Comments on "Importance of matching software with hardware"
Dear All, since I was present at the (by now infamous) ILL workshop, I will make a brief comment. For those who still wonder what this is all about, Lachlan's talk was part of a session entitled "Data acquisition and Software", in the context of the workshop "Review of Neutron Diffraction ILL instruments for Powder and Disordered Systems", held at the ILL on March 22-23 (this week). Among many other quite interesting and useful things, Lachlan said that data analysis from TOF powder instruments was hampered by the lack of adequate data analysis software, and that some users may have to throw away otherwise good data as a consequence of this. Therefore, pulsed neutron facilities should devote more resources to developing software and worry less about building new super-duper instruments. I think that there is some inaccuracy in the phrasing of Lachlan's concept. Pulsed neutron facilities have always devoted a significant amount of resources to developing data analysis software, and this has benefitted the Rietveld community at large. The most prominent example of this is GSAS, which was written primarily for TOF diffraction at Los Alamos, and has subsequently developed into a general-purpose Rietveld tool. GSAS was recently listed as one of the most significant accomplishments of LANSCE. Another example is RIETAN, developed primarily for the KEK diffractometers but later evolved into a general-purpose code. Even when the aim is not that of providing a general-purpose tool, pulsed facilities have always developed software to help their users. Beside CCSL, which was obviously at the centre of the controversy on Tuesday, let me mention the excellent Rietveld package in use at IPNS, written by Frank Rotella and others. True, it does not do x-rays and CW neutrons, but it is perfectly adequate for data from the SEPD and GPPD diffractometers. As a consequence, I don't think there is a single case when users had to throw away data because of lack of software. If they did so, it was because the high information content of TOF data was for them too hard of a nut to crack. Besides, in terms of throwing away data, who is without sin Nevertheless, there is also some truth in what Lachlan said. TOF diffraction is a much more "software-intensive" technique than CW neutron or x-ray diffraction, and it will become increasingly so. If we take no action, we may in fact soon reach the state that Lachlan so dreads. I have full confidence that pulsed neutron facility will step up their efforts to prevent this from happening. Paolo Dr. Paolo G. Radaelli ISIS Facility Rutherford Appleton Laboratory, Bldg. R3 Chilton, Didcot Oxon. OX11 0QX United Kingdom Phone : (+44) 1235-44 5685 FAX : (+44) 1235-44 5642 e-mail: [EMAIL PROTECTED] [EMAIL PROTECTED]
Rietveld analysis on layer silicates
Hi, I am interested in doing Rietveld work on layer silicates (especially pyrophyllite at the moment). I anticipate that there may be some problems in dealing with stacking disorder and different kinds of ordered stacking sequences when developing a structural model for analysis. Being new to Rietveld I am wondering if anyone has worked on this problem. Thanks, Roger Mason
Re: Rietveld analysis on layer silicates
On Thu, 25 Mar 1999 09:42:50 -03-30, Roger Mason wrote: >Hi, > >I am interested in doing Rietveld work on layer silicates (especially >pyrophyllite at the moment). I anticipate that there may be some problems >in dealing with stacking disorder and different kinds of ordered stacking >sequences when developing a structural model for analysis. > >Being new to Rietveld I am wondering if anyone has worked on this problem. Have a look at: J. Bergmann and R. Kleeberg Rietveld Analysis of Disordered Layer Silicates Proceedings of EPDIC 5 (5th European Powder Diffraction International Conference, Parma, Italy, May 25-28 1997) Materials Science Forum 278-281 (1998) part 1 pp. 300-306 R. Kleeberg frequently does quant analyses of samples containing layered silicates. They commonly shows b/3 stacking faults and related disorder. Therefore, I have designed two special real structure models for use within the Rietveld program BGMN (see http://www.bgmn.de). Originally, I have designed real structure models for Chlorite and Kaolinite. But R. Kleeberg has successfully used them for some more layer silicates. Sincerely yours J. Bergmann [EMAIL PROTECTED]
Re: Comments on "Importance of matching software with hardware"
Paolo wrote: >... GSAS... RIETAN...Frank Rotella package... Yes, and the list is too short. But are there some ISIS software among them, or is it modesty that force you to not give their names ? Lachlan would not play correctly his role at CCP14 if he did not required more and more good freewares... Thanks to Lachlan for defending orphans and poor guys against commercial lobbies closing their codes. Thanks to all those developpers which let things slide, making software available to our community, without any request, apart glory (request may consist in being added as co-author for instance, this occuring naturally when the sofware is not at all distributed - a very tempting way to be a scientist in this jungle). Best Armel Le Bail - Universite du Maine, Laboratoire des Fluorures, CNRS ESA 6010, Av. O. Messiaen, 72085 Le Mans Cedex 9, France http://www.cristal.org/
Re: Comments on "Importance of matching software with hardware"
Paolo wrote: >since I was present at the (by now infamous) ILL workshop... As local organiser of the workshop, I thought it went rather well :-) I certainly enjoyed all the talks, especially those of Lachlan, Bill, Paolo, Juan and other Rietveld list speakers - though I thought Paolo's suggestion that we build a TOF machine on our high flux reactor went a bit far :-) as did another ISIS scientist when he suggested we should plot against Q rather than 2-theta ! >Beside CCSL, which was obviously at the centre of the controversy... For the younger ones here, CCSL stands for Cambridge Crystallography Subroutine Library, and was written by Jane Brown, who has been an ILL scientist for almost 25 years, together with Judy Matthewman. It was successfully adapted for TOF diffraction by Bill David et al., and I doubt that it can be called 'controversial'. If there is a 'controversy' it is about making software freely available, which is the key to the success of GSAS and FullProf, both distributed as compiled code. And of course both owe much to the original Rietveld code - which was also freely available - as source code. Alan Hewat, ILL Grenoble, FRANCE <[EMAIL PROTECTED]> tel (33) 4.76.20.72.13 ftp://ftp.ill.fr/pub/dif fax (33) 4.76.48.39.06 http://www.ill.fr/dif/
Re: Comments on "Importance of matching software with hardware"
Plotting in Q sounds good to me.
Another comments on "Importance of matching software with hardware"
Dear All: >Among many other quite interesting and useful things, Lachlan said that data >analysis from TOF powder instruments was hampered by the lack of adequate >data analysis software, and that some users may have to throw away otherwise >good data as a consequence of this. Therefore, pulsed neutron facilities >should devote more resources to developing software and worry less about >building new super-duper instruments. As a developer of RIETAN-98T, which is a Rietveld-refinement program for the TOF neutron powder diffractometer Vega at KENS, I will refer to only Rietveld analysis of neutron diffraction data taken with Vega. I believe that the above comments of Lachlan are far from true at least in the case of Rietveld refinements of intensity data measured with Vega. Most excellent fits are attained using a feature of partial profile relaxation. We can escape from local minima easily in most cases by the conjugate-direction method. Very stable convergence is usually attained with a modified Marquardt method adopting Fletcher's algorithm and s special technique for avoiding divergence. In addition, related softwares are available for (1) calculation of interatomic distances and bond angles, (2) Fourier/D syntheses, (3) calculation of Madelung energies and site potentials, (4) converter for a structure drawing program ATOMS. Now, we are routinely using RIETAN-98T without serious problems. However, I admit that RIETAN-98 is applicable only to intensity data taken with Vega at present and that other Rietveld-refinement programs fail in obtaining satisfactory fits in those data. Fujio IZUMI National Institute for Research in Inorganic Materials 1-1 Namiki, Tsukuba, Ibaraki 305-0044, Japan TEL: +81-298-51-3354 (ext. 511); FAX: +81-298-52-7449
