adjoint restart for scalar problems (turbulence, species) do not recover residuals

[bigfooted]

It looks like in general, an adjoint restart does not recover the residuals of the adjoint variables when the problem uses additional scalar variables like turbulence or species.
This leads to additional runtime when restarting an adjoint solver.

For instance trying to restart testcases/disc_adj_incomp_rans leads to a jump in residuals, or the Inc_Species_Transport tutorial.

[Cristopher-Morales]

It looks like in general, an adjoint restart does not recover the residuals of the adjoint variables when the problem uses additional scalar variables like turbulence or species. This leads to additional runtime when restarting an adjoint solver.

For instance trying to restart testcases/disc_adj_incomp_rans leads to a jump in residuals, or the Inc_Species_Transport tutorial.

Hi Nijso,

Do you also observe similar behavior for the heat solver? specifically the testcases/disc_adj_heat/

Thank you so much in advance!!!