Many bridges and special structures no longer have construction drawings, or their use has changed so that the loads differ from those they were designed for. Back-calculation from drawings alone therefore often gives an overly conservative answer, or one with high uncertainty. Testing with real loads provides direct data on how the existing structure behaves. This article explains the types of load test, how they are carried out, what instruments are used and how the results are read.
What is a bridge load test?
A bridge load test is a test of the load-carrying capacity of an existing structure, carried out by applying precisely known loads in stages and measuring how the structure responds. The main quantities measured are deflection and strain, which reveal both the actual stiffness of the structure and how forces are distributed between its members.
A load test does not test how strong the concrete or steel is; it tests how the structural system as a whole works together under real loads. This includes the effects of the deck, railings, wearing surface and structural continuity that models often ignore. The result is often that the bridge is stiffer than calculated.
Non-destructive testing, such as assessing concrete quality or locating reinforcement, answers questions about material condition but not about the load-carrying capacity of the whole system. In practice the two are therefore used together: assess the condition first, then decide whether a load test is needed.
Which bridges should be load tested?
Bridges that should be load tested are those where data is insufficient for a confident calculation, or where the conditions of use have changed. Not every bridge needs a load test. There are 6 common cases.
- Old bridges without construction drawings — with no data on the original reinforcement or concrete strength, calculations rely on many assumptions; testing greatly reduces the uncertainty.
- Increasing the load or changing the type of use — for example, opening the bridge to heavy trucks, or converting a footbridge to carry vehicles.
- After repair or strengthening — when numerical evidence is needed that the strengthening works as designed.
- Structures that have been struck or damaged — for example, a girder hit by an over-height truck, or after flooding and scour of the foundations.
- Moving exceptionally heavy loads across the bridge — one-off transport of transformers, machinery or large structural components.
- Acceptance of new bridges — some contracts require testing to confirm behavior before opening to traffic.
Before deciding to load a structure suspected of deterioration, always assess the condition of the concrete and reinforcement first, so that the test itself does not accelerate the damage. See the methods available at this stage in the NDT guide to non-destructive testing of concrete structures, and for old buildings or structures whose use will change, read the assessment sequence in What structural checks does an old building need before an extension?
Diagnostic load test vs proof load test: what is the difference?
A diagnostic load test uses a safe load level to observe actual behavior and then calibrate the model, whereas a proof load test uses a higher load to prove directly that the structure can carry the target load. Both approaches are covered in the AASHTO Manual for Bridge Evaluation, the main and widely used reference for bridge evaluation.
← Scroll to see the full table →
| Topic | Diagnostic Load Test | Proof Load Test |
|---|---|---|
| Purpose | Understand actual behavior, then calibrate the analysis model to match reality | Prove directly that the target load can be carried without damage |
| Load level | Below or equivalent to normal service load | Above service load, increased in stages up to the target |
| Risk to the structure | Low | Higher; requires a test-stop plan and contingency measures |
| Output | Actual stiffness, load distribution, model calibration factors | A load level confirmed directly by testing |
| Best for | Bridges with drawings whose calculated capacity appears overly conservative | Bridges without drawings, or that fail by calculation yet still perform normally in service |
| Preparation | Medium | Extensive; requires prior analysis and clearly defined stop criteria |
Most projects start with a diagnostic test, because it yields a lot of data at low risk, and then decide whether a proof load test is needed. Going straight to a proof test on a structure whose behavior is still unknown should be avoided.
What are the on-site test steps?
The test has 7 steps, starting with data collection and prior analysis, followed by staged loading, unloading and reporting. The desk work before going to site is therefore as important as the test day itself.
- Collect data and assess condition — construction drawings, repair history, current damage, and NDT results if available.
- Prior analysis — build a model to predict the values that should be measured at each stage, and use it to set instrument locations and test-stop criteria.
- Plan loading and traffic — define the number of stages, parking positions, access routes and traffic-closure windows.
- Install instruments and take zero readings — install according to the analyzed layout, then take initial readings with no load.
- Apply load in stages — at each stage, wait for readings to stabilize before recording, and compare immediately with the predicted values before moving to the next stage.
- Unload and read recovery — unload in stages as well, then take readings after full unloading to check residual deflection.
- Analyze results and report — compare measured with calculated values, calibrate the model, and have the responsible engineer summarize the recommendations.
