Last reviewed 30 Sept 2026 · 7 min read
Purpose of a load test
A load test applies known loads to a bridge and measures its response. It answers questions that calculation alone cannot:
- Acceptance of a new bridge — does the completed structure behave as designed?
- Assessment of an existing bridge — what is its real stiffness and capacity, in the presence of deterioration, undocumented changes or uncertain materials?
- Proof of a new or unusual system — cable-stayed and long-span bridges, new bearings or connections.
- Calibration of a computer model — to update the stiffness and boundary conditions.
- Assessment after damage, rehabilitation or strengthening.
- Speed increase or heavier loading on a railway bridge.
A load test never replaces analysis: it complements it. It gives the behaviour at the test load; extrapolation to the ultimate capacity needs judgement.
Static load test
Planning
FormulaPlanning a static test
- Objectives and acceptance criteria decided in advance (deflection, strain and recovery limits).
- Analysis of the bridge to predict the responses to the test loads (deflection, strain, rotation) at each point and stage.
- Test load: in India the guidelines for load testing (the IRC special publication on load testing of bridges and the codes) specify the magnitude and duration, generally up to the design live load with impact reduction or a service value; the load is applied in increments (for example 25, 50, 75, 100 % of the test load) with holds.
- Load application: heavy trucks or tippers of known weight (weighed at a weighbridge), water tanks, concrete blocks or sand bags, or hydraulic jacks against reaction frames; positioned to produce the maximum effect (moment, shear) in the member being tested.
- Instrumentation plan: deflection, strain, rotation, crack width, bearing movement and temperature at defined points; redundant sensors at critical points.
- Safety plan: stop criteria, evacuation, shoring beneath the bridge, personnel, traffic diversion, weather.
Measurements
- Deflection — dial gauges or LVDTs on a fixed reference (a scaffold not connected to the bridge), precise levelling, laser or total-station for large bridges.
- Strain — strain gauges (foil or vibrating-wire) on girders, bars and slabs; neutral-axis position and composite action are found by strain profiles.
- Rotation — tiltmeters at the ends.
- Bearing and joint movement, and crack opening.
- Temperature — because the bridge deforms with temperature, tests are run at stable temperature (early morning or night) and the readings are corrected for temperature or a control gauge is used.
Procedure
- Record the zero (initial) readings with no test load — for a sufficient time to establish the drift.
- Apply the load step by step; after each step hold until the readings are steady, and compare with the predictions before proceeding.
- At the maximum test load, hold for the specified period (for example a period of hours for a proof test) and note any creep.
- Unload in stages and record the recovery.
- Record final zero readings.
- Repeat the cycle if required.
Evaluation and acceptance
The bridge is judged by comparing measurements with predictions and with limits:
| Criterion | Meaning |
|---|---|
| Linearity | The load–deflection curve should be nearly linear (no sign of yielding or cracking beyond that expected) |
| Measured vs predicted | Measured deflections and strains should not exceed the predictions by more than a set margin (they are usually lower than predicted because real bridges are stiffer) |
| Recovery | On unloading, a high percentage of the deflection should recover — typically 75–80 % or more for concrete, higher for steel (check the code for the exact value) |
| Crack behaviour | New cracks or crack widening beyond the allowable limit is a warning |
| Stability | Readings should stabilise at each hold, without continuing creep |
Worked ExampleExample — recovery
A girder mid-span deflection under the maximum test load is measured as 10.5 mm. After unloading, the residual deflection is 1.2 mm.
Recovery — satisfies a typical recovery criterion of 75–80 %. The predicted deflection was 12.0 mm; the measured 10.5 mm is 12 % lower — indicating that the bridge is stiffer than the model (composite action of the kerbs, and the real modulus of concrete).