anne-glerum

Mmm. not sure about that. Does the RMS velocity decrease over time? Is elasticity on?

Without elasticity, does it get to an equilibrium?

``` ----------------------------------------------------------------------------- -- This is ASPECT, the Advanced Solver for Problems in Earth's ConvecTion. -- . version 2.3.0-pre (master, c268d7d) -- . using deal.II 9.3.0-rc1 -- . with 32 bit...

I'll dig in a little deeper.

Low res models (1 proc): - On one proc with a smaller problem size, the norm of the rhs still differs ~1 OOM. - When melt transport is turned off,...

For the convection box, both the matrix and rhs are different after restart, but in the second timestep after restart, they have recovered. For the mass_reaction_term test, only the rhs...

I don't see differences in linear iterations for the convection-box model either, only in the norms. With the test mass_reaction_term on 1 proc in debug, the nonlinear residuals of the...

Thanks @gassmoeller ! 1. That makes sense, sorry for the side track. 2. I still see some differences after a restart with a custom melt model. 1. The nr of...

I'll rebase my custom branch to master (and keep the linearization point fix) and try again.

Using `current_linearization_point = solution;` gives identical timesteps before and after restart for both master and the rebased custom branch for the `mass_reaction_term.prm`. However, the custom melt model & prm with...