Top Page > Browsing

Parameter setting for molecular dynamics(NVT_NH) Date: 2020/04/23 12:04 Name: cpniu   <cpniu@theory.issp.ac.cn> Dear OpenMX community, When I was running molecular dynamics operation with OpenMX, I was not quite clear about the setting of several parameters. I hope you can give me some advice. Thank you very much. I need to run a molecular dynamics simulation under the condition of 3000 K and 200 GPa. The relevant parameters set are as follows: MD.type NVT_NH MD.maxIter 10000 MD.TimeStep 1.0 <MD.TempControl 5 1 3000 100 3000 1000 3000 5000 3000 10000 3000 MD.TempControl> NH.Mass.HeatBath 2016 MD.applied.pressure 200 The not quite clear parameters set are as follows: 1. scf.ElectronicTemperature This parameter is 300 K by default, should it be modified? If the atom is hot, do I have to raise the temperature of the electron? 2. scf.EigenvalueSolver I'm going to run the liquid water with 40 cores. The ordinary method is ok. Could I set it as "Band"? What is this parameter set for general? 3. NH.Mass.HeatBath I have 112 water molecules in my system. This parameter is set to 2016(112*18), is it correct? 4. MD.applied.pressure Whether this parameter is valid in the process of molecular dynamics? Is it only applicable for structural optimization? Waiting for your suggestions. Thank you. cpniu

Page: [1]

Re: Parameter setting for molecular dynamics(NVT_NH) ( No.1 ) Date: 2020/04/23 14:55 Name: Naoya Yamaguchi Hi, >1. scf.ElectronicTemperature >This parameter is 300 K by default, should it be modified? If the atom is hot, do I have to raise the temperature of the electron? This is the temperature of Fermi-Dirac distribution function and it is related to counting the number of occupied electrons. https://en.wikipedia.org/wiki/Fermi%E2%80%93Dirac_statistics >2. scf.EigenvalueSolver >I'm going to run the liquid water with 40 cores. The ordinary method is ok. Could I set it as "Band"? What is this parameter set for general? I am not familiar with molecular dynamics, but ideally, the liquid system should not be periodic, and I think that "Cluster" is enough. >3. NH.Mass.HeatBath >I have 112 water molecules in my system. This parameter is set to 2016(112*18), is it correct? No, it is an imaginary mass, not based on the atomic mass. You can refer to the original articles in the reference 31: http://www.openmx-square.org/openmx_man3.9/node226.html#NH And, for example, https://en.wikipedia.org/wiki/Nos%C3%A9%E2%80%93Hoover_thermostat >4. MD.applied.pressure >Whether this parameter is valid in the process of molecular dynamics? Is it only applicable for structural optimization? Maybe, you may use it even for molecular dynamics. Regards, Naoya Yamaguchi

Page: [1]