Pull-off testing is the only quantitative way to prove your prepared concrete will actually hold the coating. CSP verification tells you the geometry is right. Moisture testing tells you the slab is dry. Pull-off testing tells you the bond will mechanically survive — the rigid epoxy system needs at least 1.5 MPa of tensile adhesion or the EN 1504-2 warranty is automatically void.
Why the Pull-Off Test Exists
A clean, dry, profiled slab can still fail under a coating if the top millimeter of concrete is itself weak. The most common reason is laitance — the thin, cement-rich, fines-rich layer that floats to the top during finishing. Laitance can look perfectly profiled after grinding or shot blasting but cohesively fail under shear or thermal cycling, dragging the coating off with it. Carbonated surface concrete behaves the same way; so does heat-damaged concrete near a fire or near a long-standing oil leak.
Visual inspection cannot detect a 0.5 mm weak layer. The pull-off test can — and it does so with a single number you can put in a project file and reference in a warranty claim two years later.
ASTM D7234 — The American Standard
ASTM D7234 — "Standard Test Method for Pull-Off Adhesion Strength of Coatings on Concrete Using Portable Pull-Off Adhesion Testers" — defines the test procedure used worldwide. Procedure summary:
- Site preparation. Surface dry, clean, at ambient temperature. If testing the bare substrate before priming, the area must be representative of the prepared floor — not a freshly scrubbed patch.
- Dolly attachment. A 50 mm aluminum or stainless steel dolly (sometimes called a loading fixture) is bonded to the surface with a fast-cure adhesive. Methacrylate (5-minute) or two-part rapid epoxy is standard. Curing time per the adhesive TDS — typically 30–90 minutes.
- Scoring. A core drill or hand tool scores the perimeter of the dolly through the coating (or through the surface concrete if testing bare substrate) to a depth of ~5 mm. This isolates the test area from the surrounding floor so the pull-off force acts only on the bonded plug.
- Pull. A portable hydraulic or motorized tester engages the dolly and applies a perpendicular tensile load at a controlled rate (~0.5–1.0 MPa/s). The instrument records the peak load at failure.
- Reporting. Peak load is converted to pull-off tensile strength in MPa (force ÷ dolly area). Failure mode is recorded: substrate cohesive, adhesive at coating/substrate interface, coating cohesive, glue failure (dolly came off the coating).
EN 1542 — The European Standard
BS EN 1542 — "Products and systems for the protection and repair of concrete structures — Measurement of bond strength by pull-off" — is the European equivalent. Procedure is functionally identical to ASTM D7234; some details differ:
- EN 1542 specifies a 50 mm dolly diameter and a circular partial core to ~5 mm depth.
- Loading rate is specified as 0.05 ± 0.01 MPa/s, slower than ASTM D7234.
- The result is reported alongside the failure mode and the depth at which the failure occurred. EN 1542 distinguishes "100% cohesive in substrate at 0–10 mm depth" from "cohesive at 10+ mm" — the former indicates surface weakness, the latter indicates structural concrete strength is the limiting factor.
For practical purposes, a tester certified to one standard will read identically to a tester certified to the other. Most modern instruments (DeFelsko PosiTest AT-A, Proceq DY-2) are rated for both.
Acceptance Criteria
EN 1504-2 — the European standard for surface protection systems for concrete — defines the pull-off acceptance thresholds for coatings:
- Rigid systems (epoxy 100% solids, hard-cured PU): ≥ 1.5 MPa (the most common spec).
- Flexible systems (elastomeric PU, MMA with elastic component): ≥ 0.8 MPa.
- Cementitious overlays (microcement, screeds bonded to substrate per EN 13813): ≥ 1.0 MPa.
- Polymer overlays > 6 mm: ≥ 1.5 MPa with substrate cohesive failure as the mode of failure.
Sample 3 dollies per 100 m² minimum. The worst result governs — averaging is not acceptable under EN 1504. If one of three dollies fails to threshold, the area is re-prepared, not just averaged into the report.
