Overview
A pull-off test measures how firmly a coating is bonded. A metal disc is glued to the cured film and pulled straight off. The force at failure, and the place the failure happened, are the result.
What the test measures
The test measures tensile bond strength: the stress needed to separate the coating from what it is on. It is reported in N/mm², which is the same unit as MPa.
It is the standard adhesion test for thick coatings and for anything applied to concrete.
How the test is carried out
The general method for coatings is ISO 4624. For products on concrete, EN 1542 is commonly used.
- A metal disc, called a dolly, is glued to the cured coating.
- On concrete, the coating is cut around the dolly down to the substrate, so only the area under the dolly is tested.
- A tester pulls the dolly at right angles to the surface at a steady rate until something breaks.
- The stress at failure is recorded, with a description of the break.
How to read the result
The number alone is half the result. Where the break happened tells you what the number means.
| Where it broke | What it tells you |
|---|---|
| In the substrate | The concrete failed before the bond did. The bond is stronger than the figure shown. |
| Between coating and substrate | The bond itself failed. The figure is the adhesion. |
| Within the coating | The coating pulled apart inside itself. The figure is the strength of the film, and the bond is stronger than that. |
| In the glue | The glue let go first. The test shows only that the bond is at least this strong. |
A result of "2.5 N/mm², concrete failure" therefore says more than the figure suggests. The coating stayed on and the concrete gave way.
For comparison, EN 13813, the European standard for screed materials, declares bond strength in classes such as B1.5 and B2.0. The number is the minimum bond in N/mm².
What it means in use
Adhesion is what keeps a coating down when it is loaded sideways. Tyres turning on the spot and pallets being dragged both try to shear the film off the surface. So does the movement of a slab as it warms and cools.
What the result does not tell you
A published figure comes from a prepared test surface. On site, adhesion depends first on preparation. Laitance, dust, oil, old sealers and moisture in the slab all reduce it, whatever the coating.
The test also cannot exceed the strength of what is underneath. On weak or friable concrete the substrate breaks at a low load, and no coating can raise that.
For thin films on metal and wood, see Adhesion (Cross-Cut).
On our product pages
Where a product has been tested, its page shows the result, the method and, where the method has one, the rating. Results published on this site come from internal testing to the stated method. They are given for technical reference and comparison.
If a figure you need is not shown, ask our technical team.


Share:
Slip Resistance (Ramp Test)
Light Reflectance Value (LRV)