Geotechnical Exploratory Data Analysis For Fault Rock Mass Classification

In Namibia,Windhoek city,considerable investment is needed for geotechnical data acquisition and rock mass exploration for safe foundation design.Windhoek city is experiencing increased civil development.However,geotechnical investigations around the Windhoek central business district are faced by challenges such disparate data sources,costly geotechnical survey,and inefficient data analysis techniques.This thesis focuses on utilizing systematic exploratory data analysis for the rock mass parameters of the Pahl Fault Zone passing through Windhoek city,central business district.Importantly,there is insufficient research on the Pahl Fault Zone rock mass parameters such as RQD,fracture frequency and fault gouge depth.The thesis aims to improve research on the Pahl fault by employing geomorphologic analysis,scanline mapping,descriptive statistics,exploratory data analysis and regression modelling for rock mass classification and characterization.The main statistical tools used were histograms,scatterplots,stereonets,statistical programming,margin plots and maps.Additionally,regression modelling was used to establish a linear multivariate linear relationship between RQD,depth and fracture frequency.Scanline mapping survey yielded a total length of 120.5 m from 10 scanline points.Likewise,the average joint spacing of the scanline points is 0.664 m with a standard deviation of0.3828.Joint orientation data indicated steep dips(NW and NEE)with angles between 40-90o.Scanline RQD* and joint spacing histograms showed skewed negative distribution.The skewed distributions indicate lack of normality inherent with geological data.Moreover,Schmidt hammer tests were tested for the rock mass uniaxial compressive strength which ranged from 18-35 Mpa,indicating a weak rock.However,comparing scanline and borehole compressive strength data,the analysis showed the rock mass becomes stronger with increasing depth.Multivariate regression modelling was analyzed to establish a relationship between RQD,fracture frequency and depth.The result yielded a power relationship between RQD and borehole depth with a brink point at 10 meters whereby a steep increase in RQD is associated with depths greater than 10 meters.The power relation is caused by high fracturing,anisotropy,and weathering at shallow depths(below 10 meters)and low fracture zone above 10 meters.From statistical correlation coefficient and p-values analysis shows there is a high correlation between RQD,depth,fracture frequency.Mathematical formulations were proposed for RQD,depth and fracture frequency.Finally,spatial maps were produced using ArcGIS to classify thesurficial rock mass based on the spatial influence of parameters(RQD,stability,weathering etc.).The geotechnical maps can help geotechnical engineers in planning cost-effective geotechnical investigations and saving excavation input in poor rock zones.