Slab Cracks Appear Weeks After Pour: Shrinkage vs Settlement Decision Tree

Why this matters

A crack that opens three to six weeks after a slab pour is the single most common callback in flatwork, and it carries real liability because the owner reads any crack as a defect. The diagnostic job is to separate benign drying-shrinkage cracking, which is a material property of every Portland cement mix and is almost never a structural problem, from settlement cracking, which signals a failed subgrade or a void that will keep moving and eventually telegraph through tile, break utility lines, or drop a load-bearing edge. Getting this wrong in either direction is costly: calling a settlement crack "normal shrinkage" leaves a moving defect under warranty, while ripping out a slab over hairline shrinkage burns margin and trust. The distinction is readable in the field from crack geometry, location relative to joints and loads, elevation change across the crack, and the pour history. This tree walks the branch points a senior finisher uses before committing to a remediation path.

Symptom presentation

Drying-shrinkage cracks are typically narrow at the surface (often under 1/16 inch), run in roughly straight or gently meandering lines, and frequently originate at re-entrant corners, around column blockouts, or wherever a control joint was missed, shallow, or spaced too far apart. They appear within the first weeks as the slab loses mix water, and crucially they show no vertical offset: drag a straightedge across the crack and both sides sit at the same elevation. Settlement cracks, by contrast, often show a measurable lip, one side has dropped relative to the other, and they tend to track over a buried trench, a recently backfilled utility line, a soft spot in the subgrade, or the edge of a slab that cantilevers past compacted fill. Settlement cracks may also widen over time and can run diagonally toward a corner that is sinking.

Quick checks

Before branching, gather five data points. First, run a straightedge or string line across the crack and feel for vertical offset; note the offset in 1/32 inch increments. Second, measure crack width with a crack comparator card at several points along its length. Third, map the crack against the control-joint layout and any blockouts, columns, or re-entrant corners. Fourth, pull the pour records: mix design and slump, ambient temperature and wind on pour day, time to first cure, and joint sawing timing. Fifth, locate buried utilities and the backfill/trench history under the crack line. Photograph everything with a scale reference.

Isolation tree

Start at vertical offset. If there is NO measurable offset and both faces sit flush, branch toward shrinkage. Confirm by checking joint spacing: ACI 302.1R guidance is roughly 24 to 36 times slab thickness in feet, so a 4 inch slab wants joints at about 8 to 12 feet. If the crack falls midway between joints that were spaced too far apart, or if joints were sawn late (after the slab had already cracked), you have classic restraint shrinkage and the diagnosis is essentially closed. A crack running straight from a re-entrant corner with no offset is also shrinkage driven by stress concentration, not settlement.

If there IS measurable vertical offset, branch toward settlement and isolate the cause. Map the crack against utilities: a crack tracking a backfilled trench points to under-compacted trench fill. A crack near a slab edge over fill points to inadequate subgrade compaction or fill that was placed without lift compaction. A crack with offset that is also actively widening week over week (re-measure at two-week intervals) confirms ongoing movement and rules out a stable shrinkage crack that merely looks offset because of curling.

Watch the curling trap: a slab edge that curls upward from differential drying can mimic settlement offset near joints and edges. Distinguish it by checking WHERE the high side is. Curling lifts the slab edge UP at joints and perimeter as the top dries faster than the bottom; true settlement drops a side DOWN over a soft zone. If the high side is the free edge and the slab rocks under load, suspect curling, not settlement.

Confirming diagnosis

For a suspected settlement crack, confirm the void. A nondestructive first pass is a chain drag or hammer sounding across the suspect zone: a hollow, drummy return over an area that sounds solid elsewhere indicates a void under the slab. To quantify subgrade support, a plate load test or a dynamic cone penetrometer reading in an exposed area adjacent to the crack tells you whether bearing capacity is deficient. For shrinkage confirmation, core or chip the crack and inspect depth and aggregate fracture: shrinkage cracks usually penetrate full depth on a thin slab and the crack passes around aggregate near the surface, with no evidence of a moving plane. Document slump and mix water against the ticket; a high-slump, water-added load shrinks more and cracks more.

Remediation

Shrinkage cracks that are tight, flush, and stable are cosmetic. Rout and seal with a semi-rigid polyurea or epoxy joint filler for interior floors that will see hard-wheel traffic per ACI 302.1R, or seal with a flexible sealant on exterior flatwork. Add or deepen control joints if a pattern of mid-panel cracking shows the layout was deficient, so future panels relieve correctly. For settlement cracks with a confirmed void, the slab must be re-supported: polyurethane or cementitious slab-jacking to fill the void and re-level, or remove and replace the affected panel with corrected subgrade compaction to 95 percent standard Proctor and proper trench backfill in compacted lifts. Never simply fill a settlement crack and walk; the void will reopen it.

References

• ACI 302.1R, Guide to Concrete Floor and Slab Construction (joint spacing, curling, sawing timing)
• ACI 224.1R, Causes, Evaluation, and Repair of Cracks in Concrete Structures
• ASTM C1583, Standard Test Method for Tensile Strength of Concrete Surfaces (overlay/repair bond)
• ASTM D6951, Standard Test Method for Use of the Dynamic Cone Penetrometer in Shallow Pavement Applications
• ASTM D1557, Test Methods for Laboratory Compaction Characteristics (modified Proctor, subgrade compaction targets)