Calculating the self-diffusion from Forcite MSD is a standard technique. For example, I've calculated the self-diffusion of my polymer chains in solvents by applying the instructions in an Accelrys tutorial. Now I'm pursuing the calculation of mutual-diffusion.

I've identified a technique by Krishna and van Baten (LINK) which requires the x,y,z displacements of both polymer and solvent. Prior to taking the effort of writing a script for extracting these displacements and calculating the given equations, I'm wondering if I can automatically export displacement data (or data which leads to these values). My Accelrys version is 4.3.

The approach I'm currently trying but having trouble with is as follows:

1) Create a set of polymer centroids.

2) Extract MSD data from Forcite (including anisotropic terms XX, YY, ZZ) for centroids

3) Reverse the cumulative sum on each XX, YY, ZZ value and keep sign (+/-)

4) Square root abs(XX), abs(YY), abs(ZZ) and incorporate sign

5) Now I think I have the X, Y, Z coordinates (?) <-- Please advise.

What is the relationship of XX, YY, ZZ (and XY, XZ, YZ) MSD anisotropic components with position coordinates X, Y, Z?

Is there a way to access the code how Accelrys calculates MSD? Though the formula for MSD is given in the help files, I have a feeling the trick is in the averaging of MSD to remove noise.