, based on the cantilever span
lengths shall be 1/300 and 1/375.
118
A7 1970 NAASRA - Prestressed concrete
4 (i) The previous mandatory span/depth ratios and live load deflection rules have been
deleted. However, where significant numbers of pedestrians are likely to use a bridge, the
designer is required to check for undesirable vibration effects.
A10 1936 Country Roads Board Code for the state of Victoria.
The ratio of depth to span for the various types shall be preferably not less than the ordinary
values and shall be in no case less than the limiting values shown in Table 13, unless sufficient
extra metal is used in web and flanges of joists and girders to reduce the deflection to that
which would occur with the limiting values, or unless the structure is continuous over
supports, when the depths may be reduced to two-thirds of the tabulated values.
Table 13. Ratios of Depth to Span
Type Ordinary Limiting
value value
Rolled joists 1/15 1/20
Plate girders 1/12 1/15
Trusses 1/8 1/10
Longitudinal stringers for trusses 1/12 1/15
119
120