BRIDGING THE MILLENNIA

Recommendations for Proof Load Testing in Australia

Authors

ANGELA RANSOM

Angela Ransom graduated from the University of Queensland with a Bachelor of

Engineering in August 1990 and has been employed with the Queensland Department of
Main Roads since March 1990. During this time she has worked as both a Design
Engineer in Bridge Branch and as Site Engineer for construction of the New Burnett River
Bridge in Bundaberg. In 1993 Angela was awarded the Queensland Young Engineer of
the Year Award by J.EAust, and the Young Achiever Award (Career Achievement
Category) by DEFT and Channel Nine. In 1996 she was awarded a Science and
Technology Awareness Scholarship by IEAust. Angela used this scholarship to travel to
Canada and the USA to further her research into Proof Load Testing of Bridges and
related bridge technology. She is currently undertaking a part-time Master of Engineering
Degree at QUT.

DR. ROBERT HEYWOOD,

Dr. Robert Heywood is a Director (Research and Development) of Infratech Systems and
Services Pty Ltd and Principal Researcher at the Physical Infrastructure Centre, School of
Engineering, Queensland University of Technology. He graduated from the University of
Queensland in 1974 and undertook his MEngSc and PhD study at the same University. Dr
Heywood joined the QUT in 1985 after a decade of consulting experience in Australia and
overseas.

Dr Heywood is a recipient of the WH Warren Medal (Institution of Engineers, Australia)

and the Babcock Australia, Institution of Engineers Australia Centenary Scholarship.

Dr. Heywood's interests centre on improving the efficiency of transport by considering

heavy vehicles, bridges and roads as inter-related elements in the transport system. Recent
highlights include the development and application of instrumentation to assist in the
evaluation of bridges, the development of the next generation AUSTROADS Bridge
Design load, participation in the OECD DIVINE (dynamic interaction between bridges
and vehicles experiment) international expert group, identifying damaging features in road
profiles, and the development of bridge health monitoring technology.

231
 BRIDGING THE MILLENNIA

Recommendations for Proof Load Testing in Australia

AL Ransom, B.E., MIEAust, NPER-3

R.J.Heywood, BE (Bons), MEngSc, PhD, MIEAust, NPER-3

ABSTRACT

Throughout the world many countries are faced with the problem of ageing bridge
infrastructure that is being called upon to carry increasing loads. With the difficulties
associated with gaining funding to replace or rehabilitate these bridges, it is essential that
the most efficient use is made of the existing infrastructure. Many of these bridges have
load limits imposed on them which do not accurately reflect their capacity. Various
methods of assessing bridge capacity are adopted by road authorities throughout the
world. These forms of assessment include analytical rating, calibration of analytical
models by supplementary load testing and assessment by proof load testing.

Proof load testing is a valuable tool that can be used to accurately determine the capacity
of these bridges to carry the required loads. The procedures used in proof load testing do
not vary greatly between countries but the magnitude of the loads applied and the load
factors used to calculate a load rating vary significantly. This paper discussed the use of
proof load testing as an assessment tool and presents a case study of a proof load test in
Queensland.

232
 BRIDGING THE MILLENNIA

Recommendations for Proof Load Testing in Australia

INTRODUCTION

1. Proof load testing of bridges as a means of assessing bridge performance and load
capacity has received considerable attention in Australia in recent years. With the ever
present threat of increased legal loads, a means of assessing the capacity of Australia's
bridge infrastructure to withstand these loads is necessary. A mobile proof load testing
facility has been developed and has been used to evaluate a number of bridges in New
South Wales, Queensland and Victoria. This facility was predominantly based on the
system used by the Ministry of Transportation in Ontario, Canada which has been in
operation for the past 28 years. Proof testing of bridges is relatively new in Australia and
there is a need to develop a set of suitable guidelines if it is to be adopted as an asset
management tool.
2. A review of approaches to proof load testing used internationally has highlighted the
fact that the test methods used and the procedures for determining a bridge load rating
from test results vary significantly between countries.
3. This paper presents an overview of bridge load testing internationally including the
philosophies behind proof load testing. The procedures use to convert measured data to a
bridge rating are also outlined.
4. The load testing and evaluation of the South Pine River Bridge in Queensland is
presented as a case study. The bridge was one of Queensland's first steel and concrete
composite bridges constructed in 1941. The performance of the bridge under a proof load
is discussed and the rating of the bridge is presented using various rating methods adopted
internationally.

