I am using the compute msd command in LAMMPS to calculate the diffusion coefficient of LiF with the Lennard-Jones (LJ) force field.
The resulting MSD curve is horizontal at low temperatures (below 500 K) and shows a step-like pattern at high temperatures, sometimes even with decreasing segments. How can I address this issue?
Hi @cababe ,
What is the system you are studying? It is impossible to define how “unreasonable” your results are if we don’t know which physical behavior to expect. Which are the atoms you are computing the MSD from? In what context? Your results simply look like diffusion with discrete hopping between low energy sites and local vibration in between. This could be expected in the context of surface diffusion of diffusion in a solid with defect. Below a certain temperature, your atoms would simply lack the kinetic energy to overcome the energy barrier during the simulation time. Nothing wrong with that.
@alphataubio I think a LLM trained specifically on LAMMPS data could be very useful in some cases. And I think it is totally appropriate to advertise a new tool, and even point users to the tool’s URL. However, for a MatSci reply, can you please not copy/paste a large amount of output from the AI model directly into a reply? I think this is applicable for any AI model, not just LammpsGPT. There can be subtle mistakes or inaccuracies that get propagated. This is the policy for other forums like Stack Overflow as well, see What is this site’s policy on content generated by generative artificial intelligence tools? - Help Center - Stack Overflow.
We are planning to add a new forum guideline regarding the use of AI for responses here: Please Read This First: Guidelines and Suggestions for posting LAMMPS questions. You didn’t know because we didn’t have the rule posted, so please don’t feel bad. Thanks
Have you visualized your trajectory? Is the system behaving liquid-like?
Please note that the shape of the MSD curve will always be “reasonable” since it is correctly applying the algorithm to the positions in your system. This has been confirmed many times.
The question is now: does the system behave the way you expect? or are your expectations unreasonable?
See this thread and my comment here: Question related to MSD - #11 by srtee
My system is a LiF bulk with 998 atoms, missing one Li+ and one F-. I’m calculating the MSD of Li ions to get the diffusion coefficient. I understand why atomic motion is restricted at low temperatures. How can I use the MSD to calculate the diffusion coefficient at high temperatures?thank you so much!
My system is a LiF bulk with 998 atoms, with one Li+ and one F- defects,the system isn’t behaving liquid like,According to the formula for MSD, I expected the curve to show an upward trend, but it did not. How should I calculate the diffusion coefficient of lithium ions? Thank you so much!
