PREDICTION OF QU AND Ru CAPACITIES OF PILE IN CLAYEY SOIL LAYER USING GEOSPATIAL ANALYSIS

Abstract

The basic purpose of foundation design is to evaluate ultimate soil bearing (Q_u) and load-carrying capacities (R_u) at the final set installation. A case study when conducting the deep foundation design of building project with 7 floors at Jakarta Pusat area, the project owner was only allowed to use CPT data. The location was not only relatively densely populated but also the soil investigation work could cause some noises. CPT has widely used in soil investigation and provides reliable subsurface soil. However, the most suitable design methods must be selected and easy in predicting the Q_u value of a single pile. Schmertmann's method in determining Q_u is a selected method for static loading and relatively suitable for the clay layer in Indonesia. Two random points at one of the configurations of single piles were evaluated by PDA test in determining the real load capacity (R_u). Values of (Q_u) and (R_u) were analyzed by spatial autocorrelation through Moran’s I as well as the prediction and interpolation by semivariogram and ordinary Kriging on (Q_u) and (R_u) data. Then, it compares other prediction and interpolation models (gaussian, exponential, and spherical). The Gaussian model is a model that gives better predictions than other models. Application of the ordinary Kriging method for interpolation of the spatial distribution of (Q_u) and (R_u) yields good results for other location points of single piles with an error of around 13.2% at the final set installation.