Compressive strength testing of concrete

Compressive strength is the single most-quoted property of any concrete pour. The strength class on the design drawing — C25/30, C32/40, C40/50 — is shorthand for a specified characteristic compressive strength that the structure relies on. Verifying that strength on as-built concrete is one of the most important things construction NDT and laboratory testing do. Here is how it is measured, what the numbers mean, and what to look for in the deliverable.

What “strength” actually is

Compressive strength is the maximum load per unit cross-sectional area that a concrete sample can sustain before failure in compression. It is measured in megapascals (MPa) or, equivalently, newtons per square millimetre (N/mm²).

A strength class such as C32/40 specifies two values: the cylinder strength (32 MPa) and the cube strength (40 MPa) at 28 days. The two correspond to slightly different geometries of test sample. Both are widely used in the UK; cube strength is the historic norm, cylinder strength is more common in mainland Europe.

The “characteristic strength” is the strength below which only 5% of the population would fall — a statistical floor, not the average. Real concrete is variable; the design specification accounts for that variability.

How strength is measured

There are several methods to measure or estimate compressive strength:

Cube testing during construction. Cubes are cast from each significant pour, cured under controlled conditions, and tested at 7 and 28 days in a UKAS-accredited laboratory. This is the standard quality control during construction.

Core extraction and testing on existing concrete. A cylinder of in-situ concrete is extracted, prepared, and tested in the same way as a cast cube, with adjustments for the geometry and any defects in the core. This is the gold-standard method for measuring strength on existing concrete.

Pull-out testing. A calibrated in-situ test that produces a quantitative strength estimate. Defensible for most engineering work; less definitive than core testing.

Rebound hammer testing. A surface-hardness measurement that correlates loosely with compressive strength. Useful for screening; not definitive.

Ultrasonic pulse velocity. Velocity through the concrete correlates with stiffness and, indirectly, with strength. Used as part of multi-method assessment.

The right method depends on the brief. For definitive strength on existing structures: core testing. For broad characterisation: pull-out plus rebound. For quality control on new work: cubes alongside in-situ pull-out.

What the deliverable should contain

A defensible compressive strength assessment deliverable includes:

• A statement of brief.
• The sampling plan and rationale.
• A pre-drill scan record for any cores extracted.
• Photographs of every sample.
• UKAS-accredited laboratory test certificates.
• Statistical analysis of the results.
• The estimated characteristic strength at the agreed confidence level.
• Comparison against the design specification.
• A narrative interpretation by a qualified engineer.
• A surveyor / engineer sign-off.

The characteristic strength estimate is the headline number. It is what an engineer uses in subsequent calculations. The estimation method (typically the standard k-factor approach or a bespoke statistical analysis) should be stated and defensible.

Why a single test is not enough

Compressive strength on as-built concrete is variable. A single core, or a single pull-out test, gives a single point measurement. A defensible characteristic strength estimate needs:

• Enough samples to support a statistical estimate (typically 6–10 minimum, depending on element and method).
• Random or representative sampling, not cherry-picked points.
• Treatment of any outliers in line with documented method.
• A confidence level appropriate to the engineering decision.

A test programme that produces only a few values, or values from non-representative locations, cannot support a defensible characteristic strength estimate. The engineer using the data should be involved in the sampling plan.

Common mistakes

• Mistaking mean strength for characteristic strength. They are different statistical quantities. The characteristic strength is significantly lower than the mean for a typical population.
• Inadequate sample size. Small samples produce wide confidence intervals. The number of tests must match the engineering question.
• Mixing methods without calibration. Combining pull-out, rebound, and core results without proper calibration produces a confused dataset.
• Skipping reinstatement. Cores must be reinstated; pull-out tests leave a small visible mark; both should be addressed in the deliverable.
• Reporting without interpretation. Raw test numbers without statistical analysis and without engineering interpretation are not a defensible deliverable.

Where compressive strength fits in a wider assessment

For structural assessment of an existing concrete structure, compressive strength is one of several inputs:

• Reinforcement layout and cover (from GPR and ferro scanning).
• Compressive strength (from this kind of programme).
• Durability properties (carbonation, chlorides) where relevant.
• Geometry (from LiDAR or measured survey).
• Loading and environment.

The engineer combines these into a structural assessment. A concrete strength programme that ignores the wider context (because, for example, the engineer was not consulted on the sampling plan) often produces data that is technically correct but inadequate for the assessment it was meant to support.

Practical advice

If you are commissioning compressive strength testing for the first time:

1. Engage the structural engineer in the sampling plan.
2. Specify UKAS-accredited laboratory testing for any cores.
3. Plan an adequate sample size for the question.
4. Insist on pre-drill clearance before coring on reinforced elements.
5. Receive the deliverable with a clear characteristic strength estimate, not just raw numbers.

Compressive strength testing, done well, is the bedrock of defensible structural assessment on existing concrete. Done badly, it produces numbers that confuse rather than clarify. The brief, the sample, and the statistics are what separate the two.