THE NEED FOR BRIDGE TESTING AND EVALUATION

5. Australia has an extensive road transport network that constitutes a substantial

investment. Bridges form a significant component of the value of the transport network
with approximately 40,000 bridges throughout the country (Austroads 1994a). The road
network is increasingly being called upon to carry heavier loads due to pressure to
improve transport productivity by increasing legal loads. Invariably, the load limits for the
network are controlled by the capacity of the bridges. Many in-service bridges on our
national road network were designed for significantly lower vehicle loads than the current
legal loads and many of these bridges have load restrictions imposed upon them due to
their age or condition. In many cases the load restrictions do not accurately reflect the
actual capacity of the structure. There is therefore a need for asset managers to have anrcss
to a reliable and accurate tool to determine the actual capacity of these bridges. This would
enable the most efficient use of the existing infrastructure without compromising safety
requirements (Heywood and McDonnell 1996).
6. In 1994, as a result of an AUSTROADS Structures and Technology Group strategic
initiative, a pilot proof load testing program was initiated in Australia (Austroads 1994b).

233
 BRIDGING THE MILLENNIA

The aim of this program was to adapt proof load testing methods used overseas, to suit
Australian conditions. The results of this program were then intended to be utilised in
developing a national bridge testing program. This project was a co-operative effort
between the state road authorities in Queensland, New South Wales and Victoria,
Queensland University of Technology and Blastronics. Pilot tests were conducted in New
South Wales, Queensland and Victoria.
7. The evaluation of bridges by proof load testing is not new but it is not widely used
throughout the world. The approaches to load testing adopted by road authorities
throughout the world varies. The facility developed in Australia was primarily based on
the methods used by the Ministry of Transportation in Ontario which has conducted
research and development in bridge testing since 1969. Proof load testing has also been
used to varying degrees in Switzerland, The United Kingdom, The United States of
America and New Zealand. The methods of testing, the loads applied and the method of
calculating a rated load from the test results varies from country to country.
8. Most of the countries investigated have bridge infrastructures with a relatively high
level of deficiency. Factors that have contributed to this level of deficiency are the age of
the infrastructure, inadequate maintenance, increasing loads and environmental conditions
which lead to deterioration of the bridges. There are various courses of action available to
the infrastructure owners to alleviate this problem. These options include replacement,
rehabilitation, strengthening, load posting and evaluation of load carrying capacity to
determine if the bridge can carry unrestricted traffic. In many cases, the high cost of
rehabilitation or replacement can be avoided by accurate evaluation of the structure's
capacity. The assessment of load carrying capacity can include analytical rating methods,
evaluation by load testing, materials testing, structural analysis and probabilistic methods
(Nowak and Tharambala 1988).

DEFINITION OF PROOF LOAD TESTING

9. The aim of the proof loading facility is to practically determine realistic bridge load
ratings for structures which might otherwise require posted load limits or major
rehabilitation. While analytical assessment of bridges is commonly used, this often lends
to conservative estimates of bridge capacity. It is difficult to accurately determine load
carrying capacity through theoretical analysis. Factors such as load distribution, composite
action, material properties, unintended continuity and end restraint effects are difficult to
model accurately. These factors often combine to produce strengths significantly in excess
of that indicated by theoretical analysis. Experience has shown that the predicted structural
behaviour is rarely in close agreement with results measured during physical tests (Phillips
and Wood 1987).
10. Proof load testing is a non-destructive load testing procedure in which the bridge is
incrementally loaded to a level significantly higher than the maximum legal loads. The
bridge response is monitored and recorded. The test load applied is generally two to three
times the legal load level. This allows dynamic and overload effects to be accounted for. A
rated load capacity is determined from the observed structural response under the test load
with reductions applied to cater for live load factors and dyirmic increments. (Fu 1995;
Agarwal 1988; Heywood et al 1996).

