Determination of interface shear strength is crucial in the design of many geotechnical structures. To study the effect of interface roughness, particle geometry (size and shape), and sand gradation on the interface friction angle, direct interface shear tests were performed for 10 sands with varying particle sizes, shapes, and gradations and four steel surfaces with different levels of rusting. When sheared along the same interface, the interface friction angle was greater for sands with smaller particle sizes and more angular or elongated particle shapes. For a given sand, the interface friction angle increases with increasing surface roughness. For sands with uniform particle size, a unique relationship was found between the normalized surface roughness and the ratio of the sand–steel critical-state interface friction angle to the internal critical-state friction angle of the sand. Given the same surface roughness and mean particle size, smaller critical-state interface friction angles were mobilized for graded sands than for sands with uniform particle sizes.

This paper presents the results of axial load tests performed on instrumented model piles pre-installed in a large-scale, half-circular chamber with a viewing window in its flat-side wall. Uniform silica sand samples were prepared with different densities using dry pluviation. The effects of pile surface roughness and soil density on the response of the soil during loading of the model piles were studied by analysing sequences of digital images using the digital image correlation technique. Test results show that the extent of the zone next to the pile that is affected by loading of the pile increases as the pile surface roughness and soil density increase. The development of a shear band next to the pile shaft was also studied by carefully analysing images taken with a digital microscope during loading of the model piles. The average thickness of the shear band is in the 3·2D₅₀–4·2D₅₀range for rough model piles, whereas no shear band was observed for smooth model piles. Understanding of shear band formation along the pile–soil interface provides insights into the calculation of the shaft resistance of the pile as a function of initial soil density and stress as well as pile surface roughness.

A full-scale, double-wall, open-ended pipe pile with a diameter, B = 660 mm was densely instrumented with a combination of electrical-resistance and vibrating-wire strain gauges. The test pile was driven into a gravelly sand soil profile to a depth of 30·48 m and then statically and dynamically load tested. The double-wall instrumentation allows independent measurements of the outer shaft, inner shaft (plug) and annulus resistances during a static load test. The residual loads locked in the pile after driving were measured and considered in the correction of the resistance components. The load–settlement curve, load transfer curves and the profile of the unit limit shaft resistance were obtained from the static load test measurements. Two cone penetration test-based methods were used to estimate the resistance of the test pile, which was compared with that obtained from the static load test. Both methods provide good estimates for the unit shaft resistance in sandy layers with low gravel content (30%). The estimated base resistances by both methods match well that measured in the static load test.