I have calculated the flow field in a duct using a mesh that had a low number of cells (4000), and I had no problems. But after mesh refining (32000 cells) the calculation stops and gives the following error:

" negative densities found at more than 100 cells"

What kind of problem is it? How can I fix it?

Thanks everybody,

Fabrizio

(1). I think, the person who wrote the code knew that this negative density problem could happen in the code, so he put in a line of code to check the density and also print the message to give the user a warning. As expected, he was right. Now you are having the same negative density problem, and the code works fine because you are getting the printed message of negative density at certain cells. (2). I suspect that if he had a solution to it, he would have removed the printed message already. When he was writing the code, I don't think he knew that you will be doing the same thing later on. (3). I think, he ran into a man hole before, so he decided to put a warning sign there. Now you are in the same man hole, and the sign is still there. (4). So, it is your business to either avoid the man hole, or complain to the city hall.

Redefine your inlet and outlet boundaries and pressure reference cell number after the refinement.

Fabrizio, look at the file (case.info) and see what went wrong in the solution to have (-ve) pressure or temperature causing these (-ve) densities, then come back to the forum and discuss your findings. Don't worry, it's not a big deal.

I have had the same problem! It seems to depend on the mesh quality. I suggest that you look for bad cells and try to fix them.

It is much more important to have a good mesh when you solve for the temperature.

Regards,

Lennart Wern

Swedish Meteorological and Hydrological Institue