234
 BRIDGING THE MILLENNIA

11. The rated load calculated from the test results is not necessarily the maximum safe
load that the structure can resist. Proof load testing does not enable determination of the
ultimate capacity of the structure as the nature of the test is to load the structure within the
linear elastic range. Rather, it enables a lower bound on strength to be determined because
the application of a significant test load effectively truncates the possibilities of lower
strength levels.
12. Proof load testing carries a small risk of damage or failure of the bridge under these
large loads. However, it is considered that the risk associated with a carefully planned and
controlled test is significantly lower than the risk of failure due to vehicular loads on a
bridge for which the capacity is unknown or questionable (Phillips and Wood 1987).

LOAD TESTING OF BRIDGES INTERNATIONALLY

CANADA

13. The Ministry of Transportation, Ontario, Canada, (MTO) has a long history of
assessment of bridges by proof load testing. Over 225 bridges have been proof tested since
the inception of the program in 1969 (Bakht & Jaeger 1990). These tests have consistently
shown that the bridges have reserves of strength not indicated by theoretical analysis.
MTO are the recognised world leaders in this technology and the procedures developed in
Canada have been adapted for use in New Zealand, USA and Australia.
14. The testing facility consists of two specially built test vehicles which are capable of
carrying the high loads required during a proof test with a maximum capacity of
approximately 100 tonnes. Loading is applied through the use of concrete kentledge
blocks. Stockpiles of these blocks are stored in various locations throughout the Province
to minimise the costs of transportation between test sites. Concrete blocks are used in
preference to other methods of loading such as gravel or water, because of the stability
offered by solid blocks and more importantly, the ability to accurately plan the loading
levels and to know the load that is distributed to each axle (Bakht et al 1981).
15. Bridge response is measured by mechanical strain gauges and linear voltage
displacement transducers. Weldable strain gauges are used in some instances but the use
of the mechanical gauges is preferred due to the minimal time required to install the
instruments. These gauges can be clamped to the structure in a matter of minutes
compared to the considerable surface preparation required for contact strain gauges.
16. A recreational vehicle (mobile home) is used to house the data acquisition facility
and all the associated computer equipment required during the tests. This facility also
provides a mobile office in which the test engineer can evaluate the bridge response during
testing.
17. The test loads applied during a proof test are based on the ultimate limit state live
load according to the Ontario Highway Bridge Design Code. A live load factor of 1.4 and
a dynamic load allowance is applied to the code evaluation vehicle which is equivalent to
the maximum permissible vehicle. Note that the load factor applied for the Serviceability
Limit State in the Ontario Highway Bridge Design Code is less than unity.

235
 BRIDGING THE MILLENNIA

18. Rated loads are calculated from the proof load by reducing the test load by a live
load factor, a load shift allowance, an overload allowance and a dynamic load allowance.
These factors have been incorporated to produce a series of posting charts from which the
rated load can be read directly. The development of the posting charts is discussed in
Agarwal (1978).
UNITED STATES OF AMERICA

19. The United States, has a well developed and formalised system for inspection,
evaluation and rating of bridges. Gordon and O'Connor (1992) cite the collapse of the
Silver Bridge in West Virginia in the early 1970's as the catalyst in establishment of the
National Bridge Inspection Standards (NBIS).
20. The NBIS requires that all states maintain an up-to-date bridge inventory, that
inspections are carried out at least every two years, and that inventory and inspection data
is reported to a national database annually. This information is then used to manage
federally funded bridge rehabilitation and replacement programs. The information now
contained in the National Bridge Inventory indicates that of approximately 600,000
bridges in the USA about 50% are classed as deficient and in need of replacement or
rehabilitation.
21. In the last decade the New York State Department of Transportation has conducted
research into the use of proof load testing in assessment of bridges. Most of this work has
concentrated on the use of reliability analysis and its application in the selection of rating
factors and determination of target proof loads.
22. Fu (1995) recommends a procedure for determining a target proof load for
assessment of bridges. The target proof load is calculated by applying a live load factor
and a dynamic load allowance to the nominal design live load and factoring this load by
the inverse of the capacity reduction factor. The determination of an appropriate live load
factor is based on reliability theory and is calibrated to produce a safety index of 2.3. This
safety index was selected as it has been estimated that theoretical evaluation and rating of
bridges according to the AASHTO Manual (1983) assures a safety index of about 2.3 for
primary components such as girders. This procedure results in live load factors ranging
from 1.35 to 2.0 depending on the condition of the bridge and the average daily truck
traffic.
Lichtenstein (1993) recommends a range of factors to be used in calculation of a rated load
from the applied proof load. The basic reduction factor is 1.4 with adjustments to account
for fracture critical details, single lane structures, the frequency of inspection following the
proof load and the performance of the bridge during the test. These adjustments lead to a
live load factor ranging from 1.3 to 2.2. This load factor is applied to the actual proof load
level to achieve a bridge Operating Level Rating. The Operating Rating is the absolute
maximum permissible load level for the bridge. An Inventory Level Rating is calculated
as a fraction (approximately 73%) of the Operating Level Rating. The basic load factor of
1.4 was selected to give the same level of reliability as that from theoretical calculation of
the operating capacity.

