The application of parallel computer technology to the dynamic analysis of suspension bridges

Beith, Jason Gordon (1997) The application of parallel computer technology to the dynamic analysis of suspension bridges. PhD thesis, University of Glasgow.

Abstract

This research is concerned with the application of distributed computer technology to the solution of non-linear structural dynamic problems, in particular the onset of aerodynamic instabilities in long span suspension bridge structures, such as flutter which is a catastrophic aeroelastic phenomena.

The thesis is set out in two distinct parts:-

Part I, presents the theoretical background of the main forms of aerodynamic instabilities, presenting in detail the main solution techniques used to solve the flutter problem. The previously written analysis package ANSUSP is presented which has been specifically developed to predict numerically the onset of flutter instability. The various solution techniques which were employed to predict the onset of flutter for the Severn Bridge are discussed. All the results presented in Part I were obtained using a 486DX2 66MHz serial personal computer.

Part II, examines the main solution techniques in detail and goes on to apply them to a large distributed supercomputer, which allows the solution of the problem to be achieved considerably faster than is possible using the serial computer system. The solutions presented in Part II are represented as Performance Indices (PI) which quote the ratio of time to performing a specific calculation using a serial algorithm compared to a parallel algorithm running on the same computer system.

• Item Type: Thesis (PhD)
• Qualification Level: Doctoral
• Subjects: T Technology > TG Bridge engineering; T Technology > TA Engineering (General). Civil engineering (General)
• Colleges/Schools: College of Science and Engineering > School of Engineering
• Supervisor's Name: Agar, Dr. Alan
• Date of Award: 1997
• Depositing User: Mrs Marie Cairney
• Unique ID: glathesis:1997-1108
• Copyright: Copyright of this thesis is held by the author.
• Date Deposited: 02 Sep 2009
• Last Modified: 10 Dec 2012 13:33
• URI: https://theses.gla.ac.uk/id/eprint/1108