Abstract
The aim of this thesis is to use numerical modelling to evaluate practical guidelines for soil and large diameter rigid pipe interaction problems. A detailed critique is also given on current design methods, followed by an explanation of the alternative general criteria for field performance. Conventional design methods are generally based on Marston-Spangler theory developed for small diameter concrete non-prestressed pipe in the 1920's, which take no account of the actual material properties, true geometry and field construction procedure. Numerical analyses have been performed using the geotechnical finite element package CRISP, together with data from full scale site installation trials and three-edge bearing load tests. Practical guidance is given for various construction conditions. A brief summary of the main factors influencing the behaviour of buried rigid pipe are: the elastic properties of both the pipe and the surrounding soil, the placement, compaction, and any subsequent volumetric changes in the soils placed around the pipe, the variation of earth cover, geometry and bedding conditions, the weight of water, internal pressures, and prestressing effects from helically-wound steel wires. From this study, it was found that the prestressing and internal water pressures significantly affect the behaviour of buried pipe, leading to the change of bending stresses in the pipe wall. The improvement of construction techniques is also very important in terms of bedding condition and compaction of side fill, etc. In conclusion, current design practice is highly deficient, causing very uneconomical designs for buried large diameter rigid pipe. The finite element method has provided useful insights into an important practical engineering problem, and has indicated the critical design factor for large diameter buried rigid pipe.