236
 BRIDGING THE MILLENNIA

SWITZERLAND

23. Switzerland has a long history of load testing new bridges before they are
commissioned for service. The Federal and Cantonal Administration requires a load test of
every major bridge in Switzerland before it is opened to traffic (Favre et al 1992). As a
result of this requirement, over 200 bridges have been load tested in Switzerland.
24. The load tests conducted to date in Switzerland have been described by Moses et al
(1994), as 'Acceptance Proof Tests' and are primarily used to quantify the global
behaviour of the structure. The tests are conducted to verify the design assumptions
regarding structural behaviour by comparing the measured behaviour with analytical
predictions. The major objective of the test is to ensure that structural behaviour under the
applied load is linear and that the measured displacements are fully recoverable. The
serviceability performance is investigated by monitoring crack widths in the concrete and
ensuring that they remain within acceptable limits.
25. According to Favre, et al (1992) substantial test loads are applied to the structure.
The actual loads applied are approximately 5 kPa or 250 IN per vehicle which is
considerably lower than proof loads applied in other countries. The loads are applied by
dump trucks loaded with gravel or sand with each vehicle representing a load increment.
These loads are relatively low in terms of proof testing and generally lead to prestressed
concrete structures remaining uncracked.
NEW ZEALAND

26. Following the approach adopted by MTO, authorities in New Zealand proof tested
five bridges between 1977 and 1987. These tests were all performed on bridges
theoretically assessed to be substandard but in all cases the load restrictions were removed
or raised significantly (Phillips and Wood 1987). No literature was found to indicate that
further proof testing has been conducted.
27. According to the New Zealand code of practice NZS3101:1982, the target proof
load is the normal legal load factored by a live load factor of 1.4 and including a dynamic
load allowance. The legal load is approximately 85% of the code design live load.
Typically this equates to a target proof load of 16 t to 24 t on the rear axle group. Loads
are generally applied using vehicles loaded with gravel or coils of prestressing wire.
28. The test is considered to be successful and no load limit is required if the following
criteria are met : the target load is applied without signs of distress, the bridge behaviour is
within 20% of linear-elastic response and the crack widths do not exceed 150% of the
normal allowable widths. If the test is successful no further rating calculations are
required. Two repetitions of the loading sequence are recommended to ensure that the
structural response is repeatable.
29. NZS 3101 : 1982 recommends that at maximum load, the measured deflection
should be within 20% of the deflection obtained by extrapolating the straight portion of
the load-deflection curve to the target load. This allowable non-linearity appears to be very
generous and Phillips and Wood (1987) suggest that the test should be stopped if the
extent of non-linearity reaches 10%. A detailed inspection of the bridge should then be
conducted before proceeding with further testing.

