New Slab Curls at Edges: Mix vs Cure vs Subgrade Decision Tree

Why this matters

Curling, the upward warping of slab edges and corners away from the subgrade, is a structural and serviceability problem that drives callbacks ranging from rocking joints and broken joint edges to forklift damage and tile cracking. It happens because the top of the slab dries and shrinks faster than the bottom, so the slab tries to assume a dish shape and lifts its free edges. The reason this matters to diagnose precisely is that curling has three controllable inputs: the mix (water content, shrinkage potential, aggregate), the curing regimen (how fast the top dries relative to the bottom), and the subgrade and vapor-retarder detail (which controls the moisture gradient through the slab). The remedy differs by cause, and on a slab that already curled you must also decide whether to grind joints flat, fill voids, or accept it. Reading the slab tells you which input failed: how much it lifted, where, how the joints behave under load, and what the placement records show.

Symptom presentation

Curling presents most strongly at slab corners, then along free edges and at sawn or formed joints, where the unrestrained edge can lift. You will see joint edges sitting proud, a measurable gap under a straightedge laid across the joint, and often a hollow, drummy sound when the lifted edge is sounded because it has separated from the subgrade. Under traffic, curled joints rock and spall: hard wheels hammer the unsupported lip and break it down. Curling is worst on thin slabs, on slabs placed over a low-permeability vapor retarder directly under the concrete (which holds water at the bottom while the top dries), on high-water/high-shrinkage mixes, and on slabs that were poorly cured so the top dried rapidly. The lift typically develops over the first weeks to months as the moisture gradient establishes and can worsen seasonally in dry indoor environments.

Quick checks

Measure the curl: lay a 10 foot straightedge across the joint or from the edge inward and measure the gap at the lifted edge in 1/16 inch increments; record several locations. Sound the lifted edges for hollowness to confirm separation from subgrade. Check whether joints rock under foot or wheel load. Pull the records: mix slump and water-cement ratio, aggregate size and gradation, slab thickness, joint spacing, the curing method and how soon it was applied, and the subgrade/vapor-retarder detail, specifically whether a vapor retarder was placed directly under the slab or with a sand/blotter layer over it. Note the indoor humidity and HVAC status if interior. Photograph straightedge gaps with a scale.

Isolation tree

Curling is almost always a combination, so the tree ranks contributors rather than picks one. Start at the moisture gradient driver: was a low-permeability vapor retarder placed in direct contact with the bottom of the slab? If yes, that single detail strongly promotes curling because it blocks bottom drying while the top dries freely, maximizing the gradient. This is a top suspect on interior slabs that curled badly. Next branch to curing: was the top cured wet or with compound promptly and kept moist for the specified duration, or did it dry quickly in wind/low humidity? Rapid top drying with a wet bottom is the direct engine of curl, so a missed or short cure pushes the diagnosis toward cure failure.

Next branch to mix shrinkage: high water content, high-slump water-added loads, and excessive paste/fines all increase drying shrinkage, which increases curl magnitude. Pull the ticket and any field water adds. A high w/c mix on a thin slab with a poor cure is the textbook severe curl. Finally branch to geometry: thin slabs curl more for the same gradient, and widely spaced joints leave longer unrestrained edges. If joints were spaced too far apart per ACI 302.1R (roughly 24 to 36 times thickness in feet), each panel has more length to lift. Rank the contributors you found; the remediation addresses the dominant ones and prevents recurrence by fixing the detail (vapor retarder placement, cure) on the next pour.

Confirming diagnosis

Confirm the moisture gradient by checking how the curl responds to wetting the surface: re-wetting the top can partially relax curl because it swells the dried top back toward the bottom's moisture state, confirming a drying-gradient mechanism rather than settlement. Confirm subgrade separation (void under the lifted edge) by sounding and, if needed, drilling a small test hole to measure the gap. Confirm shrinkage potential of the mix against records or, going forward, by drying-shrinkage testing of the mix per ASTM C157. Confirm slab moisture state with relative humidity probes per ASTM F2170 if a coating is planned over the curled floor. Rule out true settlement by checking the SHAPE: curling lifts edges UP relative to panel centers; settlement drops a side DOWN over a soft zone. The high-edge, dished-panel signature is curling.

Remediation

On an already-curled slab in service, restore joint performance: diamond grind the proud joint edges flat to stop rocking and spalling, then fill the void under the curled edge with polyurethane or cementitious slab-jacking grout to re-support the lip so traffic no longer hammers an unsupported edge. Semi-rigid joint filler per ACI 302.1R supports the joint arris against hard-wheel traffic. For severe, ongoing curl, sawing additional intermediate joints reduces panel length and curl per panel. Prevent recurrence on future pours: place the vapor retarder correctly for the application (a blotter layer detail where appropriate, or accept that direct-contact retarders trade vapor protection for more curl), cure the top promptly and for the full duration per ACI 308.1, lower mix water and shrinkage, and tighten joint spacing on thin slabs.

References

• ACI 302.1R, Guide to Concrete Floor and Slab Construction (curling mechanism, joint spacing, joint fillers)
• ACI 308.1, Specification for Curing Concrete
• ASTM C157, Test Method for Length Change of Hardened Hydraulic-Cement Mortar and Concrete (drying shrinkage)
• ASTM F2170, Test Method for Determining Relative Humidity in Concrete Floor Slabs Using in situ Probes
• ACI 360R, Guide to Design of Slabs-on-Ground (curling, vapor retarder placement effects)