Am I correct in thinking that CFD can only model steady flows? There are a number of components in an exhaust system whereby pulsating flows can be exploited, for example Kordinas and de Laval nozzles. Also I think the inertia of the gas in the long chimney of a Lempor will have an effect. If the velocity of the gas is rising and falling all the time, what happens to the energy when it is reducing? Atmospheric pressure isn’t going to change, but the “peristaltic or piston” effect could cause some additional vacuum in the smokebox.