Alright, this is the last bit we need for initial dG support. The pull request does two things:
First, it exposes the "relaxation order" r used for the incidence matrix for collocated degrees of freedom, roughly speaking: ((h_i + h_j)/2)^r
Second, it introduces a configuration parameter for the Indicator (currently implemented for Euler and EulerAEOS) that allows to switch to a more aggressive split for the denominator. Using this split recovers expected convergence rates for the isentropic vortex, rarefaction and LeBlanc. Furthermore, nonsmooth problems such as the Mach 3 Cylinder show a comparable resolution for dG Q1 versus cG Q1 for the same number of degrees of freedom (i.e. dG computed on a once coarsened mesh compared to cG).
Let's get this in. I will follow up with
some test vectors for the testsuite and some test vectors for the benchmarks category for dG Q0 and dG Q1.
implementing the same Indicator logic for shallow water and scalar conservation (which still need some testing).
Alright, this is the last bit we need for initial dG support. The pull request does two things:
Indicator(currently implemented for Euler and EulerAEOS) that allows to switch to a more aggressive split for the denominator. Using this split recovers expected convergence rates for the isentropic vortex, rarefaction and LeBlanc. Furthermore, nonsmooth problems such as the Mach 3 Cylinder show a comparable resolution for dG Q1 versus cG Q1 for the same number of degrees of freedom (i.e. dG computed on a once coarsened mesh compared to cG).Let's get this in. I will follow up with
benchmarkscategory for dG Q0 and dG Q1.