Raft Foundation on Partially Saturated Clay After Groundwater Drawdown
A five-storey RC building with a 25 m x 18 m raft (thickness 900 mm) sits on a 12 m thick layer of stiff-to-very-stiff silty clay over dense sand. The design assumed a groundwater table at 1.5 m depth and an allowable bearing pressure of 180 kPa. Two years after construction, dewatering for a neighboring excavation lowered the regional water table to 9 m. Monitoring shows 22 mm of settlement concentrated near the shared boundary, with differential settlement of 14 mm across the raft. The structural engineer reports new cracks in the ground-floor beams at the column faces. The geotechnical report's consolidation tests were run on samples saturated at the original water table, giving a compression index Cc = 0.28 and recompression index Cr = 0.06.
1. Explain the mechanism driving the settlement and assess whether the observed 22 mm is consistent with a drawdown-induced consolidation of the clay layer.
Read the worked answer
Mechanism. Lowering the water table from 1.5 m to 9 m reduces pore pressure within the clay over that interval. The effective stress increases by roughly Δσ' = γw × Δh ≈ 9.81 × 7.5 ≈ 74 kPa in the dewatered zone, and less below 9 m. That increase in effective stress produces consolidation settlement in the compressible clay.
Order-of-magnitude check. For a 12 m clay layer with the water table starting near the surface, initial effective stress at mid-depth (6 m) is about (γsat − γw) × 6. Using γsat ≈ 19 kN/m³: σ'v0 ≈ (19 − 9.81) × 6 ≈ 55 kPa. With the water table drawn to 9 m, the upper 7.5 m loses buoyancy, so σ'vf rises substantially - a rough estimate puts the mid-depth increase around 35–50 kPa depending on the stress distribution.
Using Cr = 0.06 for the reloading path (the clay was already consolidated under the original profile) and e0 ≈ 0.8, H = 12 m:
s = [Cr·H/(1+e0)] · log10(σ'vf/σ'v0) ≈ [0.06 × 12/1.8] × log10(100/55) ≈ 0.40 × 0.26 ≈ 0.10 m
That upper bound is larger than the observed 22 mm, but the estimate is sensitive to the actual stress increment, the depth of the dewatered zone, and the fact that only part of the layer is affected. 22 mm is entirely plausible - the observed value sits comfortably within the range the calculation brackets. The key point for the exam is that the settlement is drawdown-induced consolidation, not bearing-capacity failure: the raft pressure did not change, the effective stress in the soil did.
2. The cracks are at column faces in the ground-floor beams. Is this consistent with the reported 14 mm differential settlement, and what investigation would you order before recommending remediation?
Read the worked answer
Consistency check. A 14 mm differential over the span between two columns induces angular distortion β ≈ Δ/L. For a typical 6 m bay, β ≈ 14/6000 ≈ 1/430. Published damage thresholds (Skempton & MacDonald; Boscardin & Cording) place 1/500–1/300 in the zone where cracking in finishes and lightly reinforced elements begins. Cracks concentrated at column faces - where the beam is restrained and moment demand is highest - are exactly the pattern expected from differential settlement, not from overload.
But column-face cracking also occurs from shrinkage, from hogging moment at supports, and from a soft-storey effect. Do not accept settlement as the sole cause without evidence.
Investigation to order:
- Level survey of the raft and floor slabs on a stable benchmark, repeated over 4–8 weeks to establish whether settlement is ongoing or arrested. Ongoing settlement changes the remediation strategy entirely.
- Crack mapping with tell-tales or DEMEC gauges to measure whether cracks are active.
- Piezometers in the clay and sand to confirm the current water table and whether recovery is occurring.
- Additional boreholes with high-quality sampling and consolidation tests on reconstituted-at-current-stress specimens, because the original Cc/Cr were measured at the old water table.
- Structural review of the cracked beams: check reinforcement, measured crack widths against code limits, and whether the beams need strengthening independent of the foundation.
Judgement. If settlement has arrested and cracks are stable and within serviceability limits, monitoring plus crack repair may suffice. If settlement is ongoing, remediation (compensation grouting, or controlled re-wetting) is needed before structural repair - repairing cracks over a moving foundation is wasted money.