City Research Online

Genetic algorithm optimisation of load cell geometry by finite element analysis

Robinson, Gordon M. (1995). Genetic algorithm optimisation of load cell geometry by finite element analysis. (Unpublished Doctoral thesis, City University London)


The objective of the work described in this thesis was to numerically model the influence of end-loading conditions on column strain gauge load cells and to develop the model into a program for optimising the geometry of column load cells.

It is shown that, for most practical loading conditions, load celis with large numbers of equispaced strain gauges are only sensitive to the axisymmetric components of contact stress distribution. The problem of cylindrical load celis subject to frictionless equipollant annular loading is analysed by the method of superposition of the homogenous solutions. A plot of the relationship between sensitivity, load radius and cylinder aspect ratio is presented.

An experimental study using a strain gauged cylinder was then undertaken. The measured strains confirmed the results of the analytic solution and provided results for other loading conditions, more representative of those to which real load celis are subjected.

In order to extend the study to more typical contact conditions and to load celis of more complex shape, a simple finite element (FE) program incorporating contact analysis, automatic meshing, and infinite elements was developed. The results from the program are shown to be in good agreement with those from the analytic solution, from benchmark problems, and from the experiments. The FE program was used to predict the end-loading sensitivity of hollow and solid cylindrical load cells. Plots of their end-loading sensitivity against aspect ratio are presented.

The finite element routines were then used as the core of a Program for Optimising the Geometry of Load cell Elements (POGLE). This program combines genetic algorithm optimisation with finite element analysis to optimise a load celi's shape so as to minimise its end-loading sensitivity.

In order to test the capabilities of the POGLE program it was used to optimise the shape of a 3 MN low-profile load cell. A prototype with the predicted optimum shape was manufactured and tested. The tests confirmed the program's predictions that the new design would have much lower end-loading sensitivity than existing designs.

Publication Type: Thesis (Doctoral)
Subjects: T Technology > TK Electrical engineering. Electronics Nuclear engineering
Departments: School of Science & Technology > Engineering
Text - Accepted Version
Download (15MB) | Preview



Downloads per month over past year

View more statistics

Actions (login required)

Admin Login Admin Login