Dear all,
I'm a new user to apply a ReaxFF force field. Before I use it, I'm doing some tests of some simple systems. When I was doing a simulation of a Diamond system under NVT (actually under NVE it also has similar problems), the diamond can not maintain the cubic lattice structure. I checked the thermo output and found the temperature increased sharply at the first few steps, with the following warning.
Step Volume PotEng c_eatoms Temp
1 23231.096 -780141.38 -780141.38 300
WARNING: Fix qeq/reax CG convergence failed after 200 iterations at 8 step (../fix_qeq_reax_omp.cpp:571)
WARNING: Fix qeq/reax CG convergence failed after 200 iterations at 8 step (../fix_qeq_reax_omp.cpp:571)
WARNING: Fix qeq/reax CG convergence failed after 200 iterations at 9 step (../fix_qeq_reax_omp.cpp:571)
WARNING: Fix qeq/reax CG convergence failed after 200 iterations at 9 step (../fix_qeq_reax_omp.cpp:571)
WARNING: Fix qeq/reax CG convergence failed after 200 iterations at 10 step (../fix_qeq_reax_omp.cpp:571)
WARNING: Fix qeq/reax CG convergence failed after 200 iterations at 10 step (../fix_qeq_reax_omp.cpp:571)
10 23231.096 -475827.66 -475827.66 445128.37
WARNING: Fix qeq/reax CG convergence failed after 200 iterations at 11 step (../fix_qeq_reax_omp.cpp:571)
WARNING: Fix qeq/reax CG convergence failed after 200 iterations at 11 step (../fix_qeq_reax_omp.cpp:571)
WARNING: Fix qeq/reax CG convergence failed after 200 iterations at 12 step (../fix_qeq_reax_omp.cpp:571)
WARNING: Fix qeq/reax CG convergence failed after 200 iterations at 12 step (../fix_qeq_reax_omp.cpp:571)
I searched and was awared that it usually means a bad initial geometry. But in my current test system, it only consists C atoms in cubic lattice structure. I can't identify where the error is.
I also have tried several ReaxFF from different references. They shows different strucures but all with similar problems at maybe different few steps.
Have anyone met similar problems before and solved it?
Thank you very much for your help!
Dear Shengyao Yang ,
Could you share your LAMMPS parameter, so others can help to analyze the problem and give suggestions? It is difficult to find the problem without seeing the input parameter.
I expect there is a problem when applying the fix qeq. However, I can't suggest if I don't have the complete line of fix qeq.
Regards
There might be a lot of things that can get you these type of results.
Firstly, is this forcefield used previously for the diamond structure.
Check timestep size it should be small around 0.5 or 0.2 fs try to work that out
Units with most reaxx files should be in real
Also there might be other problem,
You can share input script so that we can help you
Dear all,
I think I have found the reason. The problem has been solved by adjust the timestep from 1 to 0.25 or lower.
Thank you all.
Dear Nitin Luhadiya ,
You are right, I found this because when I change unit from metal to real and forgot to switch the timestep, the simulation can run smoothly. This tells me the problem might be timestep things, just like you mentioned.
Thank you for your answer!
Dear Khabib Khumaini ,
Thank you for your reply. I have solved this, currently it is not the problem regarding the fix qeq, it is about the timestep.
Your suggestion is good, if I have any problems in the future, I'll have my LAMMPS parameter included.
- Badges
- Science topic