Cyclic shear testing maps how pile–silty clay interfaces weaken under repeated loading

Pile foundations do not fail only at the level of gross bearing capacity; they often begin to lose performance at the contact zone where concrete or steel meets soil. In this PLOS ONE study, that interface was examined under repeated shear to define how roughness, water content, displacement amplitude, normal stress, and loading mode reshape the mechanical response of pile–silty clay systems. The work matters because silty clay is water-sensitive and structurally distinct from the sand-based interfaces that dominate much of the existing literature.

The experiments were built around large-scale cyclic direct shear tests on a rigid steel plate representing the pile surface and remolded silty clay sampled from northern China. Two boundary conditions framed the mechanical problem: constant normal load and constant normal stiffness. Those two settings are not interchangeable. Under a fixed load, the interface can dilate or contract against a steady vertical force; under fixed stiffness, the vertical response changes with deformation, which better resembles soil-structure interaction in real foundations where confinement evolves as the structure moves.

The interface behavior followed the Mohr–Coulomb failure criterion, but the cyclic response was far from simple. Hysteresis loops, volumetric deformation, stiffness loss, and energy dissipation all changed as the interface was driven through repeated cycles. The result is a mechanical picture in which damage accumulates early, then slows and eventually approaches a more stable state. The stiffness degradation was especially concentrated in the early stage of cycling and began to stabilize after about 25 cycles.

Moisture and stress exerted the strongest control over peak resistance

Soil moisture content had a striking effect on strength. Raising moisture from 9% to 15% reduced the initial peak shear strength by 48.9%. In practical geotechnical terms, water softened the silty clay, reduced interparticle friction, and weakened the interface’s ability to mobilize resistance against the steel surface. That finding is consistent with the broader behavior of water-sensitive geomaterials: as pore water increases, the soil fabric loses resistance, and cyclic loading accelerates rearrangement and softening at the contact plane.

Normal stress pushed in the opposite direction. Increasing normal stress from 100 to 300 kPa increased the initial peak shear strength by 164.2%. Greater confinement improves particle interlocking and frictional resistance, especially when the soil is dense enough to transmit load efficiently across the interface. The study therefore shows a sharply dual control: water weakens the interface, while confinement strengthens it, and cyclic loading exposes how those two forces compete over time.

One sentence captures the mechanical consequence plainly: the interface did not merely deform, it progressively lost capacity in a load-dependent way.

Roughness improved initial resistance, but peak strength did not rise in a straight line

Interface texture width changed the early response, but not in a monotonic manner. Increasing roughness enhanced initial shear resistance compared with the smooth interface, yet peak shear strength did not continue to rise steadily as texture width increased. That pattern suggests that mechanical interlock helps at first, but excessive texture does not necessarily translate into proportionally greater mobilization of shear resistance. Once surface geometry becomes too pronounced, local stress concentrations, soil detachment, or premature sliding along preferential planes may limit further gains.

This is one of the study’s most useful engineering messages. Surface roughness matters, but more roughness is not automatically better. For pile design, that means interface machining or surface texturing should be treated as a calibrated variable, not a default assumption that coarser always equals stronger.

Damping and energy loss followed a three-stage evolution during cycling

The cyclic response of the interface stiffness and equivalent damping ratio followed a three-stage pattern: rapid degradation, gradual transition, and stable convergence. In physical terms, the first cycles reorganized the contact fabric, the middle phase adjusted the damaged interface toward a new equilibrium, and the later cycles settled into a more repeatable response. Energy dissipation mirrored these changes. The cumulative dissipated energy depended on both boundary conditions and loading parameters, with rougher interfaces and higher normal stress improving dissipation capacity.

CNS conditions reduced cumulative dissipated energy by about 19.7% compared with CNL conditions. That difference reflects how boundary restraint changes the work done by the interface during repeated sliding and dilation. Under a stiffness-controlled boundary, part of the deformation energy is absorbed in the evolving boundary response rather than being fully expended as interface frictional loss. In the setting of seismic or traffic-related cyclic loading, such details are not abstract. They determine how quickly a pile-soil system softens and how much of the imposed movement is converted into irreversible damage.

Damage evolution was translated into a constitutive model for simulation

To move beyond descriptive testing, the investigators built an improved Mohr–Coulomb constitutive model that introduced a damage variable to capture degradation of the interface mechanical parameters. The model was implemented in ABAQUS and checked against cyclic shear tests at different moisture contents. The simulations reproduced the main trends in shear stress-displacement response and normal deformation evolution. For the normal displacement-cycle relationships, the reported R² values ranged from 0.836 to 0.852.

Those values are not perfect, and they do not claim to eliminate all mismatch between test and model. They do show that the main trends of moisture-driven weakening and cyclic deformation can be represented in a calibrated numerical framework. For foundation analysis, that is the real payoff: a damage-aware constitutive description that can be used to estimate how pile interfaces evolve under repeated loading instead of assuming a fixed frictional parameter throughout service life.

What this means for pile foundations in silty clay

The study connects a set of small-scale mechanics to a larger structural problem. Pile foundations in silty clay regions face repeated actions from earthquakes, traffic, and long-term service loads. If the interface loses stiffness early, the foundation may redistribute stress, accumulate settlement, and undergo changing lateral or vertical response even before overt failure appears. The observed three-stage pattern of stiffness degradation and the moisture sensitivity of peak strength both point to the same design implication: water condition and cyclic loading history should be treated as central variables in pile foundation assessment.

The work also narrows a gap in the existing literature. Much of the prior experimental attention has gone to sand-concrete or red clay-steel systems, while the combined influence of roughness and water sensitivity in silty clay has been less systematically examined. By linking large-scale testing to a cyclic damage model, this study provides a mechanically coherent view of how the pile-soil interface evolves under repeated shear.

The remaining challenge is not whether the interface can weaken under cycling. It can, and the data show it clearly. The harder question for engineering practice is how to translate those damage curves into design limits for real piles subjected to variable field loading, variable groundwater conditions, and long service durations.


Reporting by Dr. A.R. Athira, MBBS for Medicine World Foundation, with additional data provided by PLOS ONE (Medicine).
PRIMARY CLINICAL EVIDENCE & REGISTRY DATA