Despite continuous progress in computer technology, modelling neutron flux transients accurately remains challenging in terms of resources and runtimes with existing neutron transport codes. This facilitates the need to develop new high-fidelity, time-dependent transport solvers and examine their performance. This PhD project explores novel time-dependent neutron transport approaches based on the random ray method, a stochastic variation of the method of characteristics. Two distinct versions of time-dependent random ray methods are being developed. The first version is a time-implicit method, in which the spatial distribution at a given time step is brought to convergence before moving to the next time step. The second version applies an unconventional approach: it converges the entire space-time shape over a number of iterations using time-continuous rays. Both versions are to be tested on various benchmark cases.

Project team: Maximillian Kraus and Eugene Shwageraus