Search

Meso-scale simulations of pressed energetic materials are performed using synthetic microstructures generated using deep feature representation, a deep convolutional neural network-based approach. Synthetic microstructures are shown to mimic real microstructures in the statistical representation of global and local features of micro-morphology for three different classes of pressed HMX with distinctive micro-structural characteristics. Direct numerical simulations of shock-loaded synthetic microstructures are performed to calculate the meso-scale reaction rates. For all three classes, the synthetic microstructures capture the effect of morphological uncertainties of real microstructures on the response to shock loading. The calculated reaction rates for different classes also compare well with those of the corresponding real microstructures. Thus, the article demonstrates that machine-generated ensembles of synthetic microstructures can be employed to derive structure–property–performance linkages of a wide class of real pressed energetic materials. The ability to manipulate the synthetic microstructures using deep learning-based approaches then provides an opportunity for material designers to develop and manufacture pressed energetic materials that can yield targeted performance. [ABSTRACT FROM AUTHOR]

The effectiveness of several metamodeling techniques, viz. the Polynomial Stochastic Collocation method, Adaptive Stochastic Collocation method, a Radial Basis Function Neural Network, a Kriging Method and a Dynamic Kriging Method is evaluated. This is done with the express purpose of using metamodels to bridge scales between micro- and macro-scale models in a multi-scale multimaterial simulation. The rate of convergence of the error when used to reconstruct hypersurfaces of known functions is studied. For sufficiently large number of training points, Stochastic Collocation methods generally converge faster than the other metamodeling techniques, while the DKG method converges faster when the number of input points is less than 100 in a two-dimensional parameter space. Because the input points correspond to computationally expensive micro/meso-scale computations, the DKG is favored for bridging scales in a multi-scale solver. [ABSTRACT FROM AUTHOR]