Dear Yaroslav, Dear Bob and all, thanks for your hints. I don't have some difficulties with negative FWHM^2.
I try to refine a pattern of LaB6 reference material with GSAS. Data were obtained with a Siemens D5000 diffractometer. So after refining background, zero, Lx, Ly and lattice dimensions i try to refine GU, GV and GW simultaneously (Lx, Ly, background, lattice dimensions and zero are flagged for refinement too). Starting values are GU=2, GV=-2, GW=5. After refining the plot looks better and CHI^2 is better too (nearly 1.6). THe FWHM^2 calculated by widplot is around 0.06°. GU and GW are positiv but the GV parameter is now positiv too (between 2 and 4). Is this trustable or should i change something in my refinement procedure (may be unflag Lx and Ly...)? Thanks in advance, Stefan -----Ursprüngliche Nachricht----- Von: Von Dreele, Robert B. [mailto:[EMAIL PROTECTED] Gesendet: Mo 6/25/2007 19:27 An: rietveld_l@ill.fr Betreff: RE: Caglioti U V W parameters Dear Stefan (& all, I suppose), >From the original formulation by Caglioti, et al. U>0, V<0 & W>0 for a "nonfocusing" neutron CW instrument and describes a parabolic curve with the minimum at roughly the 2-theta angle that matches the monochromator "take-off" (really 2-theta) angle. For a Bragg-Brentano powder diffractometer the curve is usually quite flat yielding very much smaller FWHM especially for the low angle portion rising only at high angles. I suppose the minimum is about the 2-theta for the analyser crystal. Thus, U, V & W are usually quite small but the relationship of U>0, W>0 & V<0 still holds. They can be a bit hard to determine unless a very high quality pattern is used for the calibration. The expression using U, V & W yields (FWHM)^2 so it REALLY can't ever have a negative result as Yaroslav noted in his message. Rietveld refinement codes will protect against the possibility of a negative square root in various ways (if not they can crash on a "negative sqrt" error). Bob Von Dreele R.B. Von Dreele IPNS Division Argonne National Laboratory Argonne, IL 60439-4814 -----Original Message----- From: Stefan Berger [mailto:[EMAIL PROTECTED] Sent: Monday, June 25, 2007 11:18 AM To: rietveld_l@ill.fr Subject: Caglioti U V W parameters Dear All, i am a little bit confused about the magnitudes of the U V W profile parameters of the caglioti function for instrumental broadening. Is it true, that U and W are always positiv (larger 0) and V is always smaller 0 (negativ)? Thanks for your hints and explanations. Best Regards, Stefan
<<winmail.dat>>