I am currently performing free energy calculations for liquid using thermodynamic integration (TI). For these metastable states, maintaining a strictly fixed density (fixed simulation cell volume) is critical to ensure the thermodynamic path connects the correct initial and target states.
For some metastable states, maintaining a strictly fixed density (fixed simulation cell volume) is critical to ensure the thermodynamic path connects the correct initial and target states.The ProblemBased on the runtime logs, calphy appears to execute an automated equilibration step at the beginning of the workflow:
calphy.helpers INFO At count 1 mean pressure is -9.610814 with 10.202168 vol/atom
This internal equilibration process (presumably employing an NPT ensemble or pressure relaxation) induces significant volume fluctuations. For supercooled liquids, any lattice or volume relaxation deviates the system from the pre-optimized input configuration, effectively rendering the calculated Gibbs free energy invalid for the specific density of interest.
To maintain the integrity of the metastable thermodynamic integration, I need calculate TI without any initial box relaxation.
Dear developers, could this equilibration process be banned? How can it be banned?
I am currently performing free energy calculations for liquid using thermodynamic integration (TI). For these metastable states, maintaining a strictly fixed density (fixed simulation cell volume) is critical to ensure the thermodynamic path connects the correct initial and target states.
For some metastable states, maintaining a strictly fixed density (fixed simulation cell volume) is critical to ensure the thermodynamic path connects the correct initial and target states.The ProblemBased on the runtime logs, calphy appears to execute an automated equilibration step at the beginning of the workflow:
calphy.helpers INFO At count 1 mean pressure is -9.610814 with 10.202168 vol/atomThis internal equilibration process (presumably employing an NPT ensemble or pressure relaxation) induces significant volume fluctuations. For supercooled liquids, any lattice or volume relaxation deviates the system from the pre-optimized input configuration, effectively rendering the calculated Gibbs free energy invalid for the specific density of interest.
To maintain the integrity of the metastable thermodynamic integration, I need calculate TI without any initial box relaxation.
Dear developers, could this equilibration process be banned? How can it be banned?