M&C 2017
The reduced order model methods are very useful for real-time or many-query simulations of nuclear reactors. The common way for the reduced order models construction is a reduction of the preliminary calculated snapshots by the proper orthogonal decomposition methods. But an implementation of this approach for the high-fidelity real-size problems becomes very expensive due to a large computation cost of the one snapshot. To escape this problem the reduced basis element method is suggested to use. The method is based on a spatial decomposition of the reactor core and computation of the reduced bases on each subdomain. An efficiency of the method is confirmed by calculations of the two-dimensional core with pin-level resolution.
