This research project was concerned with the investigation of material parameters associated with the ‘residual soil’ profile of South East Queensland (SEQ) and their variation across the full weathering profile (residual soil and highly weathered rock materials). Based on an assessment of their prevalence throughout the SEQ region, the specific ‘residual soils’ investigated by this study were subsurface materials that exhibited properties that were primarily determined by the soil skeleton and remnant rock structure, contained a significant coarse sized component, and did not demonstrate behaviours influenced by their fines (clay mineralogy) component. A review of historical records held by a major Queensland government department revealed that such materials were becoming increasingly underrepresented in the frequency of material testing, regardless of the importance their insitu deformation and strength properties could have on geotechnical design projects.
The research initially identified a lack of a simple field tests able to determine the insitu deformation parameter of a ‘residual soil’ as a void in the current geotechnical knowledgebase. To overcome this shortcoming, the initial section of this thesis details fieldwork in which the use of a Light Falling Weight Deflectometer (LFWD) for the direct measurement of insitu modulus parameters was developed. Comparative testing was undertaken with the Plate Load Test at three (3) discrete SEQ sites, and with the DCP at an additional seven (7) SEQ sites. The results of this testing demonstrated the suitability of the LFWD for use an investigative tool for residual soil and highly weathered rock materials. The results of the LFWD characterisation of the near-surface produced an insitu modulus parameter of a higher sensitivity and with lower inherent variation when compared to the results of other conventional site investigation techniques.
The focus of the research was transferred deeper into the typical SEQ weathering profile, and studied material parameter variation across the gradational change between ‘soil’ and ‘rock’ materials. Although the distinction is frequently subjective, and often recorded at an arbitrarily depth during site investigations, the field designation of a material into either of these two (2) categories considerably affects the design material parameters assigned. By analysis of the penetrative ability of various types of drilling bit used for borehole drilling in site investigation, this research demonstrated that drill bit ‘refusal’ does not typically occur at the soil / rock interface. Instead, ‘refusal’ generally occurs within highly to moderately weathered rock. The completed analysis has demonstrated that the depth of ‘refusal’ is actually a function of (a) the type of drill bit; (b) rock strength; and (c) defect spacing within the rockmass. The project presented the likely (statistical) material parameters associated with the ‘refusal’ depth of various drill bit types, and that drill bit ‘refusal’ could realistically occur at any point within a rockmass up to a moderately weathered (MW), medium strength rock material. This finding was significantly different to the conventional assumption that such drill bit ‘refusal’ would typically occur at the soil / extremely weathered rock interface.
Similarly, this research showed that no change in material parameters occurred at the logged soil / rock interface in SEQ weathering profiles. By analysis of the results of alternative site investigation techniques (e.g. shear wave velocity) of SEQ sites, typical material parameters associated with residual soils and both extremely weathered and highly weathered rock types were quantified. Instead of an near-instantaneous increase in strength and deformation properties occurring at the soil / rock interface – as frequently indicated within borehole records at the point drilling technique is changed – all observations indicated a gradational increase in strength and deformation parameters as the effects and extent of weathering processes were reduced.
The completed work also allowed a reassessment of SPT to strength relationship to be completed for ‘weak’ rock materials. The finding of this study showed that SPT’s completed by the adoption of the Australian Standard test method and interpretation would result in an overestimation of rock strength of Highly Weathered (HW) rock. This highlighted the limitations of the existing Australian Standard methodology, and this project recommended an appropriate SPT ‘N’ value for HW rock could be found by simply multiplying the ‘N’ value applicable to XW rock by 160%.
The third distinct section of the research project was concerned with the properties of ‘weak rock’ associated with weathering profiles within SEQ. These materials are known to be significantly affected by the insitu weathering processes applied, and have been assessed by this project in terms of the strength variation by weathering category. Extensive strength index testing was completed on 18 SEQ rock units, and characteristic rock strengths based on material origin and weathering classification was produced. The presence of strength anisotropy behaviour for each material unit and weathering category was also assessed. The results of this study demonstrated that once a rock material was weathered into a ‘moderately weathered’ classification, an observable change in material strength and strength anisotropy parameters could be expected. The magnitude of the loss of strength and presence of anisotropy was found to be specific to the parent rock unit’s material origin.
The method used to determine a typical ‘characteristic’ strength for use in subsequent foundation design – in this study rock socket pile design – was completed, based on the results of O-Cell instrumented large-scale test piles installed within a SEQ weathered rock profile. Rock strength and insitu deformation parameters required to be input into various rock socket design methodologies in order to replicate the field observations were found to be specific and varied from values below the 5th percentile to significantly above the median (50th percentile) value determined by extensive rock strength testing.
An important finding repeated throughout all studies related to the SEQ residual soil and weathered rock profile was the statistical distribution of the compiled datasets. For most analysis, a high coefficient of variation (CoV) value was returned, demonstrating a high degree of variability in results (even after the data had been controlled for material type and weathering classification) and indicating the heterogenetic nature of SEQ materials. It was repeatedly determined that a non-normal distribution was applicable to the compiled material parameter datasets, with the datasets largely exhibiting ‘right-skewed’ behaviour. This finding is extremely important in the context of determining ‘characteristic’ design parameters, as the assumption of normality would likely provide inappropriate values that could either significantly over or under-estimate the insitu condition of the residual material.
In its entirely this research project evaluated variation in material behaviour across the full residual / weathering profile – residual soil and weak rock materials – with each chapter either providing specific guidance on the suitability / limitations of site investigation techniques within such profiles or assessing a specific parameter exhibited by such materials. Typical material parameters, and variation therein over the full weathering profile, are presented throughout this thesis, which are suitable for use to characterise the SEQ residual soil / weak rock profile. These research findings were progressively implemented upon ‘live’ projects over the period of this thesis.