We are pleased to announce the availability of OpenMD version 2.4. In addition to bug fixes and performance improvements, there are new features that have been added to the code since the 2.3 release:

New Features:

  • Added Mie potentials for non-bonded interactions
  • We now have an initial implementation of fluctuating-density EAM models
  • A new omd2omd utility replaces the md2md script. omd2omd is able to do box repeats and translations without splitting up individual molecules.
  • Added a normalization option to vcorr2spectrum
  • All of the analysis programs (StaticProps, DynamicProps, SequentialProps) now provide revision and meta-data in their output files
  • Added arithmetic mixing rule option to GB (Gay-Berne) potentials.
  • Added support for selecting Bonds, Bends, Torsions, Inversions, and Molecules in parallel environments.
  • Added “potentialSelection” keyword for reporting Potential Energy for a selection in a stat file.
  • Added an option to use force field files with or without the scaled bond and bend force constants (e.g. Amber and Charmm use k (r-r0)2 while TraPPE and OpenMD default to k/2 * (r-r0)2).
  • New modules for StaticProps:
    • Distance-dependent Kirkwood factors (dipolar and quadrupolar)
    • Tetrahedrality matrices for hydrogen bonds
    • Generalized Coordination Numbers (GCN)
    • Added PotDiff module
  • New modules for StaticProps:
    • Much (10-1000x) faster for many correlation functions
    • HydrogenBond Jump time calculator
    • RCorrFuncZ: does root mean squared displacements binned by z coordinates

Build Fixes:

  • Build is now git-aware
  • Many fixes to allow building with different versions of the qhull library
  • Updated force field files and samples

Bug fixes

  • Bug-fix in pack2md (contributed by Hythem Sidky)
  • Fixes in InteractionManager to catch rare cases for explicit non-bonded interactions could erase themselves
  • Simple change to allow printing information from final frame in dump and stat files
  • Fixed typos in how Stats module displayed column titles
  • Fixed some long-standing issues with bin counting and fixed normalization of g(r,cos(theta))
  • Fixed a parallel constraint bug in Rattle/Shake
  • Fixed correlation functions with changing populations in selections
  • Fix for directional atom types and Gay Berne potential (contributed by Hythem Sidky)
  • P2 order parameter is now calculated from row for directional atom types. This is consistent with the Gay-Berne module.
  • Atom spacing in constructed polyhedra in icosahedralBuilder was off by 1/sqrt(2).
  • A number of fixes to allow fluctuating charge potentials to use the multipolar Ewald sum.

Go grab it at the Download page!