A molecular dynamics run can finish without answering the question you actually care about: is the system ready for the next stage? During NVT equilibration, the goal is to bring the system to its target temperature while keeping the number of particles and volume constant. Moving on too early can leave the temperature still settling; waiting for a perfectly flat temperature trace, on the other hand, is not the goal either. Small fluctuations around the target are normal.
The useful question is whether the temperature’s running behavior has reached a stable plateau. That gives you a practical way to decide whether to continue equilibration or proceed to NPT, where density is stabilized.
Read the trend, not just the last value
A temperature trace naturally varies over simulation time. A single point near the target does not show that the system has settled, just as a few small deviations do not automatically mean something is wrong. Look at the overall behavior: does the running average approach the target and then remain around it, without a continuing trend toward or away from that value?
In the example below, the plot shows temperature stabilizing around 300 K, the default reference temperature in the example’s advanced parameters. Your target may differ, so compare the plot with the temperature you intended to use.
A stable temperature plot is a practical checkpoint for continuing your workflow; it is not a reason to expect every fluctuation to disappear. If the temperature has not yet stabilized, the documented approach is to run additional NVT equilibration using the results of the previous NVT run as input.
Check the temperature settings together
When reviewing the run setup, compare the reference temperature used for temperature coupling with the temperature specified for velocity generation. The tutorial notes that these should correspond. For many projects, it gives velocity rescaling with a stochastic term (v-rescale) and a time constant of about 1 ps as a common starting point. These are typical values, not a substitute for checking whether the settings suit your system and workflow.
If you are coupling separate groups, such as Protein and non-Protein, make sure the associated time constants and reference temperatures match the groups listed. For a general system, the default System coupling can be restored with the reset control. Position-restraint options are filled automatically based on the input system.
A practical NVT check in GROMACS Wizard
- Start from the right input. Use the GRO file or batch project produced by minimization, or the results of a previous equilibration run. The auto-fill button can populate the input path from the previous successful step.
- Review the run length and parameters. The tutorial notes that 50–100 ps is typically sufficient, but the time required depends on system size and contents. Treat the temperature plot, rather than a fixed duration alone, as the check for whether the target has stabilized.
- Run locally, in the cloud, or generate inputs. GROMACS Wizard offers all three options. Cloud computations require computing credits; local jobs can be followed from the Local jobs list.
- Inspect the plot after the job finishes. Confirm that the temperature is fluctuating around the intended target and that the running behavior looks stable enough to continue. If not, start another NVT run from the current NVT results.
- Proceed to NPT when appropriate. Once the temperature is stable enough for your workflow, NPT equilibration is the next step for stabilizing density.
This check helps avoid two common interpretation mistakes: treating every small fluctuation as a failure, and treating a brief arrival at the target as proof that equilibration is complete. Focus on the trend, verify that the temperature settings agree, and continue NVT if the plot has not settled.
For the complete workflow and interface details, see the NVT Equilibration tutorial.
Note: SAMSON and all SAMSON Extensions are free for non-commercial use. Get SAMSON.