Slow compression in HPMC simulations.

[joaander]

Description

Add a method that slowly compresses in HPMC simulations.

Proposed solution

Either implement SlowCompress or add optional arguments to QuickCompress. In either case, apply the QuickCompress algorithm to a variable box size.

Additional context

QuickCompress works extremely well, but always quenches to the chosen box size. Some simulation workflows require slow annealing. MD simulations can use BoxResize. HPMC simulations need an alternate method that accounts for particle overlaps.

[joaander]

@tcmoore3 you asked about this once. Would you be interested in implementing it?

[tcmoore3]

I am interested, but since I don't have a research need for this at the moment I cannot commit to implementing it.

[tcmoore3]

Now that I'm thinking about it, it should be fairly straightforward. Unless I'm missing something, shouldn't this consist of adding a variant to UpdaterQuickCompress and using that variant to compute the new box in UpdaterQuickCompress::getNewBox by modifying the minimum scale factor?

[joaander]

Yes, we can replace target_box with a box that varies over time. Work on this should wait pending the outcome of #1640 where I am suggesting a new BoxVariant API.

[joaander]

Now that #1685 is merged, we make target_box a BoxVariant in QuickCompress. @tcmoore3 do you plan to implement this? I would like to put it in the next release.

[tcmoore3]

I'm not sure why GitHub didn't automatically close this when #1736 was merged, but #1736 resolves this issue.

[joaander]

I'm not sure why GitHub didn't automatically close this when #1736 was merged, but #1736 resolves this issue.

GitHub automatically closes issues only when a pull request is merged into the default branch (trunk-patch).