Need help from LAMMPS/AtC community

LAMMPS LAMMPS Mailing List Mirror
1

Dear LAMMPS/ATC users:

I guess my questions should be addressed to Drs. Zimmerman, Jones, Templeton or other AtC folks.

I’m playing with AtC package trying to find the pros and cons of using a coupled FE/MD simulation for predicting the mechanical and thermal properties of solids. Most of the tests and input scripts provided with the package work fine and I am truly impressed with the ATC code, but sometimes I face with lack of some useful functionalities. Here are two examples:

  1. After a few days of playing around with the simple input files found in the examples/USER/atc/thermal sub-directory of the LAMMPS public release (10 Aug 2015), I was confident that I could produce any script with minimal effort. Unfortunately I cannot visualize my results when I use the thermal fix and the binary output (Ensight *.case file): the temperature field is present in the text .DATA file, however, it is not displayed by Paraview. How can I sort out this problem? In particular, how can I generate with Paraview/EnSight software the temperature fields like those shown in Fig.7 in the paper of Wagner et al. (Comput. Methods Appl. Mech. Engrg. 2008, 197, 3351)?

  2. My second question is related to the in.eshelby_static input file from the examples/USER/atc/hardy sub-directory. I understand that the FE region can be larger than the MD region, but how can I visualize the coarse-grained stress fields outside the MD domain, not only inside it? Rephrasing my question, is it possible to analyze the whole FE+MD domain when FE >> MD? In this regard I’d like to refer to a nice picture from Zimmerman’s presentation (2013) where J-integral calculations are discussed (see e.g. slide #9). I would be grateful for examples of the corresponding input scripts.

Thanks and best regards,

Helmut Grauer

The University of Ulm

89069 Ulm,

Germany

2

Dear LAMMPS/ATC users:

I guess my questions should be addressed to Drs. Zimmerman, Jones, Templeton
or other AtC folks.

since nobody has yet responded, i would guess the answer to this is: yes
and the recommendation to contact them directly.

axel.

3

I forwarded your message to the AcT folks. You can
wait a few days and see if they respond.

Steve

4

Hi Helmut,

re: Q1, if you are using ATC in coupling mode e.g.

fix AtC internal atc thermal Ar_thermal.mat

and specify Ensight compatible output:

fix_modify AtC output bar1dFE 100 text binary

then the *.case file should have a line like:

scalar per node: 1 1 temperature bar1dFE.temperature

which means you can load the case file in paraview and then select “temperature” as the nodal field to display on the mesh. If this isn’t true something is very wrong

re: Q2, this is pretty much the same issue. The “fields” are nodal values associated with the mesh and can be used in paraview to create the lovely rainbow plots you mention. If in addition you want to display the point data directly associated with atoms you can use the attached script that is a thin wrapper around the LAMMPS dump.py and other python code. You give it the dump file and get an Ensight compatible file/s in return.

Hope this helps,

Reese

​

dump2ensight.py (2.09 KB)