What is a good sequence for refinement of (U,V,W) of a multiphase sample ?

[Huy LE-QUOC]

Dear Rietvelders,

I'm currently doing Rietveld refinements (with FullProf) for a 
multiphase sample whose 3 phases with space groups Fm3m, P63/m and 
I41a/md. The problem is the reflections of these 3 phases are very 
superposed and hence I have found that their (U,V,W) refined parameters 
seem to be very correlated.

I have read somewhere that for each phase we should begin refine W, then 
V and finally U. But I don't know in my case whose 3 phases having their 
own  (U, V, W) we should follow which sequence ? First refine (U, V, W) 
of the strongest phase and then for others phases? Or should we refine 
simultaneously W for all phases and then V and U for all phases?         
      

Do anyone of you have any experience on this case ?

Thank you in advance for your kindly helps.

Best regards,

---
Huy LE-QUOC,
Doctorant
LPSC/CNRS - Centre de Recherche Plasmas-Matériaux-Nanostructures
et Institut Néel/CNRS
53 rue des Martyrs, Grenoble 38026, FRANCE