Simulations of High Energy Density Materials
Siddharth Dasgupta

High Energy Density Materials such as of HMX, RDX, and TATB play an important role in rocket engines, excavations, and as primary explosives in nuclear weapons.  We are collaborating with several groups at Caltech to establish an understanding of the basic mechanisms responsible for initiating and sustaining explosions in such materials. In this project we are studying:

For initial study, we are using a generic force field (DREIDING II with the exponential-6 van der Waals terms) with Charges derived from the Charge Equilibration generic method.  This should give an adequate equation of state but may not be accurate at high temperatures and pressure. Thus, we are starting to develop new force fields based on quantum chemical studies of these systems.

The starting point for the EOS is the Cold compression curve, where the minimum energy structure is calculated for each external pressure. Next, we performed molecular dynamics as a function of temperatures to obtain the volume versus pressure data at elevated temperatures.

There are two approaches,

We find that with NVE the pressure fluctuations damped out much more quickly so that 100 ps dynamics is sufficient to give statistically significant data.  However, as the melting temperature is approached (at 0 external pressure) the fluctuations become quite large requiring the use of large supercells.

The current calculations have used the Cerius2 suite of MD and analysis programs. Cerius2 has convenient analysis tools but uses the Ewald summation method for non-bonded interactions which makes it slow for large systems (> 2000 atoms/unit cell).  We are developing the analysis tools for MPSim program, which uses the much faster and more accurate Reduced Cell Multipole Method (RCMM) for the non-bond part.  MPSim is also well parallelized and runs efficiently on Intel, SGI, and other multi-node systems.

To obtain accurate specific heats and Gruniesen parameters at low temperatures (< Debye temperature), it is necessary to analyze the trajectory using the velocity autocorrelation functions. At high temperatures, reliable properties can be obtained directly form the MD trajectory.

See HE for a PowerPoint presentation.