Release Notes

Changes, bug fixes or other relevant information for users and developers of Cassandra

Cassandra V1.2


Fixed: error in Rotate when no rotatable molecules are in the box

Fixed: error in Rigid_Dihedral_Change when some species don't have dihedrals


Fixed: gfortran bug in read_checkpoint

Fixed: over/underwhelming P_bias for molecules with >10 fragments


Fixed: gfortran_openMP bug in Ewald energy calc

Fixed: bug in gemc_particle_transfer effecting GEMC_NVT and GEMC_NPT ensembles

added cvf improper to participation.f90, modify printed energy in ring fragments file

removed vdw_type NONE from examples

Fixed: bug when reading chemical potential for sims with non-insertable species


Fixed: computed pressure was a factor of 166.054 too high

Fixed: error when using the half-box cuttoff

added error checking for box number in section # Property_Info

removed Update_Reservoir code


Fixed: gemc_particle_transfer, error when selecting a molecule when some species are not insertable


added SiO2.pdb used in Examples/GCMC/Methane_Silicalite


Fixed: volume_change for GEMC-NPT, error in allocating sin_sum/cos_sum ewald arrays when smaller box changes size

Fixed: pressure calculation for atoms wtih vdW type 'NONE' in section "# Atom_Info"

Fixed: Loop over Mie potential vdw_param3_table by atom_type


Initial release. New features and changes include:

  • Damped shifted force method for computing long-range electrostatic energies (faster than ewald)
  • Fixed errors in gfortran-compiled Cassandra that occured if:
    • attempting volume change moves with Ewald summation
    • generating ring fragment libraries
  • Added tail corrections for Mie potential
  • Standardized format of input and output files, including:
    • Each word of a section header begins with an initial capital, e.g. # Run_Name
    • Keywords and arguments are all lowercase, e.g. cubic, units steps, coul ewald

For those familiar with previous versions of Cassandra, some of the sections and keywords that have changed are:

  • In Section # Simulation_Length_Info, the keyword MCsteps has been replaced with run. The keyword units takes the argument sweeps in addition to steps and minutes. The number of steps per sweep can be defined using the keyword steps_per_sweep, e.g. steps_per_sweep 100 defines a sweep as 100 steps.
  • Section Simulation_Length_Info ends with a blank line (like other sections) instead of the section footer # Done_Simulation_Length_Info.
  • In Section # Start_Type, the keyword read_old has been replaced with read_config and new keyword add_to_config has been added. Along with make_config, read_config and add_to_config take arguments on the same line as the keyword. For example, the command to fill a box with 100 molecules of species 1 has changed from



make_config 100

This format allows you to mix and match start types for different boxes. For example, you can read in a equilibrated configuration of liquid water (species 2), but make the corresponding equimolar gas phase configuration:

read_config 0 300
make_config 100 100

  • For grand canonical and Gibbs ensemble simulations, the # Prob_Insertion or # Prob_Swap subsections have changed. In V1.1, to insert molecules of a species 1 but not species 2 requires the following commands

# Prob_Insertion
insertion method

insertion method

But in V1.2,

# Prob_Insertion
cbmc none

This format now parallels the format of the # Prob_Translation and # Prob_Rotation subsections in which parameters for each species are listed on the same line.

  • For Gibbs ensemble simulations, the temperature and pressure for each box in sections # Temperature_Info and # Pressure_Info must be given on consecutive lines. For example, in V1.1 the temperature would be specified as 

# Temperature_Info
298 298

but in V1.2, this must be replaced with

# Temperature_Info

This format is now consistent with the # Start_Type section and # Prob_Translation and # Prob_Rotation subsections in which properties for each box are listed on consecutive lines.

Cassandra V1.1


Fixed when reading coordinates from .cml

Fixed when reading .pdb with non-consecutive numbered atoms


Make verbose options compatible with gfortran compiler.

Fixed warnings when accessing mie matrices.

Fixed compatibility of with new code names.


Initial release