How do you determine pile failure from test results?

Pile failure in a load test is identified when a pile can no longer sustain increasing load without continuous, uncontrolled settlement. In practice, engineers rarely observe true plunging failure directly. Instead, they apply defined failure criteria to the load-settlement curve to determine the interpreted failure load, which then informs the pile’s design capacity.
The specific criterion used depends on the test method, the applicable standard, and the project’s design approach. Different criteria can produce different failure load values from the same test data, so understanding how each one works is important for anyone reviewing or commissioning a pile load test.

What criteria are used to define pile failure in load tests?

Pile failure in a load test is defined by a criterion that sets a threshold beyond which the pile is considered to have reached its ultimate capacity. No single universal criterion exists. Engineers apply different definitions depending on the test standard, pile type, and design purpose, and the chosen criterion directly affects the interpreted failure load.

The most widely used criteria fall into three categories:

Settlement-based criteria: Failure is defined when total pile head settlement reaches a fixed value, such as 10% of the pile diameter. This is straightforward to apply and commonly used in European practice under standards such as EN 1997.
Offset limit criteria: Methods such as the Davisson offset limit define failure as the point where the load-settlement curve diverges from the elastic compression line by a defined offset. This approach is widely used in North American practice and tends to give a more conservative result than settlement-based methods.
Rate-of-settlement criteria: Failure is defined when the rate of settlement under a given load increment exceeds a specified threshold, such as 0.2 mm per minute. This approach captures time-dependent behavior and is particularly relevant in cohesive soils.

Some standards also permit graphical methods, such as the log-log plot of load versus settlement or the Chin-Kondner extrapolation, to project a failure load when the test has not been taken to full plunging. Each method produces a different numerical result, so your team should agree on the applicable criterion before testing begins, not after the data is collected.

How is a load-settlement curve used to identify failure?

The load-settlement curve is the primary output of a static pile load test. It plots applied load on one axis against measured pile head displacement on the other, and the shape of that curve tells engineers whether the pile is behaving elastically, approaching its limit, or has exceeded its capacity.

In the early stages of loading, the curve is nearly linear. Settlement increases proportionally with load, and most of the displacement is recoverable when the load is removed. As load increases, the curve begins to bend, indicating that soil resistance is being progressively mobilized and some permanent settlement is accumulating.

Failure is identified at the point where the curve changes character in a way that matches the chosen criterion. For settlement-based criteria, you read off the load corresponding to a defined displacement. For offset limit methods, you draw the elastic compression line and find where the measured curve diverges from it by the specified offset. For rate-based criteria, you examine the time-settlement records at each load increment rather than the curve shape alone.

A curve that continues to steepen sharply with no sign of stabilizing indicates that the pile is approaching or has reached plunging failure. A curve that flattens and stabilizes at each load increment suggests the pile has capacity remaining beyond the maximum applied load, which means the test result represents a lower bound rather than a true failure load.

What’s the difference between interpreted failure and actual plunging failure?

Actual plunging failure occurs when a pile undergoes continuous, uncontrolled settlement under constant or decreasing load. The pile has fully mobilized all available soil resistance and can no longer carry additional load. The load-settlement curve turns nearly vertical. This is the true geotechnical ultimate capacity, and it is rarely reached in practice because the risk of damaging the pile or the test setup is too high.

Interpreted failure is an engineering construct. It is the load value derived by applying a defined criterion to a test that was stopped before plunging occurred. Because most load tests are designed to reach 1.5 to 2 times the design load rather than true failure, the failure load must be inferred rather than directly observed.

The gap between interpreted and actual failure can be significant. A conservative criterion such as the Davisson offset limit may identify failure at a load well below what the pile could physically sustain. A less conservative criterion applied to the same data may produce a higher interpreted failure load. Neither is wrong in isolation. What matters is that the criterion is appropriate for the design standard being used and is applied consistently.

For design purposes, the interpreted failure load is then divided by a safety factor or combined with partial factors under a limit state framework to arrive at the allowable or design capacity. Understanding which failure definition was used is therefore important when comparing results across projects or test methods.

How do dynamic and static load tests compare for failure determination?

Static load testing and dynamic load testing approach failure determination in fundamentally different ways. Static testing measures load and displacement directly and produces a load-settlement curve from which failure can be read or interpreted using standard criteria. Dynamic testing derives capacity indirectly through signal matching analysis and does not produce a load-settlement curve in the same sense.

What static load testing provides

Static load testing applies a sustained, slowly increasing load and measures pile head displacement at each increment. The result is a direct, unambiguous load-settlement curve. Engineers can apply any standard failure criterion to this curve and arrive at an interpreted failure load with a clear evidentiary basis. The test also captures time-dependent settlement behavior, which is important in cohesive soils where creep and pore water pressure dissipation affect long-term performance.

