I have 3.5.0 & 3.6.0 versions of Phoenics. Why my solution obtain the convergence in 3.5.0 but in 3.6.0 is not? May I ask - what version is "more right"?

Phoenics 3.6. In-form formula (PROPERTY of PRNDTL(TEM1) is 40 with IMAT=111). As result conductivity is 20. Why?

If you are a supported user send it in to support.

These are the type of problems that can take ages to resolve.

Partly because users have such large grids.

You could investigate this yourself if you are systematic

I would advise 1/ reducing mesh density 2/ simplifying problem until results match 3/ Can try and switch off PARSOL see if this makes a difference 4/ make sure you compare results on convergence.

Thank you Mick. But you advise me to solve my problem again. I have no time to reducing mesh density (and i cann't understand why, if i have phi file from ver 3.5)... I try to simplifying my problem and switch off PARSOL (my problem hasn't partial solids) - it cann't helps... I think - problem in k-e model in ver. 3.6.0. And what you can saying about INFORM question?

Hi Timur, about the convergence problem Its impossible to say without examining your case, however a lot of improvements and fixes have been made to v3.6. I would examine the RESULT file for each case and look for any differences. It might be that some default setting has changed. Alternatively, if it deoes not diverge wildly, perhaps its just a matter of a small change to the relaxation settings.

about the inform.. I have just had a look at POLIS and I believe that to set the conductivity using PRNDTL and inform inform9begin (property of PRNDTL(TEM1) is 400 with IMAT=111) inform9end

should work, it does with the latest v6.0 that I am using.

If it does not work a way around is to edit the props file for material 111 and change the conductivity.

mick

Thanks Mick. I can send you my q1 and phi files if you have the time. I have ver. 8.3.2.4 of INFORM. Unfortunately in props file I cann't set conductivity which depends of temperuture (there is no GRNDX for conductivity). I also can send you another q1 file.