Revamp Ellipsoid Class, use fixed bonds.

[chrisjonesBSU]

The ellipsoid chain features currently in flowerMD are quite "experimental", and after doing a deeper dive with @StephMcCallum there are clearly some issues with the approach we used initially (rigid bodies + ghost particles) that make the model unstable and difficult to get up and running.

After chatting a bit, we think it would be worth revamping these classes in flowerMD to start with a more simple approach. The model here now uses fixed bond lengths between the tips of 2 bonded ellipsoids. A single ghost particle is used between ellipsoids that will handle both fixed bonds, and enforcing a 2-body angle potential (ellipsoid_center - ghost at tip- ellipsoid_center). Rigid bodies are no longer used in this approach.

Here is a minimal example:

from flowermd.library import EllipsoidChain, EllipsoidForcefield
from flowermd import Pack, Simulation
from flowermd.utils import set_bond_constraints

chain = EllipsoidChain(lengths=30, num_mols=1, lpar=1, bead_mass=1)
system = Pack(molecules=chain, density=0.01, packing_expand_factor=25)
constrain_snap, d_constraint = set_bond_constraints(system.hoomd_snapshot, constrain_value=1.0, bond_type="_C-_H")
forces = EllipsoidForcefield(epsilon=1.0, lpar=1.0, lperp=0.5, r_cut=3.0, angle_k=10, angle_theta0=2.2)

sim = Simulation(initial_state=constrain_snap, forcefield=forces.hoomd_forces, constraint=d_constraint, gsd_write_freq=5000)
sim.run_NVT(kT=1.0, n_steps=500000, tau_kt=100*sim.dt)

[codecov]

Codecov Report

Attention: Patch coverage is 97.95918% with 1 line in your changes missing coverage. Please review.

Project coverage is 92.08%. Comparing base (a600b73) to head (50ec78d).
Report is 58 commits behind head on main.

Files with missing lines	Patch %	Lines
flowermd/base/simulation.py	90.00%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #191      +/-   ##
==========================================
- Coverage   94.77%   92.08%   -2.69%     
==========================================
  Files          26       26              
  Lines        1970     2048      +78     
==========================================
+ Hits         1867     1886      +19     
- Misses        103      162      +59     

Files with missing lines	Coverage Δ
flowermd/library/forcefields.py	95.45% <100.00%> (-0.08%)	⬇️
flowermd/library/polymers.py	100.00% <100.00%> (ø)
flowermd/utils/__init__.py	100.00% <100.00%> (ø)
flowermd/utils/constraints.py	32.00% <100.00%> (ø)
flowermd/base/simulation.py	90.79% <90.00%> (-1.23%)	⬇️

[StephMcCallum]

On my initial run with the minimum working example, I am seeing that the _H-_C-_H are bending. Do we want these to stay rigid for the body of the ellipsoid?

There also appears to be one _H missing on the beginning of the chain (or maybe an extra one on the end).

[chrisjonesBSU]

Yeah I'm noticing the same thing..so we'll probably have to use rigid bodies still.