I would like to start a general discussion on relaxation factors in PHOENICS, something that all users of PHOENICS could benefit with. I have often found that relaxation factors are very difficult to determine, and ultimately dictate whether or not convergence is achieved.

If any one has advice/suggests for obtaining the proper relaxation factors, and in obtaining these in reasonable amount of time, I invite you to post your strategies. It will be something where we can all learn from each other, and make our lives easier in this field.

Tim,

A very interesting automated approach via fuzzy logic was presented by Prof. Norberto Fueyo at the PHOENICS Users conf. in Luxembourg, 2000. He promised me and several others to send the paper and source code.

Since then I had tried to remind him many times by e-mail (Norberto.Fueyo@posta.unizar.es) but received no response. CHAM had confirmed the e-mail address was correct, and so does his website.

Maybe you will be luckier. I really think his treatment can be very useful.

Rami

Rami,

Thanks for your input. I will see what I can do about obtaining a copy of the paper.

Tim

The paper cited in the conference was

Fueyo N, Blasco JA, Relaxation control in the solution of CFD problems, INTERNATIONAL JOURNAL OF COMPUTATIONAL FLUID DYNAMICS, 13 (1): 43-63 1999.

Good luck!