[abt2024]

Hey all. I'm trying to see if I can simulate a completely sealed pressure vessel that has a fuel and air mixture experiencing an ignition event. This is for a safety investigation at my university. An illustration of the scenario, for now in 2D, is shown in the attached screenshot.

I want to conduct a transient simulation using the Finite Rate: Chemkin-CFD chemistry solver (NH3 mechanism from Stagni) to see at what max temperature and pressure the inside of the vessel reaches and the time it takes to reach that point. I'm using the K-Epsilon Realizable turbulence model with Enhanced Wall Functions and Production Limiter. For Pressure-Velocity Coupling I'm using PISO. All Discretization is Second Order Upwind. The Gradient is calculated using the Green-Gauss Node-Based method. The Residual values are left at default with the Iterations Per Timestep being 50. The ignition source is a 2000 K patch during the initialization of the simulation, so the ignition occurs at the very start of the simulation.

So far my attempts have been shakey, even when using both the combustion preset AMR and adaptive timestepping (CFL = 1). The main problem is the continuity residual, as it often struggles to get below a residual of 1 or even flatout tries to diverge, whereas all other residuals tend to be comparatively stable. This occurs even when I set the CFL target to less than 1 (never tried 0.5 or below). At first, I figured this sort of simulation would be relatively straightforward, but now I'm finding it to be a pain without fully understanding why. Any help on my setup and understanding will be greatly appreciated.