EXPLORING TRANSIENT, NEUTRONIC, REDUCED-ORDER MODELS USING DMD/POD-GALERKIN AND DATA-DRIVEN DMD

There is growing interest in the development of transient, multiphysics models for nuclear reactors and analysis uncertainties those models. Reduced-order (ROMs) provide a computationally cheaper alternative to compute uncertainties. However, application ROMs transient systems remains challenging task. Here, 1-D, twogroup, time-dependent, diffusion model was used explore potential three different ROMs: intrusive POD-Galerkin DMD-Galerkin methods purely datadriven DMD. For problem studied, exhibited by far best accuracy selected further uncertainty propagation. Perturbations were introduced initial condition cross-section data. A greedy-POD sampling procedure construct reduced space that captured much variation uncertain these parameters. Results indicate relatively few samples parameters are needed produce basis leads distributions quantities match well with obtained from full-order using brute-force, forward sampling.