237
 BRIDGING THE MILLENNIA

UNITED KINGDOM

30. Load testing of bridges has been carried out in the United Kingdom for over 100
years but according to Thompson (1981) the practice remains controversial.
Predominantly, the load tests that have been performed are used to calibrate analytical
models, or are ultimate failure tests of bridges that have been removed from service.
31. On January 1, 1999 highways in the United Kingdom will be subjected to an
increase of legal loads to 40t. There is currently a strong push to have all bridges assessed
and, where necessary, strengthened or replaced by that time (Silke 1994).
32. The UK Department of Transport Standard BD44/90 is aimed specifically at
theoretical assessment of bridges. It was developed primarily because of changes in codes
and loadings over the years. According to Tilly (1992) the design codes are not suitable
for assessment as they deal with maximum levels of uncertainty.
33. A seminar was held in London in 1994 to debate the use of proof tests in this
assessment process but the Department of Transport (DoT) refuses to allow proof load
testing as an acceptable tool for assessment of these bridges. This is due to concerns that
either observed or hidden damage may occur during a proof test that would lead to
deterioration and loss of capacity. The DoT will only accept supplementary testing where
loads no greater than the Serviceability Limit State loads are used to refine or calibrate a
mathematical model (Cullington 1994). The Department favours work being carried out in
New York to link reliability theory to load testing practice (Fu and Tang 1995).
34. Challdey (1994) describes a proving load test where groups of lanes on a seven lane
bridge were loaded using steel weights. An ultimate envelope was then created by
summing the responses for each load case and comparing the results with the maximum
measured deflection at each position. There has been some debate over the validity of the
superposition principle in this instance as the envelope created may be within the non-
linear range.

CASE STUDY : PROOF LOAD TESTING OF SOUTH PINE RIVER BRIDGE

35. The South Pine River Bridge (1941) is a six span simply supported bridge located in
the northern outskirts of Brisbane. The superstructure consists of six 12m spans each with
four rolled steel I-girders and a cast in-situ concrete deck. The deck is made composite
with the girders by means of shear bars welded to the top flange of the girders. The bridge
was designed for a 20t live load. Figure 2 shows a cross section of the bridge.
PRELIMINARY ANALYSIS AND BRIDGE INSPECTION

36. The bridge was analysed by a simple grillage analysis, using theoretical material
properties, and assuming composite and simply supported behaviour. The preliminary
analysis indicated that the maximum possible test load of 106 tonne on each of two test
vehicles would induce live load bending stresses of 103 MPa in one of the inner girders.
When added to calculated dead load stresses this gives a maximum predicted stress of
168 MPa in the inner girder. This is equivalent to 67% of the yield stress of 250 MPa. It
appears that although the bridge was constructed for composite behaviour between the

238
 BRIDGING THE MILLENNIA

girders and the deck, the benefits of this composite action were not utilised in the design.
Such a conservative approach is consistent with the innovative nature of the bridge. An
inspection of the bridge showed that it is generally in good condition with some corrosion
of the steel girder flanges at the piers and abutments.

7316 mm

Right lane 200 Left lane

Girder 1 Girder 2 Girder 31 24'x71/; x100# RSJ Girder 4

2134 2134 2134

><

Fig. 2 - Cross section of South Pine River Bridge

INSTRUMENTATION AND TEST SETUP

37. The target load was applied in seven load increments between 38t and 106t. For
each load level the bridge response was measured with the vehicles in 12 positions along
the bridge and three lateral positions : Vehicle A in the right lane, Vehicle B in the left
lane, and both Vehicles A and B.
38. Vertical deflections at midspan were measured using linear voltage displacement
transducers. Top and bottom flange girder strains were measured at midspan using 12052
foil stain gauges. Displacement transducers installed at the piers monitored movement at
the bearings.
39. Specially developed computer software allowed immediate assessment of the bridge
response and comparisons with the predicted response. This was a critical requirement
after each load level to determine whether the structural response was within the linear
elastic range and whether the test could proceed to the next load level. The graphs
presented in Figures 3, 4 and 5 are samples of the information available within seconds of
a field recording being undertaken.

TEST RESULTS

40. Figure 3 plots the bottom flange strain against gross vehicle mass (GVM) for girder
3 in Span 1 (compression +ve). This figure shows that the experimental strains are
approximately 65% of the theoretical strains and that the GVM versus strain plots remain
linear for each loading arrangement.

