Hi.
I am interested in getting pure component vapor pressures and am thinking of using GEMC. I have read the manual about CO2 and checked the Non-Basic/Gibbs_CO2 example and have a few questions.
1. What output property should I use to determine if I am using enough cycles and molecules?
2. The adsorbate/adsorbate energy shows a lot of variance. How much fluctuation is generally acceptable to ensure the simulation is in equilibrium?
3. What is a good rule of thumb in changing input parameters to get stable results?
(ex. increase no. of cycles > increase molecules > increase swaps)
Thanks in advance!
1) As many cycles and molecules as you need. The more the better, but the slower the simulations. That's always a trade-off.
In equilibrium, the macroscopic properties no longer change. The amount of cycles needed to reach equilibrium depends on the system.
To limit finite size effects you need a large system. In practice you can simulate various system sizes as see when your properties converge to a plateau value as a function of system size.
2) in a NVT or muVT or Gibbs NVT simulation the energy fluctuates. This energy fluctuation is related to a physical property called the heat capacity.
3) There are no rules of thumbs, but rather it will take experience to get a feel for this.
Thank you for your explanation.
I think I should take a step back and start from the top.
I want to calculate the vapor pressure of certain molecules using MC.
1. Is GEMC the most effective way to meet my objective? If not, what is a better way to obtain vapor pressure using RASPA?
2. After further studying the Non-Basic/Gibbs_CO2 example output, I found that the pressure has not been calculated. I am assuming that if I can get the pressure for both vapor and liquid systems, this will be the saturation pressure at the given temperature and calculated system volumes for each system. Please let me know if I am misunderstanding something.
3. How do I get RASPA to compute pressure for GEMC?
4. If I am simulating a molecule, where the critical properties are not known, can I use RASPA to estimate the critical properties? (The previous studies do not show data points close to the critical point.)
See the many articles of Martin and Siepmann on this topic.
To compute pressure:
ComputeMolecularPressure yes
Hi.
when I calculate the adsorption of methanol, I get that its excess adsorption capacity is negative. I know that the set pressure exceeds its saturated vapor pressure. However, I want to use Raspa to calculate the adsorption of liquid mixtures. According to the literature, it is necessary to set the configuration bias Monte Carlo, but I don't know how to set the parameters. How should I set the parameters. Thank you.