I put together a walkthrough on equilibrating polymer system to a realistic density, something I found to be one of the fiddlier parts of setting up a polymer MD run.
I basically started from a low-density box (to avoid overlaps), then drive it to the target state. I compare two ways:
Method 1 — Slow compression: hot NVT at 600 K to relax out of bad local minima → cool to 300 K → NPT with a staged pressure ramp (2% → 60% → 100% of P_max = 50,000 bar), short NVT relaxations between compressions → decompress to 1 bar with a long final NPT. The staged ramp matters — going straight to P_max tends to blow up the system.
Method 2 — Slow cooling (from a high-throughput Tg paper): equilibrate, then step-cool from high T under NPT down toward 300 K. Simpler, but a much longer trajectory.
I ran both on PMMA (~12k atoms, 3 samples each). Both landed ~1.03–1.07 g/cm³ at 300 K — same ballpark, below the experimental ~1.18 (I used openff sage force field).
Full video walkthrough: [https://www.youtube.com/watch?v=v6Z4Pz7HJ0U]
Curious how others here handle polymer equilibration. Do you have anything more efficient than these two? And how do you decide “equilibrated enough” beyond density plateauing?