Thanks for your reply Rob,

You are right. activating the operating density sounds like a better solution for that. I am using the "real gas redlich-kwong" material model for my airflow and I get the following warning when I changed my operating density to 1.1845 kg/m^3:

Warning: For compressible (ideal and real) gas models with buoyancy,
 it is recommended that you use a specified operating density value of zero:

Should I ignore it or will it cause any problems you think?