Doug Franz

Ewald force should include k-space (reciprocal space) contribution http://www.pages.drexel.edu/~cfa22/msim/node49.html

name md_NVE_example mode md ! mc or md checkpoints_option off input_atoms input.pdb ! PDB ONLY potential_form lj com_option off ! center of mass carbasis 25 25 25 90 90 90...

ATOM 1 HT h2o M 1 -1.752770 7.859250 -4.727420 default 0.4170 ATOM 2 OT h2o M 1 -1.871140 8.575620 -4.113410 default -0.8340 ATOM 3 HT h2o M 1 -1.746980 9.388710...

quaternion dynamics tests lj, no rotations, conserves in nve. [file.zip](https://github.com/khavernathy/mcmd/files/1249615/file.zip)

LJES issue fixed via ae0614c5f35cc35d44c5007e30d3d08d8726fec3 Actually, 08001392d100a583ab9ac84262ee32d89276032d k-space force needed to be doubled to compensate for half Ewald sphere Now, need to add LJES routine to GPU CUDA ...done; matches...

theta should maybe define angular position w/respect to ORIGINAL position; not previous. so theta is recursive by integration function but the actual rotation should be performed on original relative coordinates...

or, U_rotation = torque dot theta , add this to normal potential?

intertia computable via http://farside.ph.utexas.edu/teaching/336k/Newtonhtml/node64.html it seems only Ixx, Iyy, and Izz are needed, from multiple books. AMD book has example for quaternion code

maybe just remove the additive property for rotational dynamics? This seems intuitively correct. i.e. theta = angvel x dt + 0.5 x angacc x dt^2 not +=

heat capacity already done.