Hello All, I'm solving a multiple component non-linear reaction-diffusion problem with dealii and for time stepping I'm using the embedded explicit methods available within the dealii framework.
If I understand the documentation correctly, the embedded method should coarsen the time step if the error is less than the parameter 'coarsen_tol <https://dealii.org/8.4.1/doxygen/deal.II/classTimeStepping_1_1EmbeddedExplicitRungeKutta.html#aca772056c3e45631020ba0b3175ac2cf>'. This, however, does not seem to work! The error norm obtained by calling get_status().error_norm on the embedded explicit runge kutta method returns a value in the order of 1e-23(or even 0 for some methods) at all time steps, while my coarsen tolerance is set to 1e-10. The time step for the next iteration is still not coarsened, it always uses the initial time step I supply. The way I implement the time stepping is exactly the same as the way it is implemented in step 52 tutorial program. I would also like to mention that the problem I'm solving is stiff and hence I'm using the embedded methods for their ability to adapt step size. The questions I have are as follows: 1. What is wrong with the embedded method? Does the problem lie in my understanding? 2. Are embedded methods(with small time steps) a good choice for stiff problems or should I switch to an implicit method? Regards, Vaibhav -- The deal.II project is located at http://www.dealii.org/ For mailing list/forum options, see https://groups.google.com/d/forum/dealii?hl=en --- You received this message because you are subscribed to the Google Groups "deal.II User Group" group. To unsubscribe from this group and stop receiving emails from it, send an email to dealii+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
