A common moment of uncertainty in molecular dynamics comes after temperature has settled: the system is in NPT equilibration, pressure is fluctuating, and it is not obvious whether the simulation is ready to move on. The useful question is not whether every pressure value is steady. It is whether the system’s density is settling into a stable regime.
NPT keeps the number of particles, pressure, and temperature constant in the ensemble sense, while allowing the system volume to change. That volume adjustment is why this stage follows NVT equilibration: after temperature has stabilized, NPT helps the density approach a suitable value. Pressure can fluctuate widely during equilibration, so a jagged pressure trace by itself is not proof that something has gone wrong.
Look for a density plateau, not a perfectly flat line
Use the density plot to assess progress over time. The key signal is whether density’s running average reaches a plateau and the values remain reasonably stable around it. Individual fluctuations are expected; the goal is not a perfectly constant instantaneous density. The relevant target depends on the system and model, so interpret the result in context rather than treating one example value as universal.
For instance, the tutorial’s example reaches a density of about 1030 kg/m3. It notes that this is close to the experimental value of 1000 kg/m3, while the expected density for the SPC/E model is about 1008 kg/m3. These numbers illustrate why model-specific expectations matter: compare your result with an appropriate reference for your system, not simply with a number from another setup.
A practical decision: continue or proceed?
- Density is still trending: allow more NPT equilibration. A continuing drift suggests the system may not yet have reached a stable density regime.
- Density fluctuates around a steady level: assess whether that level is appropriate for the system and stable enough for your next step.
- Pressure varies substantially: remember that pressure can fluctuate widely. Read its plot alongside density rather than using pressure’s instantaneous values as the sole pass/fail test.
There is no single run duration that guarantees equilibration for every model. The tutorial suggests that 100 ps is typically a reasonable starting point, while emphasizing that the required time depends on the size and contents of the system. If density has not stabilized in the allotted time, run NPT again using the output from the previous NPT step as the input. This lets equilibration continue from the current results rather than requiring you to restart from an earlier stage.
Keep the temperature setup consistent
When moving from NVT to NPT, reuse the temperature-coupling choices unless you have a clear reason to change them. Check that the NPT temperature-coupling parameters match those used during NVT; if velocity generation is enabled, check its temperature as well. Pressure coupling is a separate set of choices. The tutorial describes exponential-relaxation pressure coupling with a stochastic term (c-rescale) and a time constant around 5 ps as a typical option for many projects—not a universal setting for every system.
Continue from the right input
In GROMACS Wizard, the NPT step can take a GRO file or a batch project from an earlier relevant step, including a previous NPT run. If you are continuing from the preceding step, the auto-fill button can populate the input path from the previous successful run. After the job finishes, inspect the plots and results before deciding whether to proceed to production molecular dynamics. The plots are generated when the job finishes and its results are loaded.
In short: judge equilibration by the density’s behavior over time, treat pressure fluctuations with context, preserve consistent temperature-coupling choices, and extend the run from the latest NPT results when needed. For the full workflow and interface details, see the NPT Equilibration tutorial.
Note: SAMSON and all SAMSON Extensions are free for non-commercial use. Get SAMSON.