I'm trying to specify several walls of varying roughness in CFX4.x. To do this I am using USRWTM to modify the wall turbulence multiplier for each wall in turn. The problem is, how do I relate RHT (the roughness height) and ELOGR (the loglayer constant) to the physical characterisitics of the roughness elements and the flow regime itself. CFX tech support have been particularly poor in answering my questions in this area. Any ideas/pointes anyone?

If all else fails, go back to the pipe friction factors found in most fluid mechanics textbooks. At least you can count on experimental verification of the results.

Most of the correlations available in the literature for the relationship between the roughness height and the intercept in the log plot (ELOG) are given in terms of non-dimensional quantities (k+), which involves the friction velocity (utau). The friction factor changes of course from point to point anf from iteration to iteration and is a part of the numerical solution, making it impossible for us to specify in advance the ELOG in terms of roughness height.

In my implementation, you can specify the roughness heights for different walls and the program computes ELOG using the friction velocity as ther solution evolves.