| |
|
Session 005 Model Identification/Inverse Problem (Monday Morning)
|
| 10:30 AM |
478 |
Keynote Presentation (40 minutes)
Identification From Full-field Measurements: A Promising Perspective in Experimental Mechanics
|
| |
M. Grédiac, Université Blaise Pascal
|
| |
Full-field measurement techniques such as digital image correlation, moiré interferometry, speckle or grid method are now widely used in the experimental mechanics community. Such techniques provide displacement, strain or temperature maps, which are very useful to analyze phenomena that occur during mechanical tests, such as parasitic effects due to boundary conditions or strain concentrations caused by some local phenomena. This wealth of data also provides resources to validate numerical models or to propose relevant constitutive equations.
Another important and very promising issue is to identify parameters governing constitutive equations by processing these fields with suitable numerical tools. The advantage is to account for a large amount of data from one test only and to process heterogeneous strain fields in which several parameters are simultaneously involved. Solving this problem is however somewhat complicated since no direct link generally exists between measurements and unknown parameters. Suitable and robust identification strategies must therefore be developed and validated to face this challenge, which is recent in the experimental mechanics community.
The keynote lecture will first shortly describe various applications of full-field measurement techniques in experimental solid mechanics. It will then focus on identification from full-field measurements, with a special emphasis on specific difficulties raised by this problem, on recent breakthroughs which offer solutions for solving it and on related questions which still remain open.
|
|
|