The Four Failure Modes — What Each One Means
1. Substrate Cohesive Failure (deep)
The failure plane is in the concrete itself, more than 10 mm below the surface. This is the ideal result: the substrate is so strong that the limit is the concrete's own tensile strength. A reading of 2.5+ MPa with deep cohesive failure means the coating will never fail at the bond; if it fails, the concrete itself will be the weak link.
2. Substrate Cohesive Failure (surface, 0–5 mm)
The failure plane is in the top millimeters of concrete. This indicates a weak surface zone — laitance, carbonation, microcracking from over-scarification, or chemical damage. If the reading is above threshold (1.5 MPa for rigid), the coating will likely hold for the warranty term but the safety margin is thin. If below threshold, re-prepare: shot blast or grind another 2–3 mm and retest.
3. Adhesive Failure at Coating/Substrate Interface
The coating pulls cleanly off the substrate with no concrete attached. This indicates a bond problem — either the prep was insufficient (under-profiled, contaminated) or the primer was applied outside its open time. A reading below threshold here is a hard fail: re-prepare the surface and re-prime; do not coat over.
4. Glue Failure (dolly bond fails)
The adhesive holding the dolly to the surface fails before the substrate or coating does. The test is invalid — the recorded value is a lower bound on the true pull-off strength. Repeat with a different adhesive (fresh methacrylate, longer cure, or a slow-cure epoxy if the surface temperature is high) and re-pull.
When to Run the Test
- Pre-coat verification: after CSP verification, before priming. Test on bare prepared substrate. This is the warranty-critical test — it proves the prep work delivered a coatable surface.
- Post-coat warranty record: 28 days after the system is fully cured. Test through the coating. Provides a baseline for any future warranty claim.
- Dispute resolution: if a coating later fails or shows blistering, a pull-off test at and around the failure area diagnoses whether the failure mode is adhesive (bond) or cohesive (substrate weakness).
The Three Testers You Will See in Israel
- DeFelsko PosiTest AT-A (automatic). Motorized hydraulic, applies a controlled loading rate per ASTM D7234 / EN 1542. Records and stores test results with date and location. Most common in commercial floor contracting.
- Proceq DY-2. Swiss-made, dual-standard (ASTM and EN), bluetooth data transfer. Premium price but lab-grade accuracy.
- DeFelsko PosiTest AT-M (manual). Hand-pumped hydraulic; cheaper but loading rate is operator-dependent, which the standards specifically discourage.
For coastal IL projects, the automatic units are non-optional. Hand-pumped testers cannot maintain the loading rate when the operator is fatigued, and the warranty paper trail demands repeatability that only the motorized testers deliver.
Common Mistakes
- Testing too soon. The dolly adhesive must reach its full bond strength. Methacrylate at 30 °C cures faster than at 15 °C — follow the TDS, not the clock.
- Skipping the perimeter score. Without scoring, the pull engages a larger area than the dolly footprint and reads artificially high. The result is uninterpretable.
- Single test. Three dollies minimum per 100 m². Single readings are not statistically meaningful and not accepted under EN 1504.
- Testing on a wet surface. Surface moisture interferes with the dolly adhesive cure. Surface must be dry to ASTM D4263 plastic-sheet criterion at minimum.
Read next: ICRI CSP 1–10: which profile for which coating · Mechanical preparation methods compared · Moisture testing: ASTM F2170, F1869, calcium chloride, Tramex.
Sources
- ASTM D7234-21 — Standard Test Method for Pull-Off Adhesion Strength of Coatings on Concrete Using Portable Pull-Off Adhesion Testers.
- BS EN 1542 — Products and systems for the protection and repair of concrete structures: Measurement of bond strength by pull-off.
- BS EN 1504-2 — Surface protection systems for concrete (acceptance criteria for coatings).
- DeFelsko — PosiTest AT-A and AT-M technical documentation.
- Proceq — DY-2 family technical data sheet.