Concrete ultrasonic pulse velocity

Ultrasonic pulse velocity testing puts a low-frequency compression wave (typically 50–150 kHz) through concrete between a transmitter and a receiver and measures the transit time. Velocity is simply v = L/t, but it is a remarkably informative number: the wave travels through the paste and aggregate skeleton, so anything that interrupts that skeleton — voids, honeycombing, cracking, frost or fire damage, incomplete compaction — slows the pulse or forces it to travel around, and the velocity falls.

The classification most inspectors quote comes from Whitehurst and is reproduced in IS 13311: above 4.5 km/s excellent, 3.5–4.5 good, 3.0–3.5 doubtful, below 3.0 poor. It is a guide, not an acceptance criterion, and it is calibrated to ordinary structural concrete of about 2400 kg/m³. Velocity depends strongly on aggregate type and content, moisture state, reinforcement in the path and age, so a single reading in isolation proves very little. Both BS EN 12504-4 and ASTM C597 make the point directly: the value of UPV lies in comparison — against a reference area of the same mix, or in a grid of readings where the outliers identify where to investigate further.

Path arrangement matters. Direct transmission, transducers opposite each other, is the only arrangement the classification really applies to. Semi-direct (adjacent faces) is acceptable with the true straight-line path measured. Indirect or surface transmission, both transducers on one face, samples mainly the surface layer and typically reads 5–20% lower; it is used where only one face is accessible, and then only for comparison or for estimating crack depth, never for classification.

Velocity also gives the dynamic modulus of elasticity through E_d = ρ·v²·(1+ν)(1−2ν)/(1−ν), the standard relation in BS 1881-203. The dynamic modulus is typically higher than the static modulus measured on a cylinder, so do not substitute one for the other in a structural check. And UPV does not measure strength: any strength inference needs a site-specific correlation against cores, ideally combined with rebound hammer readings in the SonReb approach.