Hi all,

I have implemented a quasi-1D supersonic combustor-code for scramjet applications with an implizit 2nd-Order-Upwind finite differences discretization of the unsteady conservative Euler equations, coupled with the conservation equations for the spezies concentrations. The solver shows very good stability and quick convergence (< 500 timesteps) for ideal gas or equilibrium chemistry using CFL-numbers in the range of 1-2. But for finite-rate chemistry, the timesteps have to be very small (CFL < 1e-3), because of the chemical source term, which leeds to extrem numbers of timesteps (<50000) until convergence is acieved. I wonder if this behavior is normal, or something's wrong with my code (the solution looks OK). Is there any possibility to accelerate the solver, or to increase the timestep, so that the chemical source terms do not result in an unstability. Thanks for any advice.

Cheers Daniel

Daniel,

It is quite usual that the kinetics prescribe small time step. One way to accelerate the solution is to switch off the reaction and run to convergence, then switch it on, reduce the time step as necessary and restart.

I hope this helps, Rami