Introducing the Constant-Density Ideal Gas Law Buoyancy Model

[jvangilder]

Here is post for those interested in some details of our underlying CFD technology... Recall that warm air rises and cold air sinks. We call this buoyancy. Well, we have to model that physics in ITA CFD. In fact, we just changed the buoyancy model in ITA CFD from a traditional Boussinesq approximation which uses a fluid property called the bulk thermal expansivity to what we are calling a constant-density ideal gas law model. We still assume that the air density is constant except in the buoyancy term in the vertical-direction momentum equation; for the latter term, we compute density variations due to temperature changes from the ideal gas law. As you may know, air conveniently behaves as an ideal gas in most conditions.

Our testing has shown negligible differences in accuracy for most practical applications - and no obvious impact on solution robustness. So why did we make the change? Two reasons: 1) We eliminated the need for the bulk thermal expansivity fluid property - simpler is better! 2) The new model is, at least, theoretically more accurate; this may be noticeable in scenarios with large temperature differences and strong buoyancy forces.