A posteriori analysis for dynamic model adaptation in convection dominated problems
This work provides a principled approach for model adaptivity in computational fluid dynamics, reducing computational cost while maintaining accuracy.
The authors derive an a posteriori error indicator for Runge-Kutta-discontinuous-Galerkin schemes in compressible fluid flows, enabling dynamic model adaptation between a costly complex model and a cheaper simple model. The indicator identifies regions where the complex model can be replaced by the simpler one, demonstrated with Navier-Stokes-Fourier approximated by Euler equations.
In this work we present an a posteriori error indicator for approximation schemes of Runge-Kutta-discontinuous-Galerkin type arising in applications of compressible fluid flows. The purpose of this indicator is not only for mesh adaptivity, we also make use of this to drive model adaptivity. This is where a perhaps costly complex model and a cheaper simple model are solved over different parts of the domain. The a posteriori bound we derive indicates the regions where the complex model can be relatively well approximated with the cheaper one. One such example which we choose to highlight is that of the Navier-Stokes-Fourier equations approximated by Euler's equations.