Yes, you're probably improving the conditioning of the operator by making the mass matrix in the d/dt term more important. Try switching on the monitors in the solver and see if the number of solver iterations has dropped. Regards, David On Thu, 11 May 2017 at 10:13 Anna Kalogirou <A.Kalogirou@leeds.ac.uk> wrote:

Dear all,

I have noticed that when I divide one of my equations in a mixed system by the time step dt, the simulation time drops by more than half -- from 7m13s to 3m7s for the same simulation. Is there an explanation why this might be happening?

Regards,

Anna. _______________________________________________ firedrake mailing list firedrake@imperial.ac.uk https://mailman.ic.ac.uk/mailman/listinfo/firedrake

-- Dr David Ham Department of Mathematics Imperial College London