DPD¶

class hoomd.md.pair.DPD(nlist, kT, default_r_cut=None)¶

Bases: Pair

Dissipative Particle Dynamics.

Parameters:

nlist (hoomd.md.nlist.NeighborList) – Neighbor list
kT (hoomd.variant or float) – Temperature of thermostat $[\mathrm{energy}]$ .
default_r_cut (float) – Default cutoff radius $[\mathrm{length}]$ .

DPD computes the DPD pair force on every particle in the simulation state. DPD includes a an interaction potential, pairwise drag force, and pairwise random force. See Groot and Warren 1997:

F = F_{\mathrm{C}}(r) + F_{\mathrm{R,ij}}(r_{ij}) + F_{\mathrm{D,ij}}(v_{ij})

where

\begin{split} F_{\mathrm{C}}(r) &= A \cdot w(r_{ij}), \\ F_{\mathrm{R, ij}}(r_{ij}) &= - \theta_{ij}\sqrt{3} \sqrt{\frac{2k_b\gamma T}{\Delta t}}\cdot w(r_{ij}), \\ F_{\mathrm{D, ij}}(r_{ij}) &= - \gamma w^2(r_{ij})\left( \hat r_{ij} \circ v_{ij} \right), \\ w(r_{ij}) &= \begin{cases} \left( 1 - r/r_{\mathrm{cut}} \right) & r < r_{\mathrm{cut}} \\ 0 & r \ge r_{\mathrm{cut}} \\ \end{cases}, \end{split}

$\hat r_{ij}$ is a normalized vector from particle i to particle j, $v_{ij} = v_i - v_j$ , and $\theta_{ij}$ is a uniformly distributed random number in the range $[-1, 1]$ .

C. L. Phillips et. al. 2011 describes the DPD implementation details. Cite it if you utilize the DPD functionality in your work.

DPD does not implement any energy shift / smoothing modes due to the function of the force.

To use the DPD thermostat, apply the hoomd.md.methods.ConstantVolume or hoomd.md.methods.ConstantPressure integration method without thermostats along with DPD forces. Use of the DPD thermostat pair force with other integrators will result in nonphysical behavior. To use DPD with a different conservative potential than $F_C$ , set A to zero and define the conservative pair force separately.

Example:

nl = nlist.Cell()
dpd = pair.DPD(nlist=nl, kT=1.0, default_r_cut=1.0)
dpd.params[('A', 'A')] = dict(A=25.0, gamma=4.5)
dpd.params[('A', 'B')] = dict(A=40.0, gamma=4.5)
dpd.params[('B', 'B')] = dict(A=25.0, gamma=4.5)
dpd.params[(['A', 'B'], ['C', 'D'])] = dict(A=40.0, gamma=4.5)

Members inherited from AutotunedObject:

property kernel_parameters¶: Kernel parameters. Read more...

property is_tuning_complete¶: Check if kernel parameter tuning is complete. Read more...

tune_kernel_parameters()¶: Start tuning kernel parameters. Read more...

Members inherited from Force:

additional_energy¶: Additional energy term. Read more...

additional_virial¶: Additional virial tensor term $W_\mathrm{additional}$ . Read more...

cpu_local_force_arrays¶: Local force arrays on the CPU. Read more...

energies¶: Energy contribution $U_i$ from each particle. Read more...

energy¶: The potential energy $U$ of the system from this force. Read more...

forces¶: The force $\vec{F}_i$ applied to each particle. Read more...

gpu_local_force_arrays¶: Local force arrays on the GPU. Read more...

torques¶: The torque $\vec{\tau}_i$ applied to each particle. Read more...

virials¶: Virial tensor contribution $W_i$ from each particle. Read more...

Members inherited from Pair:

nlist¶: Neighbor list used to compute the pair force. Read more...

mode¶: Energy smoothing/cutoff mode. Read more...

r_cut¶: Cuttoff radius beyond which the energy and force are 0. Read more...

r_on¶: Radius at which the XPLOR smoothing function starts. Read more...

compute_energy()¶: Compute the energy between two sets of particles. Read more...

Members defined in DPD:

params¶

The force parameters. The dictionary has the following keys:

A (float, required) - $A$ $[\mathrm{force}]$
gamma (float, required) - $\gamma$ $[\mathrm{mass} \cdot \mathrm{time}^{-1}]$

Type: TypeParameter [tuple [particle_type, particle_type], dict]