I am using LAMMPS (23 Jun 2022 - Update 4) -version
I am trying to simulate liquids on top of a graphite surface to measure the contact angle. the binary mixture of 900 water + 100 ethanol molecules is used. This system is already equilibrated. Using nanotube generator in VMD 2 layers of graphene sheet are generated. using topo tools those structures are merged. the data file and the input script are attached. while running the simulation I am getting an error: Out of range atoms - cannot compute PPPM (src/src/KSPACE/pppm.cpp:1887).
The simulation crashes at the first step itself. I tried increasing the box dimensions, but it didn’t work. the same script and procedure previously worked for me in a different systems where I used water-methanol binary mixtures.
input.lammps (1.0 KB)
graphene_eth_1.data (6.4 MB)
This is probably the most asked topic in LAMMPS’ history. Please search the forum for the countless advice already given.
I can only add that doing random changes* won’t solve the problem. You first need to understand what is causing it. In your case, one possible factor is the box boundaries:
The correct value is not a matter of guessing, but has to match the supercell indexes.
* Note
There is a more effective way to express the concept in Italian and that is: alla cazzo di cane.
1 Like
If the input has been worked before, then you probably should focus your attention to the data file, although also settings in the input may need to change if your system changes much.
Thus, I would recommend to reduce the thermo and dump output frequency to 10 (or even 1) and look at what happens. I guarantee you all kinds of unexpected things like atoms moving fast, atoms moving in unexpected ways, unexpectedly large potential energy contributions, and unexpected quick temperature rise to unreasonable values.
As @hothello mentioned, the same kind error report has been posted and the question has been asked and answered many times before, so a look into the archives (or a web search) can be very illuminating.
An important aspect of the problem, since there can be multiple causes, it to realize that “out-of-range” does not apply to the simulation box but to the individual subdomains when running in parallel. The only coincide when running in serial.
