Abstract
Engineers piece together information from multiple sources to make preliminary interpretations of ground-related uncertainty. Data is sourced from previous projects or case histories and used to inform engineering design, determine the allocation of financial resources and critically, provide confirmation that allowable ground conditions are present. In engineering projects, key decisions rely on the transfer and interpretation of this data, especially where the data of the subsurface can inform the new work. For engineers working in data-scarce regions the absence of viable geotechnical data can prove challenging.The first known campaign of Cone Penetration Testing (CPT) has been undertaken to determine and evaluate parameters relating to seismic response of the soils beneath the Kathmandu Valley, Nepal. The results corroborate with the findings of previous description of soft, low strength soils at shallow depths, previously only characterised in situ by Standard Penetration Tests. Two new rotary-open holed boreholes drilled in the two main geological sequences underlying the valley aid practical understanding of the CPT results. Parameters cone resistance, sleeve friction and Soil Behaviour Type are investigated for correlation with shear-wave velocity, whilst varying the correction for overburden stress. The results present a unique relationship for predicting this parameter but show that the effectiveness of these relationships when in a data scarce region in practice is limited. In response to this inefficacy, a framework considering; the acquired new shear-wave data and its potential for sampling bias, and implementation in geostatistical methods, is used to produce a new map of the time averaged shear-wave velocity in the first 30 m of soil (VS30) for the valley.
The necessary developments needed to better predict VS30 on a regional scale are implemented using a multi-Gaussian Bayesian geostatistical approach. This selected Bayesian updating methodology is typically used in the context of petroleum reservoir modelling. The impact of this framework on global VS30 prediction, is expected to be in simplifying the typically complex decisions made during geotechnical evaluation of a region, and provides the ability to apply a targeted response, i.e., the location of future ground investigation, to understanding better the geotechnical uncertainties typically impacting earthquake hazard assessment.
| Date of Award | 25 Jan 2022 |
|---|---|
| Original language | English |
| Awarding Institution |
|
| Supervisor | Paul J Vardanega (Supervisor) & Elizabeth A Holcombe (Supervisor) |