Dear Marcus Martin and towhee users
I am working on GEMC-NVT simulation to predict the VLE of pure compound at various temperatures. The simulation had worked smoothly for a temperature below 550K. i.e. they equilibrate well. But, when I perform simulation above 550K, I get an error of "VOLNVT: move rejected, trial volume too low" during the relaxation run. I tried changing (increasing) length of the hmatrix of vapour box but it is giving the same error again with the rcut of 12 angstroms. What parameters do I change to get out of this problem?
Thank you.
Gondu.
Marcus Martin - 2013-02-09
This is a common problem when performing vapor-liquid coexistence simulations. You have simply moved into a region where one or both of your simulation boxes wants to shrink below the allowed size.
I would suggest taking the output number densities in the two boxes and using those as the input for a new simulation using an initboxtype of 'number density' and setting box_number_density to the appropriate values. This allows you to quickly generate a box with the desired number of molecules and volume of the two phases as the starting point for a new simulation.
Marcus
This happens if the simulation wants to shrink one of the box volumes below the value of "rcut", the cutoff radius for LJ interactions.
Usually this means you need to increase the size (number of molecules) of the system, although there are other possible problems. Regards,
Jeffrey J. Potoff
--------------------------------------------------------------------------------------------------------------------------------------
个人经验：
如何避免上述情况：找出临近温度点正常程序正常运行结果中的数密度，然后根据此数密度，设置该状态点每个盒子的大小及分子数目，且仍采用dimensions设置。