Simulation versus Test
Linking Computational and Experimental Techniques in Industrial
Applications
 This event, which was held at Ratcliffe Power Station, UK, followed
a successful event on the same subject which was held a year ago.
Once again it proved to be a popular and interesting event,
attracting 73 delegates.
Tests are frequently used by analysts to help to demonstrate that
their simulations agree with reality. During this seminar, a number
of different industrial examples were presented which illustrated
the benefits of using the two techniques in conjunction with one
another, but which also highlighted many of the difficulties
involved in obtaining – and being able to demonstrate –
strong correlation between the results.
One interesting aspect of this meeting was to bring out the
parallel issues associated with testing, including various sources
of errors and the potential for significant variation in the
results from apparently identical tests. It was poignant to note
that test results are really just another approximation to real
life.
Some of the difficulties in getting good correlation between
simulation and test were also highlighted.
These included:
Having accurate material properties
Simulating the real test configuration
Difficulties for both simulation and test caused by the presence of
singularities
Approximations in both methods
Obtaining accurate measurements of the true loads applied
Being able to interpret the test results fully
This event was organised in conjunction with the British Society
for Strain Measurement ( www.bssm.org
) and the Institution of Mechanical Engineers. The organising
committee also included representatives from E-on, the University
of Southampton, Strainsense, Metec and DAMT.
GLASGOW, UK, NOVEMBER 28TH 2005
|
