Nonreflecting far-field for transonic flow

Bren

Hi guys, I've written a solver that deals with the low frequency transonic flow around an airfoil. I've recently updated the model to include the first order nonreflecting far-field boundary conditions proposed by Kwak.

My code converges most of the time but sometimes I'm noticing a rapid increase in the velocity potential at the edges of the domain. This eventually becomes the dominant area of potential and messes up the flowfield. I haven't managed to work out why this happens. I followed Kwak's paper closely when incorporating his equations into my model and cannot find where I'm going wrong. I am tearing my hair out trying to fix this problem.

If anyone else has tried to use these boundary conditions and has had similar difficulties I'd really appreciate some advice.

Many thanks Bren