[RMCProfile-users] Fitting Structure Factor and Bond Angles

[helen.playford at stfc.ac.uk]

Hi APEE,



In answer to your first question, there isn't a particular flag for general bond angle constraints. The program has several system-specific polyhedral restraints which can be used to maintain bond angles and lengths in accordance with the connectivity of a network of polyhedra (e.g. SiO2, or indeed any tetrahedral network - the atom types don't matter).



Alternatively there is the capability to use interatomic potentials (applicable particularly to molecular systems) to constrain bond angles in accordance with the energy penalty associated with bond stretching or bending.



Hopefully one of these options will be suitable for your system, and more information about both is in the manual.



Your second question is a rather difficult one to answer! There isn't an inbuilt convergence criterion, so when a refinement is 'finished' the goodness-of-fit parameters will continue to oscillate slightly around a minimum point until the time limit is reached. The number of moves and therefore the length of time you will need to run depends on many things such as the number of datasets you have, the number of atoms in your configuration box and how far from the 'answer' is your starting point. My refinements tend to run for days rather than hours, and I check periodically to see if the fit is still improving.



Please do get in touch if you need further advice.



Best wishes,

Helen



________________________________

From: rmcprofile-users [rmcprofile-users-bounces at rmcprofile.org] on behalf of Pandey, Anup [ap439111 at ohio.edu]
Sent: 17 July 2014 19:58
To: rmcprofile-users at rmcprofile.org
Subject: [RMCProfile-users] Fitting Structure Factor and Bond Angles


Dear RMCProfile Users,


I am a first time user and have couple of questions on using RMCProfile. I want to fit experimental Structure factor(SF) and fix bond angle as a constraint in a system. I have created an input file that runs quite well for fitting the SF but I could not find any direct FLAG to constrain the bond angle.


Secondly, I am wondering how long should we RUN the program or how many steps should we run before it converges to the experimental data. The time tag is in minutes and that is what confuses me. Is there a way to fix a convergence factor so that the program stops on its own once the required convergence is reached.


Any help will be highly appreciated.


Thanks !!!

APEE