9 On the Boundary Conditions and Optimization Methods in Integrated Digital Image Correlation 61 a decent solution, though using the MATLAB Optimization Toolbox should be considered because of code reliability and future-proofing, since toolboxes are commercially maintained by Mathworks. Furthermore, implementation of the Toolbox functions enables for using other optimization methods in the future, as new methods are introduced by Mathworks. Another recommendation is extend the use of an update limit in the custom Gauss-Newton implementation, such that constraints based on physical boundaries are enforced, reducing the solution domain, thereby enhancing the robustness. In practical use of the IDIC program, when choosing the initial guess is difficult, it can be useful to use a very robust, but slow, derivative-free method (e.g. Nelder-Mead) to determine the search direction and increasing the initial guess accuracy. After some iterations, getting nearer the solution, one can switch to a faster gradient-based method, e.g. Gauss-Newton. Using the implementation of MATLAB Optimization Toolbox, switching optimization methods runtime is possible. The analysis is done using a varying Young’s Modulus and Poisson ratio, but IDIC can be used to determine many other constitutive parameters. The different optimization methods may differ in performance and robustness when other mechanical properties are to be identified. References 1. Neggers, J., Hoefnagels, J.P.M., Geers, M.G.D., Hild, F., Roux, S.: Time-resolved integrated digital image correlation. Int. J. Numer. Methods Eng. 103(3), 157–182 (2015). doi:10.1002/nme.4882 2. Blaysat, B., Hoefnagels, J.P.M., Alfano, M., Lubineau, G., Geers, M.G.D.: Interface debonding characterization by image correlation integrated with double cantilever beam kinematics. Int. J. Solids Struct. 55, 79–91 (2015). doi:10.1016/j.ijsolstr.2014.06.012 3. Ruybalid, A.P., Hoefnagels, J.P.M., van der Sluis, O., Geers, M.G.D.: Comparison of the identification performance of conventional FEMUpdating and Integrated DIC. Int. J. Numer. Methods Eng. 106(4), 298–320 (2016). doi:10.1002/nme.5127 4. Ruybalid, A.P., Hoefnagels, J.P.M., van derSluis, O., Geers, M.G.D.: Image-based interface characterization with a restricted microscopic field of view. (2017). Submitted for publication 5. Björck, A.: Numerical Methods for Least Squares Problems. Society for Industrial and Applied Mathematics (1996). Retrieved from http:// epubs.siam.org/doi/book/10.1137/1.9781611971484. doi:10.1137/1.9781611971484 Sandra Kleinendorst PhD-researcher in the Mechanics of Materials group at Eindhoven University of Technology, investigating the optimization of stretchable electronics by developing algorithms that couple micro-mechanical experiments to finite element simulations. Bart Verhaegh Master student at the mechanical engineering department of the Eindhoven University of Technology. Johan Hoefnagels Associate professor at the Eindhoven University of Technology in the field of experimental micro-mechanics with a focus on thin films and interfaces. The research includes size effects, interface and damage mechanics, flexible/stretchable electronics, and MEMS. Andre Ruybalid PhD-researcher in the Mechanics of Materials group at the Eindhoven University of Technology, investigating the use of integrated digital image correlation for characterization of interface behavior in micro-electronic structures. Olaf van der Sluis Associate professor at the Eindhoven University of Technology focusing on numerical modelling of the mechanics of microelectronics, stretchable electronics, and interface mechanics. MarcGeers Full professor of the group Mechanics of Materials at the Eindhoven University of Technology, with an interest in damage mechanics, micromechanics, multi-scale modelling, generalized continua, crystal plasticity and metal forming.
RkJQdWJsaXNoZXIy MTMzNzEzMQ==