An advanced identification procedure for material model parameters based on image analysis

Inverse methods for the material strength model calibration are widespread techniques, which allow taking into account for the actual strain, strain-rate, temperature and triaxiality fields inside the specimen. An optimization procedure generally starts from experimental measurement of force-stroke time history and is based on the minimization of the difference between experimental and numerically computed quantities. In this work, the strength model identification is performed also on the basis of the specimen shape recorded during the test. This information is imposed as boundary condition, which forces the experimental profile to the external surface of the specimen. The optimization is based on the minimization of the reaction force of the imposed boundary condition. This technique could be applied both to quasi-static and dynamic tests, also at different temperatures, since the only additional requirement is a video of the test with a good compromise in terms of spatial and time resolutions. The methodology is compared with a standard numerical optimization procedure, in order to evaluate the reliability of the method and the advantages/disadvantages of this new approach.