local timestepping
Hi,
Local timestepping can be used to accelerate convergence towards a steady state solution. The steady state is reached when the diffence in values of to successive solutions is small. However, time steps being chosen locally, the time for the steady state solution then loses its physical meaning. The question I am asking? Is it possible and useful to perform time accurate local timestepping in transient computations? Does anybody know any references in regards to this matter? Many thanks. PierreYves Lesage 
Re: local timestepping
Hi PierreYves,
There's a timeadvancement scheme called "dual time stepping" method that adopts one pseudotime variable and one real time variable. The pseudotime is used for the inner iterations and local time stepping can be still used there. To answer your question, obviously we cannot get timeaccurate solution unless same time step is used for every cell. It's expensive, but is justified since in many applications the temporal resolution of unsteady flow in question requires use of sufficiently small time step. In that sense, it's useful. 
Re: local timestepping
In "dual time stepping" method, the pseudo time step is local time. The dual time method is: Put the real time step into the source term and then solve the new equations using loacal time step. You can read the following two papers for details: 1. Alonso, J. and Jameson, A. "FullyImplicit TimeMarching Aeroelastic solutions", AIAA paper 940056 2. Gaitonde, A. L., " A DualTime Method for solution of the Unsteady Euler Equations", Aeronautical Journal, Oct. 1994.

All times are GMT 4. The time now is 08:43. 