Application of Statistical Method on Field Pile Loading Test Result Analysis

Abstract

Most geotechnical engineering issues face great degree of uncertainties since it studies a natural material - soils and rocks. So the analysis of data obtained from field tests requires intensive data processing in order to draw meaningful conclusions which can be applied to actual field practice, while dealing with implied uncertainties within the obtained data.

Among various piling method, the precast bored pile is considered a relatively new piling method. The biggest advantage in using precast bored piles is that it can reduce public nuisance during installation, such as noise, vibration and dust. And for this reason, it is widely recognized as a suitable piling method for urban construction and construction projects within close proximity to other structures. Moreover, due to its unique installation procedure and maximum mobilization of the shaft resistance, the precast bored piles can be applied to wide range of ground condition. However, the behavioral characteristics of precast bored pile contains higher degree of uncertainties along the pile shaft.

In this study, the analysis of over 300 data points obtained from a full-scale pile loading test was carried out. Standard pile loading test was conducted on 20 fully instrumented steel pipe test piles. The test piles were installed following the typical procedure applied on actual steel pipe precast bored piles. From this, a new empirical formula estimating the bearing capacity of the steel pipe precast bored pile is proposed based on a intensive statistical data processing of the field test data. As a conclusion, by applying the statistical data processing method a new empirical formula which was found to be more accurate compared to the existing formula, as well as securing the safety of the foundation design.