This paper on "Nonlinear Benchmarking Involving Material and Geometric Effects" was presented at the International Conference on Quality Assurance and Standards in Finite Element Analysis - 13-15 May 1987, The Grand Hotel, Brighton.

Summary

Benchmark problems are often proposed as tools for checking the correctness and correct usage of software. In nonlinear analysis the construction of such problems is particularly difficult since the material behaviour itself is a matter of uncertainty and non-uniformity of treatment. There is a choice between artificial problems with very exact target solutions and problems based on tests in which agreement is sought with experimental results.
A benchmark exercise of the latter type is considered here, based on a structural test of a thin cylinder subjected to transient thermal loads. Experience with this problem is discussed, with conclusions concerning the precautions to be taken to ensure significance of comparisons between experimental and computational results.