[cfdonlineuser]

I am loading a table of 3D data for a user function USERdensity=f(P,T,dummy) where P is absolute pressure and T is temperature and dummy is a dummy variable needed when one only wants a 2D function. But when I plot the data for a CEL function for the user function USERdensity using the plot utility in CFXpre there are all sorts of spikes in the CEL function plot, which are (a) not in the original data, (b) not in the table of data listed in the user function for USERdensity.

Anyone had similar a experience with user function as a 3D interpolation table ?

[brunoc]

This is just a guess, but I'd say this is probably because the 3D table is meant to be used with coordinates, so it does a 3D interpolation to get your values based on the 3 closest points in the table. But since you're dealing with properties, the correct way to get the values from your table should be to intepolate between the four closest points, which in your case is not being done.

Also, since you have absolute pressure and temperature it is very likely that they're of different magnitudes and that might be causing additional trouble for the interpolation process.

Cheers

[cfdonlineuser]

thanks brunoc, I suspect you're right, theres something fundamental to the calculation that causes the wrong gradient be calculated.

Ill keep you posted