An often overlooked point is comparing values during the test. If you only look at the results at the end, there is no chance to stop before the structure enters undesirable behavior.
The two images below expand on the steps above. The first is a 2D diagram showing the bridge cross-section with the truck parked off-center, staged loading with a comparison at every stage, and two ways of setting up the deflection reference point. The second is a 3D model of the test site with the seven-step work sequence.
What instruments are used?
There are 4 main groups of instruments: deflection gauges, strain gauges, temperature sensors and data loggers.
- LVDT or Dial Gauge — measures deflection at midspan and other points per the analysis layout; requires a reference point that does not move with the structure under test.
- Strain Gauge — measures strain on the concrete surface, steel surface or exposed reinforcement, used to assess load distribution between individual girders.
- Temperature sensors — record temperature alongside, because temperature changes during the day can cause significant errors in the readings.
- Data logger — reads all channels simultaneously in real time, showing the relationship between load and response at every moment.
Where the underside of the bridge cannot be accessed, such as a river bridge with high water levels or strong currents, the usual alternative is a high-precision total station measuring movement from a distance. Its resolution is lower, but adequate for large deflections.
How are load test results read?
Look at three things together: the maximum deflection compared with the criteria set in advance, the linearity of the load-deflection curve, and the recovery after unloading. No single one of these can be used to decide on its own.
A curve that begins to bend away from the straight line, or a large residual deflection after unloading, is a sign that the structure is starting to behave inelastically.
For numerical criteria, the commonly cited references are ACI 437.2 (current edition ACI CODE-437.2-22), the code requirements for load testing of existing concrete structures, and ACI 318 Chapter 27 (current edition ACI CODE-318-25), on strength evaluation of existing structures. Both require residual deflection to be considered relative to the maximum measured deflection, together with inspection for cracks that develop during the test.
Deflection limits and recovery ratios have details and conditions of use that differ between standards and editions. This article therefore does not state numerical criteria, because the criteria actually used must be set by the responsible engineer, based on the edition of the standard in force for that project and on the prior analysis of the actual structure.
What other special structures are tested the same way?
The principle of applying load in stages and measuring the response also applies to structures other than bridges. There are 5 common types.
- Factory and warehouse floors — before adding racking or switching to storing heavier goods. Read about floor test methods in Floor Load Test: testing floor capacity with real loads, and about assessment approaches in How much load can a factory or warehouse floor carry?
- Ramps and parking decks — especially old parking structures that will carry heavier vehicles than before.
- Crane runway beams in factories — before changing the lifting capacity of an existing crane.
- Long-span roof structures — when additional building services are to be hung from the existing structure.
- Walkways and cantilevered balconies — in old buildings whose type of use is changing.
For fixing heavy items to an existing structure, such as chemical anchors or threaded studs, the test used is a local pull-out test, which is a different method from loading a whole span. See details in Pull-Out Test for chemical anchors and threaded studs
What does the price of a bridge load test depend on?
There is no unit price for bridge load testing that applies to every job, because the scope varies greatly from project to project. There are 5 factors that determine the price and duration of the work.
- Span length and number of spans to be tested
- Number of deflection and strain measurement points
- Required load level and how the load is provided
- Access to the underside of the bridge, such as scaffolding, boats or boom lifts
- Traffic closure and available working hours
Contact us for a free quotation within 24 hours. Our team will ask for basic information, such as bridge drawings if available, span lengths, site conditions under the bridge and the purpose of the test, to scope the work to match the actual site.
This price excludes travel and equipment mobilization costs, which are charged extra depending on the distance to site.
Summary
A bridge load test answers the question of load-carrying capacity from the actual behavior of the structure. It is best suited to bridges with incomplete data, bridges whose use will change, and bridges that have been repaired or strengthened and need proof of performance.
Success depends on preparation: prior analysis to know what should be measured, placing instruments at the critical points, setting test-stop criteria before going to site, and comparing readings at every stage during the test. Interpretation of the results and conclusions about the permissible service load must always go through the responsible engineer.
This content summarizes practice for general understanding only. It is not a test specification and not a project-specific work plan. Test load levels, instrument locations, test-stop criteria and conclusions on usable load capacity must be set and certified by the project's responsible engineer in accordance with the standard in force.