Substrate for a polymer floor: what to check before pricing
A conversation about a polymer floor almost always starts with the material. And ends with the concrete. Nine coating failures out of ten are not the resin but the substrate beneath it: somewhere the laitance was not removed, somewhere the moisture was not measured, somewhere the layer was laid on a cold slab on a damp morning.
The most frustrating part is that substrate preparation is the most underestimated line of the estimate. It is either not budgeted at all, or budgeted “by analogy”. And when it turns out on site that the top three millimetres of concrete crumble in your hand, the argument begins over who pays for it. That argument is settled before the work starts — with five measurements.
Five checks
1. Compressive strength — at least 25 N/mm²
This is what the whole structure stands on. Checked with a sclerometer (Schmidt hammer) over a grid of points, or from the concrete’s data sheet if the slab is new and documents exist. If the concrete is weaker, the polymer will not save it: the coating will be pressed in together with the substrate.
2. Pull-off strength — at least 1.5 N/mm²
This is the check skipped most often, yet it is decisive. Measured with a pull-off tester: a steel dolly is glued to the substrate and pulled off while the force is recorded. The logic is simple: a coating cannot hold on stronger than what it is glued to. If the top layer of concrete is weak, the failure will run through it — and no resin will change that.
It matters to look not only at the number but at the nature of the failure. It came off through the concrete a couple of millimetres deep — there is a weak surface layer that must be removed. It came off through the glue — the measurement is invalid, redo it.
3. Moisture — no more than 4 % by the CM method
The carbide method: a concrete sample is crushed, mixed with calcium carbide in a sealed vessel, and the moisture is calculated from the pressure of the released gas. Electronic moisture meters give an estimate, but CM is what is recognised at acceptance.
And separately — moisture from below. If there is no waterproofing under the slab and groundwater is near, the polymer coating will act as a membrane: it traps vapour, pressure builds up, and the coating is torn off in blisters. This is not a material defect. It is a detailing defect, and it must be solved before, not after.
4. Surface: flatness, roughness, cleanliness
Three things must be removed: laitance (a weak layer in itself), oil and grease stains (the polymer does not stick to them), and flaking, loose areas.
Flatness — per the project, usually with a two-metre straightedge. A thin-layer coating does not hide unevenness but emphasises it: a half-millimetre layer repeats the relief, including power-trowel marks.
5. Temperature and dew point
From +10 to +30 °C for air and substrate, relative air humidity no higher than 80 %, and — most importantly — the surface temperature at least 3 °C above the dew point.
The last is forgotten most often because it is invisible. The slab cooled overnight, the workshop warmed up in the morning, condensation formed on the concrete — a thin film of water nobody noticed. The coating lands on water and peels away in sheets a month later. The measurement is taken with a thermocouple hygrometer and takes a minute.
How the substrate is prepared
| Shot blasting | The main method. Steel shot knocks off the laitance and opens the pore, leaving an even rough profile. It works with a vacuum on a closed loop, the shot is collected and recirculated — important in an operating workshop. |
|---|---|
| Milling | For thick contaminated layers, old coatings, deeply ingrained oil. Removes more, leaves a coarse relief — usually a levelling screed is needed afterwards. |
| Grinding | For local fitting, edges, corners, and for inter-coat treatment if the overcoat window was missed. As standalone preparation under a coating it is unsuitable: a grinder smooths the surface instead of opening the pore. |
A phrase worth remembering: if a contractor offers to “run over it with a grinder” instead of shot blasting, he is cutting costs on the very thing the whole warranty depends on.
Repair before, not after
Cracks, potholes, expansion joints and abutments are closed before the main layer. The repair compound is made from the same epoxy binder with quartz sand: the compressive strength of such a screed can reach 90–95 N/mm² per EN 196-1 — higher than the surrounding concrete.
Separately about expansion joints: they are not filled over. A joint must remain a joint, otherwise it will show through the coating in the first season. It is opened out and filled with an elastic sealant, and the coating is cut along the joint.
What typical defects mean
It is useful to be able to read what has already happened — the defect almost always shows which of the five checks was skipped.
| Blisters over the whole area | Moisture or air from the substrate pores. Not primed, or primed over a damp surface. |
|---|---|
| Peeling in sheets, clean underside | Condensation: the coating landed on a film of water. The dew-point check was skipped. |
| Peeling with a thin layer of concrete | The laitance was not removed. Failure ran through the weak surface layer. |
| Craters and “fish eyes” | Oil or silicone on the substrate surface. |
| A crack repeating a crack in the slab | A rigid system on a moving substrate. An elastic — polyurethane — one should have been used. |
| Wear in tracks along traffic lanes | Thickness chosen without accounting for the equipment. Covered in the article “Self-levelling epoxy floor: how to choose the thickness for the load”. |
Handover checklist
What is worth recording in the report before work begins:
- compressive strength, N/mm², method and measurement points;
- pull-off strength, N/mm², and the nature of the failure;
- moisture by CM, %;
- waterproofing under the slab — yes / no / unknown;
- flatness, mm per 2 m;
- air temperature, substrate temperature, humidity, dew point;
- a list of areas requiring repair, with their areas.
Seven lines that settle nine arguments out of ten.
When it is easier to redo the substrate
Honestly about the limit. If the pull-off is below 1 N/mm² over most of the area, if the concrete crumbles a centimetre deep, if there is constant moisture from below — repairing costs more than a new screed. In such cases we say so straight away: a polymer floor laid on such a substrate is wasted money and a spoiled reputation, ours included.
Selecting a system for a specific site — epoxy floors and other types of coating; ask us for advice.