239
 BRIDGING THE MILLENNIA

Gross Mass of Vehicles (t)

250
-512 -335
200 Veh A right lane
Veh B - et lane
150
Veh A & B
100
Veh A - right (theory)
50 0 Veh B - left (theory)

0 0 Veh A&B (theory)
-600 -400 -200 0
Bottom Flange Strain (c)

Figure 3 Midspan girder strains SB1-3

41. The midspan deflection response shown in Figure 4 is similar to that of the strain.
The observed deflections are approximately 60% of the predicted values. These plots
show that the structure is behaving in the linear elastic range. The measured deflections
and strains are consistent with the principle of superposition.
250
Gross Mass of Vehicles (t)

6.8 11.6
200 Veh A - right lane
MI Veh B - left lane
150
Veh A & B
100 O Veh A - right (theory)
171Veh B - left (theory)
50
0Veh A & B (theory)
0
0 5 10 15
Deflection ( mm

Figure 4 Midspan deflections D1-3

42. The significant discrepancy between the measured and predicted responses could be
attributed to a number of factors. These include conservative estimates of material
properties, the effects of the kerbs, the effect of bearing restraints and unintended
continuity and rotational restraints at fixed ends. The effect of bearing restraint has been
shown by Bakht (1988) to have significant effects on measured strains in some slab on
girder bridges. It was observed from the test results that some continuous action was being
achieved that had not been accounted for in the theoretical analysis.
43. Figure 5 shows a plot of deflection at midspan versus vehicle position for the two
end spans. This essentially shows the midspan deflection influence lines for the test
vehicles for each girder in these two spans. These influence lines clearly indicate that
some degree of continuity is occurring at the piers even though pier 1 is nominally an
expansion joint. It would appear that the bearings at this pier have frozen giving some

240
 BRIDGING THE MILLENNIA

degree of restraint. This continuity effect coupled with the longitudinal restraint reduces
the midspan strains and deflections. There appears to be little difference in the degree of
continuity afforded by the frozen expansion joints at pier 1 and the rotational restraint
which occurs at the fixed joints at pier 2 where the girders are cast into the pier.
-1
0
Deflection (mm)

1
2 ":
3
4
5
6
7
-10 0 10 20 30 40 50
Distance to steer axle from Abut A (m)

-0-D1-1 -0-D1-2 ---D1-3 -0-01-4

--A- D2-1 -U- D2-2 -A- D2-3 --0- D2-4
Figure 5 Influence lines for midspan displacement Spans 1 and 2 Two vehicles @ 106 t

CALCULATION OF RATED LOAD FOR SOUTH PINE RIVER BRIDGE

44. Rated loads were calculated using the two methods adopted in Canada described by
Agarwal (1988; 1989). The following parameters were assumed.

Elastic limit steel = 80% of the yield strain

Applied load = 106t per vehicle
Load shift factor = 1.0
Live load factor = 2.0
Dead load factor = 1.2
Dynamic Load allowance = 40%
Overload allowance = Nil (live load factor of 2.0 applied
compared with 1.4 in Canada)

Method 1
The applied load is assumed to be the maximum load that the structure can sustain. The
applied load is reduced to a rated load by dividing by the live load factor and the dynamic
load allowance.

This method gives a conservative rated load of 37.9t.

Method 2
The strain recorded at the applied test load is extrapolated to the elastic limit. The
corresponding load is then converted to a rated load. This method requires the estimation
of dead load effects and is only suitable for use on steel structures which are ductile and
where high levels of confidence in material properties exist. Equation (1) shows the

241
 BRIDGING THE MILLENNIA

calculation of the available live load capacity. The rated load is then calculated in the same
manner as Method 1.

Applied Load
Available live load capacity x (Elastic Strain Limit - Dead Load Strain) (1)
Measured Strain

This method gives a rated load capacity of 68.9t from a test load of 106t. This rating
involves a significant degree of extrapolation of the measured test data. It should only be
used when one is confident that the behaviour will not change at higher loads.