What dynamic load testing provides

Dynamic load testing measures strain and acceleration at the pile head during a hammer impact. Signal matching analysis then derives a soil resistance model from which mobilized static capacity is estimated. The result is a capacity value, not a load-settlement curve. For end-bearing driven piles in granular soils, this estimate can fall within 10 to 20% of a static test result under favorable conditions. For bored piles or piles in cohesive soils, the difference can be 20 to 40% or more, and the method becomes considerably less reliable.

Dynamic testing cannot replicate the time-dependent behavior that static testing captures. It also cannot directly produce the load-settlement relationship that governs design when settlement, not just capacity, is the controlling criterion. For projects where the load-settlement curve governs design decisions, static or Rapid Load Testing remains the more appropriate choice.

What factors cause unexpected pile failure in test results?

Unexpected failure in a pile load test, where a pile fails at a load significantly below the predicted capacity, typically has one of several root causes. Identifying which factor is responsible requires careful review of the test data alongside the installation records and site investigation.

Inadequate soil investigation: Variable or poorly characterized soil conditions can mean that design assumptions do not reflect actual site conditions. Layers of weak material, undetected voids, or transitions between soil types that were not captured in the borehole data can all reduce actual capacity below the predicted value.
Pile installation problems: Damage during driving, inadequate concrete quality in cast-in-situ piles, insufficient penetration depth, or early refusal can all reduce the pile’s ability to mobilize the expected resistance. Monitoring data from installation, where available, often reveals these issues.
Setup time effects: For driven piles in cohesive soils, pore water pressures generated during installation temporarily reduce soil resistance. If a pile is tested before sufficient setup time has elapsed, the measured capacity will be lower than the long-term value. Conversely, testing too early in granular soils can also affect results if the soil has not fully reconsolidated around the pile.
Structural defects: Cracks, necking, or material discontinuities in the pile shaft reduce the pile’s structural capacity independently of soil resistance. These defects may not be visible at the pile head and require integrity testing to detect.
Test setup errors: Incorrect sensor placement, reference beam disturbance from the reaction system, or errors in load application can introduce measurement errors that appear as premature failure in the results.

When should a pile test result be considered inconclusive?

A pile test result is inconclusive when the data does not provide sufficient evidence to determine whether the pile has met, exceeded, or fallen short of its required capacity. This can occur for several reasons, and recognizing an inconclusive result is as important as interpreting a clear one.

The most common situation is a test that was stopped before the pile reached the target load or a recognizable failure point. If the load-settlement curve is still linear and showing no sign of approaching failure at the maximum applied load, the test confirms that the pile has at least that capacity, but it does not establish the actual failure load. This is a lower-bound result, not a failure determination.

Results also become inconclusive when the data quality is poor. In dynamic testing, a poor signal match, caused by variable pile geometry, damaged pile material, or unfavorable soil conditions, means the derived capacity estimate carries a wide bandwidth of uncertainty. Applying a safety factor to an already uncertain result may still leave the design on uncertain ground.

Other situations that warrant treating a result as inconclusive include:

Reference beam movement during a static test that compromises displacement measurements
Insufficient hammer energy in a dynamic test to fully mobilize soil resistance
Test results that fall outside the expected range without a clear explanation from installation records or site data
Conflicting results between integrity testing and capacity testing that suggest an unresolved structural issue

When a result is inconclusive, the appropriate response is not to accept the data at face value but to investigate the cause and, where necessary, retest or supplement with a different method.

How We Help You Interpret Pile Test Results

Determining pile failure from test results requires more than applying a formula to a curve. It requires experienced judgment about which criterion is appropriate, what the data quality allows, and how the result connects to your design standard and project risk profile. We support your team through every stage of that process.

Test planning and criterion selection: We help you define the failure criterion, target load, and test method before testing begins, so results are directly usable for your design framework.
Static Load Testing (SLT): We perform and interpret static load tests that produce direct load-settlement curves, giving you a clear and defensible basis for failure load determination.
Dynamic Load Testing (DLT) with signal matching: Our engineers use AllWave-DLT software and decades of experience to produce reliable capacity estimates from dynamic tests, with clear communication of the associated bandwidth and applicable safety factors.
Rapid Load Testing (RLT): Where dynamic testing alone is insufficient and static testing is logistically impractical, we offer Rapid Load Testing as a direct method that eliminates stress wave effects and improves accuracy.
Pile integrity assessment: When test results raise questions about structural condition, we combine capacity testing with integrity testing to give you a complete picture of pile performance.
Inconclusive result review: If your existing test data is ambiguous or disputed, we provide independent technical review and recommendations for supplementary testing.

If you are planning a pile testing program or need help interpreting existing results, contact our team to discuss your project requirements.