Dr. WEN, Haohua

Introduction

Dr. WEN Haohua joined the Department of Physics and Materials Science as Prof. C. H. Woo's postdoctoral fellow, started on April 14, 2013.

Education Experience

The Hong Kong Polytechnic University, Hong Kong SAR, China	2008-2013	Ph.D., Department of Electronic and Information Engineering
Institute of Physics, CAS, Beijing, China,	2005-2008	unfinished Ph.D., Group II of Condensed Matter Theory
Wuhan University, Wuhan, Hubei, China	2001-2005	B.E., Department of Physics Science and Technology

Research Interest

• Defects and their thermodynamics in crystalline solids
• Magnetic-induced anharmonic effects in iron
• Ferromagnetism and its phase transition
• Amorphous iron and spin glass

Contact

Office:	AC1 - G1417
Email:	haohuwen@cityu.edu.hk
Tel:	00852-2784 4045

About Molecular Dynamics

• The Aim of MD

Molecular dynamics (MD) simulation is one of the atomistic simulations based on classical mechanics, such as Newton's Equation of motion. Given an interatomic potential, the evolution of system involving numerous particles can be represented as the time-series phase-space trajectory, based on which the thermodynamic quantities can be obtained according to statistical thermodynamics. Accordingly, MD treats the system consisting of interactive particles, which the interatomic potential is related to the atomic electronic configurations. In this regard, the interatomic potential used in MD simulation is generally provided by ab initio calculations, like Density Function Theory (DFT) beforehand. On the other hand, the thermodynamic quantities obtained from MD simulation can be provided as the input parameters to the empirical equations about microstructure evolution, such as phase-field model, rate theory, and so on.

• Time-Scale and Length-Scale of MD

Since the atomic vibration frequency is ~10THz, the time-step of MD simulation is ~1 femto-second, 1% of the vibrational cycle. Therefore the upper limit of time-scale in MD simulation is nano-second. On the other hand, due to the limitation of computation power, the number of atoms in MD system can reach one-million, about 10~100 nano-metre along each dimension.

• Basic principle in MD

   1. Hamiltonian and Equation of Motion
      Since I don't know how to type the mathematical formula in this editor, this part is ignored.
   2. The main stream of MD 
      2.1 Give an initial coordinates and momenta of N particles
      2.2 Solve the corresponding 6N equations of motion
      2.3 re-new the coordinates and momenta, and repeat 2.2 again and again, until users stop it.

• Flow-Chat of MD

   1. pre-initial parameters
   include: 
       1.1 Prepare for build in Lattice Structure: 
           Lattice constant; number of unit cell at each dimension; coordinate of element atoms in unit cell; data structure to store these coordinates.
       1.2 Prepare for initial velocity
           Temperature; atomic mass; thermostat parameters.
       1.3 Prepare for calculation of force
           time-step; potential table; near-neighbor list.
       1.4 Others
           For parallel computation; total running time and number of simulation step, and so on;
   2. Initialization process
       2.1 Initial thermostat
       2.2 Build in Lattice structure, and defects
       2.3 Initial velocity
       2.4 Other initialization
   3. Relaxation process
       Integrate the Eqation of motion, re-new the coordinates and momenta of particles
   4. Production process
       Continue integrating EOM, calculate the corresponding thermodynamic quantities
   5. Post-production process
       Output files; process phase-space trajectory; delete the used storage.

• Important tricks to accelerate MD simulations

   1. potential table
   2. nearest-neighbor list

• Interatomic Potential in MD

   1. potential for ideal gas: L-J potential, pair-wise potential
   2. potential for metals: EAM potential, Finnis-Sinclair potential
   3. potential for bio-molecules: angle related potential

• Thermodynamic quantities

   1. Static quantities: atomic volume, energy, pressure
   2. Dynamic quantities: time-correlation function related; heat capactity; 
   3. Entropic quantities: free energy; enthalpy; entropy; chemical potential

• How to output and draw frame

   1. Softwares for MD simulations
   2. Programmable softwares for drawing frame and creat movie
   3. Softwares for draw figures for thermodynamic quantities

Just for me

A. Important Reference

  1. Heisenberg Hamiltonian:  
       W. Heisenberg, Z. Phys. 38 (1926) 441
  2. Weiss molecular field approximation for magnetism. 
       P. Weiss, J. Phys. 6 (1970) 661.