[pattim]

Has anyone used the reactingFoam "transsonic" option and gotten it to work? It seems like a pretty robust solver, but in particular the BC's might be difficult to set up. Are input/output totalpressure and velocity-zerogradient are what are required?

https://www.openfoam.com/documentati...ctingFoam.html

[dlahaye]

I am wondering about the same. It might make sense to

1/3: run rhoSimpleFoam / rhoPimple using the transonic option: see tutorials and Chapter 16 of Moukalled e.a. for setting up the bc;

2/3: run reactingFoam with the combustion switched off (see e.g. initialization of SandiaD tutorial on how to switch off the combustion). Similar (same?) results as in Step 1/3 are expected;

3/3: run reactingFoam with the combustion switched on;

[pattim]

Originally Posted by dlahaye

I am wondering about the same. It might make sense to

1/3: run rhoSimpleFoam / rhoPimple using the transonic option: see tutorials and Chapter 16 of Moukalled e.a. for setting up the bc;

2/3: run reactingFoam with the combustion switched off (see e.g. initialization of SandiaD tutorial on how to switch off the combustion). Similar (same?) results as in Step 1/3 are expected;

3/3: run reactingFoam with the combustion switched on;

That would be good. I did notice that reactingFoam generates some reasonable-looking shock structures without the transsonic option; although, I don't think I had the BC's set up correctly (should be stagnation pressure at the inlet, total pressure at the outlet).

Thanks for mentioning the SandiaD tutorial - it's nice how the adjust the run to develop the flowfield. I'll have to test that and see how well it works.

[dlahaye]

Thx. I will be happy to hear from your experiences.

[pattim]

Originally Posted by dlahaye

Thx. I will be happy to hear from your experiences.

Well, it's a little off-topic but, basically, chemical reactions can be very fast and can dominate flowfield values by orders of magnitude. Physically, you can make a fully-developed flowfield, then turn-on chemistry, and obtain a deflagration wave or a deflagration/detonation transition, both of which can easily crash solvers. Also, the flowfield with chemistry is often very different from a flowfield without chemistry. (For instance, a parcel of 5% hydrogen in air will detonate when ignition is applied.) So unless the chemistry is slow/simple/highly-dilute, calculating a no-chemistry flowfield beforehand wouldn't be of much use.