Geometric Optimization of a Functional Meso-structure with DAKOTA for Surrogate Energetic Heterogeneous Mixture*

Optimization of mesoscale (~1 mm) geometric structures is performed with a genetic algorithm framework utilizing the computational optimization tool, DAKOTA. The goal of this research is to design meso-structures into surrogate energetic systems that can enhance a system’s dynamic response by energy concentration or dissipation into localized regions. A conical shape is prescribed by DAKOTA’s Multi-Objective Genetic Algorithm (MOGA) to attain desired velocities in specified Lagrangian tracer locations of a heterogeneous mixture of a high and a low-impedance material. Objective functions are defined to generate maximum and minimum particle velocity profiles at specified locations where the input parameters are the individual blocks of material with a specified mesh that can move along the Lagrangian domain. Samples are fabricated with a conical structure of sugar embedded into the Polydimethylsiloxane (PDMS) polymer. Uniaxial strain experiments are performed with a single-stage gas gun between 300m/s and 600 m/s projectile velocity to confirm the functionality of the proposed optimized meso-structure. PDV probes are placed to monitor the particle velocity response on the specified locations obtained from the optimization results. Simulations are conducted in CTH using the 3D geometry of the sample acquired from the XCT scan to access the dynamic behavior of the shock wave under extreme loading conditions. Experimental results indicate that higher and lower velocity responses may be achieved due to the impedance mismatch of building materials directing the shock wave into the predefined directions.