What causes a concrete slab to curl at the edges?

A moisture and temperature gradient between the top and bottom of the slab, which dry and cool at different rates and cause the top to shrink more, lifting the edge.

Is plastic shrinkage cracking the same as drying shrinkage cracking?

No. Plastic shrinkage cracking happens while the concrete is still workable, from fast surface evaporation. Drying shrinkage cracking happens after the concrete has hardened, as it continues losing moisture over time.

Can I tell from a crack's appearance whether it is structural?

Not reliably from a description alone — that judgement needs someone assessing the actual crack in person, its width, pattern and any movement over time, against the specific slab and subgrade.

Do saw-cut joints stop cracking altogether?

No, they control where drying shrinkage cracking happens rather than preventing it. The concrete still shrinks; the joint gives it a planned line to crack along instead of a random one.

Why concrete floors crack and curl

Cracking and curling are related but different problems, both driven mainly by moisture and temperature moving through the slab unevenly. This explains the mechanisms — it will not tell you whether a particular crack in your floor is cosmetic or structural, because that call belongs to whoever inspects it in person.

Last updated 2026-08-24

Curling: a gradient problem

Curling happens because the top and bottom of a slab dry and cool at different rates. The top surface, exposed to air and often finished denser by trowelling, loses moisture faster than the bottom, which sits against the subgrade and stays more protected. That uneven drying shrinks the top more than the bottom, and the slab edge lifts. A temperature gradient — a cool night after a hot day — can add to the same effect.

Contributing factors include excessive bleed water or a high water content in the mix, how far apart the joints are cut, curing practice, the subgrade's moisture condition, and whether a vapour barrier was used underneath. Closer joint spacing generally reduces the area of any one curling panel, which is part of why where joints are cut and how far apart is a design decision, not an afterthought — and why curling and joint-edge spalling are two symptoms of the same underlying movement rather than unrelated defects.

Cracking: two different mechanisms

Plastic shrinkage cracking happens while the concrete is still plastic, when surface moisture evaporates faster than bleed water can replace it — very fast in hot, windy, low-humidity conditions, which describes a lot of Australian pour days across the country's different climates. Drying shrinkage cracking happens later, after the concrete has hardened, as it continues to lose moisture and shrink. Saw-cut joints exist specifically to make this second kind of cracking happen at a chosen, controlled line instead of at random.

These are genuinely different mechanisms with genuinely different timing, even though both end up looking like a crack in a finished floor, and neither is specific to one type of slab construction over another — a slab-on-ground and a post-tensioned slab are both susceptible, for the same underlying reasons.

Moisture does double duty

Moisture drives curling through the gradient described above. It also causes a separate, slower problem: ongoing moisture vapour still working its way up from the subgrade through an uncoated or improperly coated slab, long after the pour has cured. That is not primarily a structural issue — it is a durability issue for whatever sits on top of the floor, and it is a large part of why a floor coating applied over a slab still emitting vapour can blister or lift regardless of which coating chemistry was chosen.

Cosmetic or structural: not a call this page will make

Some cracking and curling is cosmetic — narrow, stable, and not indicative of anything beyond normal shrinkage behaviour. Some is a sign of a genuine subgrade or structural problem: active movement, widening cracks, a corner rocking under load. Telling the two apart in a specific case is exactly the judgement a structural engineer or an experienced floor assessor makes in person, looking at width, pattern, movement over time and what is happening underneath — not something a general explanation of mechanisms can responsibly do for your particular floor. General structural tolerance for a concrete member is addressed within the Standards Australia catalogue, but no single Australian Standard sets out a warehouse-specific crack or curl acceptance threshold the way it does for other trades.

That distinction is also why planning treats the floor slab as effectively fixed once poured: cracking and curling are ongoing processes, not one-off events, and a slab that looked fine at handover can develop either over its working life. It is one more reason the building specification a warehouse is built to matters well beyond clear height and dock doors.

Common questions

What causes a concrete slab to curl at the edges?: A moisture and temperature gradient between the top and bottom of the slab, which dry and cool at different rates and cause the top to shrink more, lifting the edge.
Is plastic shrinkage cracking the same as drying shrinkage cracking?: No. Plastic shrinkage cracking happens while the concrete is still workable, from fast surface evaporation. Drying shrinkage cracking happens after the concrete has hardened, as it continues losing moisture over time.
Can I tell from a crack's appearance whether it is structural?: Not reliably from a description alone — that judgement needs someone assessing the actual crack in person, its width, pattern and any movement over time, against the specific slab and subgrade.
Do saw-cut joints stop cracking altogether?: No, they control where drying shrinkage cracking happens rather than preventing it. The concrete still shrinks; the joint gives it a planned line to crack along instead of a random one.

Part of our guide to what a warehouse is made of.