Pile Load Transfer From Static Load Test Inversion-Article

Article

Pdf Summary

The paper develops an inversion method for estimating how load is transferred along a deep foundation using only the pile-head load–movement curve. This is intended for static load tests where strain-gage or telltale measurements are missing, incomplete, unreliable, or affected by residual loads.<br /><br />The method uses a nonlinear finite-element “t–z” model: nonlinear springs represent shaft resistance and a “q–z” curve represents toe resistance. A genetic algorithm adjusts the model parameters to minimise the difference between measured and predicted pile-head movements. The resulting parameters are then used to calculate load distribution with depth. Required site information is limited mainly to pile geometry and soil types and layer thicknesses.<br /><br />The approach was tested against three instrumented full-scale tests: drilled shafts in sand and residual soil, and a driven pile in clay. In the first two cases, the estimated load distributions generally agreed well with measurements. For the driven pile, residual loads had distorted the strain-gage-based distribution; the inversion produced results in excellent agreement with an independently interpreted “true” distribution that accounted for those loads.<br /><br />The results suggest that pile-head response can provide useful information about load transfer, including residual-load effects, when direct instrumentation is unavailable or questionable. However, inverse solutions are non-unique, and results depend on the model assumptions and chosen parameter ranges. The method is therefore best viewed as a practical estimation tool rather than a replacement for reliable instrumentation.

Keywords

pile load transfer

deep foundation

inversion method

t-z model

q-z curve

genetic algorithm

pile-head load-movement

static load test

shaft resistance

toe resistance