CONCLUSIONS

45. A review of the literature available on proof load testing has shown that two
distinctly opposite attitudes exist towards this method of load assessment. Canada, USA
New Zealand and Australia have embraced the concept of proof load testing as a means of
assessing the capacity of older structures. The extent of load testing programs in these
countries varies from well established programs in Canada where over 225 bridges have
been tested, through limited but increasing use in USA and Australia to relatively confined
programs such as that in New Zealand where only five bridges were tested between 1977
and 1987.
46. Irrespective of the extent of the testing programs, the literature definitely indicated a
high level of success in removing or raising posted limits in all these countries.
47. In direct contrast to the attitudes to proof load testing in these countries, the United
Kingdom has taken a fine stance against the use of proof load testing for assessment
purposes. The primary reason given for this decision is a reluctance to accept the risk of
damage to the structure associated with the relatively high loads applied during a proof
load test. Instead the United Kingdom relies predominantly on theoretical assessment
rating and supplements these results with the results of load testing at loads no greater than
the serviceability limit state traffic loads. Perhaps this reluctance stems from the fact that
the code-recommended live loads and load factors in the UK are significantly higher than
elsewhere. Therefore the target proof load that would need to be applied for a load test is
significantly higher than that used in other countries.
48. In general, where proof load testing is accepted as a viable method in the assessment
of bridge capacity, the methods adopted are similar. The method generally consists of the
following steps.
Preliminary analysis and inspection of the bridge
Determination of target proof load
Preparation of Instrumentation and data acquisition plans
Planning of Traffic Control
Incremental application of loading
Data analysis and evaluation
Recommendations and reporting.

242
 BRIDGING THE MILLENNIA

49. Significant variations have been found in the live load factor applied to determine a
target proof load and to determine a rated load from test results. In general, the rated load
is calculated as the applied proof load reduced by the live load factor and a dynamic load
allowance. Live load factors range from 1.35 to 2.2. These factors are dependant on
conditions such as average daily traffic on the bridge and structural type. Further research
is currently being undertaken into the application of reliability theory to determine an
appropriate live load factor for use in proof load testing in Australia.
50. The actual test procedures have been well documented by Fu (1995), Agarwal
(1988), Heywood (1992) and others. The most important factor to be considered in the
actual test procedures, aside from the magnitude of the applied loads, is the simplification
of both the instrumentation and the loading of kentledge onto the test vehicles. These two
procedures have been observed to be the critical components in terms of time and
therefore of the cost of testing.
51. The South Pine River Bridge has been proof loaded with two vehicles of 106 t each.
The bridge was designed in 1941 for design vehicles of approximately 20 t, but has been
load tested to show that its capacity is significantly higher than the current legal limit of
42.5 t for six axle articulated vehicles. The bridge remained linear elastic during testing
and composite action was confirmed up to a loading of 106 t. The load rating methods
adopted in Canada were applied to the midspan bending strains and show that the bridge
has adequate capacity to remain in service. If no extrapolation of the measured behaviour
is permitted then a conservative rating of 38 t results. If observed behaviour is extrapolated
to the elastic limit, then a rating of 69 t results.
52. Proof load testing has been shown to be a valuable tool that can be used to safely
and accurately assess the live load carrying capacity of bridges.

REFERENCES

AASHTO (1983) Manual for Maintenance Inspection of Bridges, Washington DC

AGARWAL, A. (1978) Evaluation and Posting of Bridges in Ontario, Transportation Research Record 664,
Bridge Engineering Vol. 1, Transportation Research Board, Washington DC.
AGARWAL, A. (1988) Bridge Evaluation by Load Testing Presented at Bridge Evaluation and
Rehabilitation Seminar, Canadian Society for Civil Engineering, Montreal, Technical University of Nova
Scotia, Halifax, Canada.
AGARWAL, A. (1989). Determination of Load Shift Allowance for Bridge Posting (Ministry of
Transportation Ontario, Downsview, Ontario, Canada) (Unpublished Paper)
AUSTROADS (1992). Bridge Design Code (AUSTROADS, Sydney)
AUSTROADS (1994 a). RoadFacts, Publication No AP 18/94 (AUSTROADS Sydney)
AUSTROADS (1994 b). Strategy for Structures Research and Development (AUSTROADS, Sydney)
BAKHT, B. (1988). Testing of and Old Short Span Slab - on Girder Bridge (Research and Development
Branch, Ministry of Transportation Ontario, Downsview, Canada)
BAKHT, B., & JAEGER, L., (1990) Bridge Testing: A Surprise Every Time. American Society of Civil
Engineering, Journal of Structural Engineering Vol 116 No 5 1990. pp 1370-1383
BAKHT, B., JAEGER, L., NAIK, T. (1981). Bridge Evaluation by Field Testing Proceedings of the
International Symposium on Rehabilitation of Structures, Bombay, India, December 1981, pp 2.57-2.64

