Prediction of Box-Jacking Force Using a Probabilistic Observational Approach

DOI IR (HANDLE) Open Access

Description

Box-jacking is an increasingly popular means for installing underground utilities and infrastructure. Accurately estimating the expected jacking forces in box-jacking is a key design concern, which can ensure the available thrust is not exceeded, to prevent damage to the box-culverts and/or launch shaft, and the construction efficacy of the jacking project. However, prediction of the total jacking force is complicated due to a multitude of influencing factors. The development of jacking force can be influenced by the site geology, the lubricant performance, work stoppages, shape of box culvert, and tunnel boring machine driving style. In this paper, a probabilistic observational approach is introduced aimed at prediction of jacking forces during the box-jacking process. Markov Chain Monte Carlo (MCMC) was adopted for this purpose which allows forecasts to be performed within a probabilistic framework. The proposed framework was applied to a box-jacking case histories completed in Kanagawa: a 150-m drive in fine and medium sands. The forecasts were appraised through comparisons to predictions determined using a classical optimization technique, namely genetic algorithms. The results show that the proposed framework yields highly accurate predictions for the monitored field data, and the prediction accuracy improves obviously as more data are acquired from the drive.

Journal

Keywords

Details 詳細情報について

Report a problem

Back to top