Plate load test and limitations

The plate load test is an in-situ test used to find the ultimate bearing capacity of soil and the settlement of a footing at a given load. A rigid steel plate is loaded in stages and the settlement is recorded at each stage.

Procedure

1. A pit is dug to the foundation level, with width at least 5 times the plate size. A square or circular plate of 30 to 75 cm (about 25 mm thick) is bedded on the levelled soil.
2. A load is applied through a hydraulic jack acting against a reaction (a loaded platform or a kentledge, or an anchored truss). The jack load is measured by a pressure gauge or proving ring.
3. Load is applied in increments of about one-fifth of the estimated safe load. Settlement is read on dial gauges (usually 2 to 4, supported on an independent datum beam) at set time intervals, until the rate of settlement falls to $0.02$ mm/min (or 24 h in clay) before the next increment.
4. The test continues until failure or until the settlement is 25 mm or more (about 1.5 times the proposed design load).

        Jack + gauge
           |   Dial gauge
  ====Reaction beam====
           |
        [Plate]    <- test pit
  ///////////////

Results

• The load–settlement curve is plotted. The ultimate load is the point of failure, or the intersection of tangents on the initial and final parts. For sand, $q_u(\text{footing}) = q_u(\text{plate}) \times B/B_p$ and settlement $S = S_p\left[\dfrac{B(B_p+0.3)}{B_p(B+0.3)}\right]^2$. For clay, $q_u$ is the same for plate and footing.

Limitations

• The stressed depth is small (about $2B_p$), so deeper weak layers are not tested.
• Only immediate settlement is measured, not long-term consolidation in clay.
• Size effects: results cannot be extrapolated directly to large footings.
• Represents only a small area, and is expensive and slow.