Gautham Dharuman

95d673_bd2a5c91b97142f2bb72fbdb15657243My research interests include modeling strongly coupled plasmas, effective quantum potentials for partially ionized and correlated plasmas, and efficient methods for force computation in molecular dynamics simulation of correlated plasmas.

My ongoing research topics:

  1. Non-equilibrium quantum evolution of  a large-scale correlated system is intractable. I am interested in development of an effective potential method employing momentum-dependent potentials for large-scale (~million to ~billion particles evolving over ~pico- to -micro-sec) molecular dynamics simulation of partially ionized plasmas such as any warm dense plasma, or specifically plasma occurring in experimental facilities such as the inertial confinement fusion (ICF), Z machine, Linac coherent light source (LCLS), Deutsches Elektronen-Synchroton (DESY) to name a few. More on momentum-dependent potentials can be found at this page.
  2. Highly efficient algorithms exist for computing forces due to short range and long range interactions in large-scale molecular dynamics simulation of correlated plasmas. But, there are many scenarios where the interaction can be medium range for which efficient algorithms haven’t been identified. I am interested in identifying efficient algorithms based on generalization of Ewald decomposition for screened Coulomb interaction and implementation using the highly efficient Particle-Particle-Particle-Mesh algorithm for fast force calculation in molecular dynamics simulation of periodic systems. For this project I developed a production-scale pure-Python code Sarkas under the guidance of my advisor Prof. Murillo. Though pure-Python, extensive use of Numpy arrays and optimization with Numba enables Sarkas to be high-performant with execution speeds comparable to compiled languages (e.g., C). Sarkas is currently undergoing improvements and will soon be made an open-source code.

My other interests include high performance computing through algorithms that employ hybrid parallelism involving many-core CPU and GPU.

Education

  • PhD in Computational Mathematics, Science and Engineering (CMSE) & Electrical and Computer Engineering (ECE), August, 2013– Present
    Michigan State University, Michigan, United States.
  • MS, Electrical and Computer Engineering, May, 2013
    Michigan State University, Michigan, United States.
  • BS, Electrical and Electronics Engineering, May, 2010
    National Institute of Technology, Warangal, India.

Publications

Atomic bound states and scattering properties using effective momentum dependent potentials
G. Dharuman, J. Verboncoeur, A. Christlieb and M. S. Murillo
Phys. Rev. E 94, 043205 (2016)

A generalized Ewald decomposition for screened Coulomb interactions
G. Dharuman, L. G. Stanton, J. N. Glosli and M. S. Murillo
J. Chem. Phys. 146, 024112 (2017).

Modeling and simulation of an ultrasensitive electron tunneling poisition/force nanosensor
Z. Fan, X. Tao, G. Dharuman, X. Li and L. Dong
RSC Adv., 6, 8297 (2016)

Internal electron tunneling enabled ultrasenstive position/force peapod sensors
X. Tao, Z. Fan, B. J. Nelson, G. Dharuman, W. Zhang, L. Dong, and X. Li
Nano Lett. 15, 7281 (2015)

An Inter-Segment Tunneling Nanoscale Force Sensor: Modeling and Simulation
G. Dharuman, Z. Fan and L. Dong
Proceedings of the 13th IEEE International Conference on Nanotechnology, Beijing, China, August 5-8, 2013