243
 BRIDGING THE MILLENNIA

CHALKLEY, C., (1994) Hidden Strength Load Testing For Bridge Assessment LoBEG Load Testing Task
Force, London , February 1994
CULLINGTON, D., (1994) Supplementary Load Testing and Ultimate Load Testing Part I : TRL
Experiences Transport Research Laboratory, Crowthome
Department of Transport (1990) Departmental Standard BD44/90 The assessment of Concrete Highway
Bridges and Structures
FAVRE, R., HASSAN, M., MARKEY, 1. (1992) Bridge Behaviour Drawn from Load Tests Bridge
Rehabilitation, Proceedings of the 3rd International Workshop on Bridge Rehabilitation, June 14-17,1992,
Technical University Darmstadt, Konig, G., Nowak, A., Novak, B
FU, G. (1995) Highway Bridge Rating by Nondestructive Testing for Consistent Safety, Research Report
163, New York State Department of Transportation.
FU, G., TANG, J., (1995) Risk Based Proof Load Requirements for Bridge Evaluation Journal of Structural
Engineering, Vol 121, No. 3, March 1995, pp 542-556
GORDON, S., O'CONNOR, D., (1992) Bridge Management, Evaluation and Rehabilitation - A Survey of
Current Practice Bridge Rehabilitation, Proceedings of the 3rd International Workshop on Bridge
Rehabilitation, June 14-17,1992, Technical University Darmstad
HEYWOOD, R., (1992) Bridge Testing in Canada and the United States of America - Recommendations for
Australia Research Report 92-12, (Physical Infrastructure Centre, School of Civil Engineering, Queensland
University of Technology)
HEYWOOD, R., ARIYARATNE, W., WEDGWOOD, R., OATES, R. (1996) Australia's Bridge Proof
Loading Facility - a Cost Effective Asset Management Tool Proc. Combined 18th ARRB Transport
Research Conference, Transit New Zealand Land Transport Symposium, Christchurch
HEYWOOD, R., MCDONNELL, A., (1996) Proof Load Testing and Monitoring of the South Pine River
Bridge Asia-Pacific Symposium on Bridge Loading and Fatigue, Monash University, December 1996
LICHTENSTEIN (1993) Unpublished NCHRP Manual for Bridge Rating through nondestructive testing.
Ministry of Transportation and Communications Ontario (1983) Ontario Highway Bridge Design Code,
Downsview, Ontario, Canada.
MOSES, F., LEBET, J., BEZ, R. (1994) Application of Field Testing to Bridge Evaluation Journal of
Structural Engineering, Vol 120, No. 6, June 1994, pp 1745-17620
NOWAK, A., THARMABALA, T., (1988) Bridge Reliability Evaluation Using Load Tests Journal of
Structural Engineering, Vol 114, No. 10, October 1988, pp 2268-2279
PHILLIPS, M., WOOD, J., (1987) Proof Loading of Highway Bridges, New Zealand Roading Symposium,
Vol 4, pp 803-807
SILKE, E., (1994) Testing Time for Bridges Surveyor, Vol 181, Issue 5281, March 1994, pp 20-22
Standards Association of New Zealand, NZS 3101:1982, Code of Practice for the Design of Concrete
Structures.
THOMPSON, D. (1981) Load Tests on Highway Bridges : A Review Transport and Road Research
Laboratory Report 1032, Crowthome, Berkshire
TILLY, G., (1992) Recent Advances in Structural Assessment of Highway Bridges Bridge Rehabilitation,
Proceedings of the 3rd International Workshop on Bridge Rehabilitation, June 14-17,1992, Technical
University Darmstadt, Konig, G., Nowak, A.
UNITED STATES CODE OF FEDERAL REGULATIONS, 23 Highways - Part 650, Subpart C - National
Bridge Inspection Standards

244