Hello all
I am simulating hydrogen generation in an electrolyzer compartment using Eulerian model in Fluent. The final amount of gas in the domain when the solution reaches steady-state condition is about 25% and 14% for Explicit and implicit solvers respectively. What is the reason for this difference? I have checked even very small values for the time-interval to see if two solutions become similar, but they did not. Any idea about the reason for this difference between the results? How can I get similar results?
I think you should check mesh independency for your solution, changing the solver, will result in change in the mesh dependency criteria.
I think you should check mesh independency for your solution, changing the solver, will result in change in the mesh dependency criteria.
you are right... I have done the mesh dependency tests...Actually the difference that I mentioned in my question obtains after the mesh dependency tests...
In general, using implicit methods you have to consider the accuracy in the solution of the additional system (I consider that implicit and explicit time integration are of the same accuracy order). Using iterative methods you should be careful to consider a correct convergence criteria. I suggest to decrease the convergence threshold and compare again to the explicit-based solution using the same discretization in space and time-step.
Furthermore, is the comparison done at the same criteria of steady solution?
Hi Filippo
the convergence threshold is the same for both cases. Actually, I have set a very small value (10E-12) for this purpose. And yes...all parameters including criteria of steady state solution are the same for both simulations.
I am not sure if we are talking about the same things... using explicit methods you do need not solve the system that appears conversely in the implicit method...I am talking about the stopping criteria used fro this latter. Be carefull that the solution is accurately computed.
- Badges
- Science topic