Integral Method and Eigenspace Decomposition for RANS Turbulent Mixing Flows

Integral methods can be used to obtain low-order dynamical systems which approximate the solution of RANS models. Such techniques are beneficial for calibrating coefficients and verifying a model for flows away from self-similarity. We apply an integral method approach to turbulent mixing flows and compare analytic results to full-field RANS simulations in xRage, a code developed at Los Alamos National Laboratory. Flows that exhibit late time self-similarity are considered. Eigenspace decomposition is then used to identify dominant solution characteristics and visualize higher dimensional closure models. Results are shown for the temporal shear and Rayleigh-Taylor layers.