Multi-level coupling of dynamic data-driven experimentation with material identification

We describe a dynamic data-driven methodology that is capable of simultaneously determining both the parameters of a constitutive model associated with the response of a composite material, and the optimum experimental design that leads to the corresponding material characterization. The optimum design of experiments may contain two parts. One involving the identification of the parameters that are tunable prior to performing the experiment such as specimen characteristics and a priori loading path. The other is involving the parameters characterizing the experiment during the experiment itself such as the directionality of the loading path for the case of multi-axial loading machine. A multi-level coupled design optimization methodology is developed and applied to demonstrate the concept. Essential to the process is the development of objective functions that express the quality of the experimental procedure in terms of the uniqueness and distinguishability associated with the inverse solution of the constitutive model determination. The examples provided are based on the determination of the linear constitutive response of a laminate composite material.