Abstract:
Cast-in-place bored piles are widely employed all over the world as deep foundations to support heavy structures such as high-rise buildings, bridges, and viaducts. Construction of pile foundations is costly, difficult and time consuming. It is thus important to properly understand the load-settlement behaviour of piles and to develop a rational approach for pile design that takes into account subsoil and geological data of the area and the pile construction method. This is expected to result in greater accuracy in prediction of ultimate bearing capacity and settlement of piles subjected to vertical loads. Ultimate bearing capacity of bored piles are generally evaluated using pile capacity equations or analytical methods prescribed in the literatures or given in various codes. In many cases, ultimate pile capacity is obtained directly using SPT-N value obtained from standard penetration tests. The allowable or design pile capacity is then computed using a high factor of safety on the ultimate capacity assuming that this will satisfy limiting settlement or serviceability criteria of the superstructure. The expected settlement of the pile at ultimate and allowable loads and profile of load distribution of pile with depth is not predicted by analytical methods given in the literature and in the codes. As a result, for economic and rational design of piles it is recommended to carry out expensive and time consuming full-scale static pile load tests on-site to ascertain bearing capacity and settlement of bored piles. The principle aim of the current research work is to investigate whether relatively simple design procedures may be formulated that can be used by both academics and practitioners to evaluate bearing capacity as well as the settlement of single pile from standard geological and geotechnical data that are generally available for foundation analysis and design.
With the above objective in mind, static load test data of 22 fully instrumented piles was conducted for foundation design of bridges and viaducts of Padma Bridge Rail Link Project (PBRLP). The results of these tests are used in this study to investigate load transfer and load settlement behaviour of single piles in greater detail. Finite element modeling of piles is conducted using PLAXIS-3D which incorporates advanced constitutive models like Hardening Soil model or HS model. Empirical correlations for shaft friction and end bearing resistance of piles are formulated as a function of SPT-N value using data obtained from instrumented test piles. These functions are subsequently used in load transfer method developed by Coyle & Reese (1966). An excel spreadsheet is developed incorporating Visual Basic for Applications or VBA code. The code is developed to carry out the iterative computations for determining load settlement behaviour and axial load distribution profile of the pile using the load transfer method. Load-settlement and load-distribution data generated from static load test of fully instrumented piles, are compared with simulations done by finite element analysis using PLAXIS-3D and by the semi-analytical load transfer method.
It was found that the HS model realistically predicts stress-strain behaviour of soils. From a comparison of instrumented test pile data with simulations, it was observed that predictions obtained using finite element analysis by PLAXIS-3D and by using semi-analytical load transfer method are in good agreement with the field data. Finally, ultimate bearing capacity of instrumented test piles computed using the field data, finite element model, load transfer method and pile capacity equations given in the literature and in various codes were compared to get a more complete picture and detailed understanding of the short-term capacity of single piles.