Maximum Number of Atoms allowed in *mcf?

  • jdmoore
  • jdmoore's Avatar Topic Author
  • Offline
  • Junior Member
  • Junior Member
More
7 years 9 months ago #174 by jdmoore
Hello,

I think I ran into a problem with the maximum number of atoms allowed in the *mcf. It looks like in input_routines.f90, that the maximum number of lines is 10,000, otherwise it triggers errors and stops the code. Is this a limitation of the code, or can I just up the number of lines in input_routines.f90?

Thanks.

Josh

Please Log in to join the conversation.

  • jdmoore
  • jdmoore's Avatar Topic Author
  • Offline
  • Junior Member
  • Junior Member
More
7 years 9 months ago #175 by jdmoore
If I change all the 10000 to 20000, all of the input appears to be read, but I then get a segmentation fault

forrtl: severe (174): SIGSEGV, segmentation fault occurred
Image PC Routine Line Source
cassandra.exe 0000000000B2D515 Unknown Unknown Unknown
cassandra.exe 0000000000B2B137 Unknown Unknown Unknown
cassandra.exe 0000000000ADDE54 Unknown Unknown Unknown
cassandra.exe 0000000000ADDC66 Unknown Unknown Unknown
cassandra.exe 0000000000A98E06 Unknown Unknown Unknown
cassandra.exe 0000000000A9F1B0 Unknown Unknown Unknown
libpthread.so.0 00007F62E3245100 Unknown Unknown Unknown
cassandra.exe 00000000006B070E fragment_growth_m 170 fragment_growth.f90
cassandra.exe 0000000000870398 make_config_ 119 make_config.f90
cassandra.exe 0000000000407723 MAIN__ 274 main.f90
cassandra.exe 0000000000402ADE Unknown Unknown Unknown
libc.so.6 00007F62E2E95B15 Unknown Unknown Unknown
cassandra.exe 00000000004029E9 Unknown Unknown Unknown

Please Log in to join the conversation.

More
7 years 9 months ago #176 by ryangmullen
Hi Josh,

I think you'll run into several problems trying to change the maximum number of atoms in a molecule. Can you explain what you are trying to model? It is likely there is another way to accomplish your objective.

Ryan

Please Log in to join the conversation.

  • jdmoore
  • jdmoore's Avatar Topic Author
  • Offline
  • Junior Member
  • Junior Member
More
7 years 9 months ago #177 by jdmoore
Hi Ryan,

Thanks for your reply. I wanted a rigid structure to have more than 10,000 atoms. I didn't catch the "per molecule" limitation. Perhaps I can split it into multiple molecules and read the pieces in separately since I don't need it to be bonded.

However, for other reasons as well, I may wish to have more than 10,000 atoms in a molecule. And in that case, I do need things to be bonded, so treating the structure as a single molecule would be ideal.

Thanks.

Josh

Please Log in to join the conversation.

More
7 years 9 months ago #180 by ryangmullen
Cassandra will not efficiently sample the internal degrees of freedom of a molecule with more than 10,000 atoms. If the molecule is rigid, then the atoms do not need to be bonded as there will not be any changes in intramolecular energy.

You can check out the example of methane adsorption in silicalite in Examples/Methane_Silicalite. We read in 7,776 atoms of silicalite (27 unit cells) but only define a 3-atom "molecule": SiO2.

Please Log in to join the conversation.

  • jdmoore
  • jdmoore's Avatar Topic Author
  • Offline
  • Junior Member
  • Junior Member
More
7 years 9 months ago #184 by jdmoore
This potentially might work for the solid, but I'm going to need to deal with > 10,000 atoms in a flexible molecule at some point. Will have to revisit this later. I was thinking I'd be able to equilibrate it much faster in MC than MD, but I didn't think about issues with acceptance.

Please Log in to join the conversation.

Time to create page: 0.123 seconds