Loughborough University
Thesis-1995-Dakin.pdf (13.16 MB)

Using experimental loads with finite element analysis for durability predictions

Download (13.16 MB)
posted on 2012-09-21, 13:50 authored by John D. Dakin
This research work involved the prediction of the fatigue life of an automotive rear suspension twistbeam assembly fitted to a vehicle travelling over a customer correlated durability route. This was achieved by making use of the integrated concepts of scaling and superposition of linear static finite element analysis being driven by experimental load data - the so called 'quasi-static time domain' approach. A study of the free body diagram of the twistbeam resulted in an indeterminate load set of some 24 components, with experimental data indicating that a state of static unbalance existed. Subsequent to developing a matrix-based generalised method ofload cell calibration to confum the foregoing, a modal technique was developed to partition the experimental data into a static load set, causing elastic deformations, and a rigid load set, imparting rigid body accelerations. The semi-independent characteristics of the twistbeam necessitated the coupling of large structural displacements with inertia relief. This required extensive modifications to the current techniques and led to the development and use of a three dimensional functional response matrix in place of the conventional two dimensional one. Recommendations concerning appropriate finite element boundary conditions were also formulated to handle these effects. Finally, the limitations of the uniaxial fatigue model were revealed under the application of a set of tools for analysing the biaxiality and mobility of the maximum absolute principal stress.



  • Aeronautical, Automotive, Chemical and Materials Engineering


  • Aeronautical and Automotive Engineering


© John D. Dakin

Publication date



A Doctoral Thesis. Submitted in partial fulfilment of the requirements for the award of Doctor of Philosophy of Loughborough University.

EThOS Persistent ID



  • en