[Viento]

Hello all.
My objective is to simulate a valveless pulsejet in various operating modes, using ignition of premixed or partially premixed mixture from a spark and/or from hot gases.
An important thing is the propagation of the front of flame, local zones with supersonic flow and local zones with autoignition of mixture.
I'm trying to find the best model for this task.

1. What turbulence model is better to use, and what additional factors I must take into account?
2. What combustion model is better to use in this case, BVM or Extended coherent flame model?
For example, I runned two tests of propane combustion in a closed volume. In BVM the spherical front has been created, transforming into a uniformly moving inverted cone. In ECFM has been created a more realistic sharped form and a high flame velocity along the walls, but at the end the reaction formed a cavern and almost stopped. (I dont know why.)
(Pics in attachment.)
A standard may be something like this: https://www.youtube.com/watch?v=itX7yMgCCxA

3. What else should I pay attention to?

Sorry if I missed some obvious points, I'm novice in CFD. Thanks.

[ghorrocks]

Aren't you the best person to figure out all those questions? You should test the various options and compare them against benchmark solutions or experimental results so you can work out in your case which models work best for you.

Your case is very specialised, so you are likely to know more on this topic than anybody else on the forum.

But I can answer one question: "What else should I pay attention to?" - read this FAQ: https://www.cfd-online.com/Wiki/Ansy..._inaccurate.3F and https://www.cfd-online.com/Wiki/Ansy...publishable.3F