skip to primary navigationskip to content
 

Adaptive Design of Supported Excavations - Dr Giovanna Biscontin

This project aims to fully document and make available the automated back analysis tool developed by Ms Jin to support commercial implementation, as well as investigate barriers to the adoption of data-driven adaptive design in practice through a targeted workshop. It will also provide guidance on future research developments needed to promote data-driven design in practice; promote provisions on adaptive design to be included in standards and guidance documents.

The observational method (OM), originally proposed by Peck in 1969, and formalised in Eurocode 7 in 1987, can provide a way to avoid redundancy in excavation design and deliver construction more economically and efficiently through modifications to the original design during construction. Quite often supported excavations are over-designed and it is not unusual for measured deformations to reach values only half of the predicted amounts. This clearly underscores a substantial potential for savings. However, the uncertainties still associated with ground investigations and numerical modelling do not allow for leaner designs at the start of the project without tight controls during construction. The emergence of advanced analysis tools, together with large amount of data now readily available during construction, make possible the development of a robust framework for a real-time, data-driven, decision making process based on the observational method, in which data can be best utilised to deliver real value, confidence, and control.

An automated 'real time' back analysis approach based on Bayesian inference was developed in Yingyan Jin's PhD thesis and validated with Crossrail case histories. This approach significantly improved the efficiency of back analysis compared with the 'trial and error' manual procedures used in current practice and delivers estimates of soil parameters for a given geotechnical model, updates the prediction of future excavation stages, while fully quantifying uncertainties. As a result, we moved a step forward towards adaptive design by using state-of-art machine learning techniques to implement the core tool for integrating construction data and geotechnical analysis. However, in order to be able to support adaptive design in practice during construction, factors such as logistics, construction schedule, and risk need to be integrated into the model to ensure an optimal design in terms of safety, costs, and executability. Technological development alone is not going to be sufficient to promote the adoption of adaptive design in construction of supported excavations. Broader changes to contractual arrangements are also needed to transform current, and well consolidated, practice.

This project aims to:

  1. fully document and make available the automated back analysis tool developed by Ms Jin to support commercial implementation;
  2. investigate barriers to the adoption of data-driven adaptive design in practice through a targeted workshop;
  3. provide guidance on future research developments needed to promote data-driven design in practice;
  4. promote provisions on adaptive design to be included in standards and guidance documents.

Researcher:


Department of Engineering

Filed under:

Welcome to the Centre for Digital Built Britain.  

The Centre for Digital Built Britain is a partnership between the Department of Business, Energy & Industrial Strategy and the University of Cambridge to deliver a smart digital economy for infrastructure and construction for the future and transform the UK construction industry’s approach to the way we plan, build, maintain and use our social and economic infrastructure.