Rebound hammer strength estimate

The rebound (Schmidt) hammer drives a spring-loaded mass against a plunger held on the concrete and measures how far the mass rebounds. The rebound number is governed by the surface hardness and stiffness of the top few centimetres, which correlate loosely with compressive strength. Loosely is the operative word: the test is a rapid comparative survey tool, not a strength measurement, and every standard that covers it says so.

Two corrections are needed before any conversion. First, the direction of impact: gravity acts on the moving mass over both the impact and the rebound stroke, so a hammer pointing upwards at a soffit reads high and one pointing down at a slab reads low. The manufacturer's correction is added to the reading and is itself a function of the reading: it is negative for upward impact, from about −5.4 at R = 20 to −2.3 at R = 60 for vertically upward, and positive for downward impact, from about +3.4 at R = 20 to +1.7 at R = 60 for vertically downward. Second, carbonation: a carbonated surface layer is harder than the concrete beneath and inflates the rebound number, sometimes by 50% on old structures, which is why the depth of carbonation should be checked with phenolphthalein on a fresh break and a reduction factor applied.

The conversion from corrected rebound number to strength is the weak link. The generic curve supplied with a Type N hammer was established on particular mixes, aggregates, ages and moisture conditions, and applying it blind to site concrete can be 25% out or worse. Both BS EN 12504-2 and ASTM C805 require the relationship to be established for the concrete under test — normally by taking cores at a spread of rebound values and fitting a regression. Enter that regression here and the estimate becomes defensible; use the indicative curve and it is a screening figure only.

Test discipline matters as much as the arithmetic. Take at least nine (EN) or ten (ASTM) readings on a ground, dry, smooth surface at least 25 mm from an edge and away from reinforcement, discard outliers per the standard, and use the median. Never test on a wet, painted, rendered or friable surface, over a void, or on an element thinner than about 100 mm unless it is rigidly supported.

R_corrected = R + direction correction (from the manufacturer's table)
f_c = curve(R_corrected) × carbonation factor
site regression: f_c = a × R_corrected^b