Modelling horizontally loaded piles in the geotechnical centrifuge

Abstract

Pile foundations are extensively used to support various structures that are constructed in soft/loose soils, where shallow foundations would be considered ineffective due to low bearing capacities and large settlements. The design of these structures to accommodate lateral applied loads in particular, usually imposed by winds, water and earth pressures, has gained popularity over the past few decades. The behaviour of horizontally loaded piled foundations is a complex soil-structure interaction problem and is usually concerned with the relative stiffness between the pile and the surrounding soil, where the relative stiffness is a function of both the stiffness and properties of the pile and the stiffness of the soil.

Many design assumptions and methods used for pile foundations are based on the principles observed from metal piles. This raises the question of the validity and accuracy of assumptions and methods for the use of analysing and designing reinforced concrete piles, that exhibits highly non-linear material behaviour and changing pile properties after cracking. Due to the elastic behaviour of metal sections, these methods typically only focus on the soil component of the soil-structure interaction problem, only allowing changes and non-linear behaviour of the soil surrounding the pile to take place upon load application, mostly disregarding the behaviour and response of the pile itself.

The main purpose and objective of the study was to determine whether aluminium sections in a centrifuge could be used to realistically and sufficiently accurately model the monotonic and cyclic response of reinforced concrete piles subjected to lateral loading. This was observed though a number of tests conducted in a geotechnical centrifuge on scaled aluminium and reinforced concrete piles, subjected to both monotonic and cyclic loading.

After conducting the tests on both the scaled aluminium and reinforced concrete piles in the centrifuge it was concluded that aluminium sections cannot be used to accurately model and predict the lateral behaviour of reinforced concrete piles. Both the scaled aluminium and reinforced concrete piles proved to model the concept of laterally loaded piles quite well regarding bending at low loads. However, even at low lateral loads, the observed response of the scaled reinforced concrete was significantly different than that observed from the scaled aluminium pile. Furthermore, as the magnitude of the applied load and bending increased, the scaled reinforced concrete pile cracked, resulting in non-linear behaviour of the section under loading, which was not the case for the scaled aluminium pile that remained uncracked. This contributed to the difference in behaviour between the piles studied, therefore, the true material behaviour and failure mechanisms involved with reinforced concrete piles were not replicated by using a scaled aluminium pile section. The non-linear behaviour of the scaled reinforced concrete pile after cracking affected both the behaviour of the pile, as well as the response of the soil surrounding the pile, in contrast with the behaviour observed from the scaled aluminium pile.