64 |
|
that is easy to learn. |
65 |
|
|
66 |
|
\tableofcontents |
67 |
< |
%\listoffigures |
68 |
< |
%\listoftables |
67 |
> |
\listoffigures |
68 |
> |
\listoftables |
69 |
|
|
70 |
|
\mainmatter |
71 |
|
|
497 |
|
{\tt minimizer} & string & Chooses a minimizer & Possible minimizers |
498 |
|
are SD and CG. Either {\tt ensemble} or {\tt minimizer} must be specified. \\ |
499 |
|
{\tt ensemble} & string & Sets the ensemble. & Possible ensembles are |
500 |
< |
NVE, NVT, NPTi, NPAT, NPTf, NPTxyz, and LD. Either {\tt ensemble} |
500 |
> |
NVE, NVT, NPTi, NPAT, NPTf, NPTxyz, LD and LHull. Either {\tt ensemble} |
501 |
|
or {\tt minimizer} must be specified. \\ |
502 |
|
{\tt dt} & fs & Sets the time step. & Selection of {\tt dt} should be |
503 |
|
small enough to sample the fastest motion of the simulation. ({\tt |
1906 |
|
& (with separate barostats on each box dimension) & \\ |
1907 |
|
LD & Langevin Dynamics & {\tt ensemble = LD;} \\ |
1908 |
|
& (approximates the effects of an implicit solvent) & \\ |
1909 |
+ |
LangevinHull & Non-periodic Langevin Dynamics & {\tt ensemble = LHull;} \\ |
1910 |
+ |
& (Langevin Dynamics for molecules on convex hull;\\ |
1911 |
+ |
& Newtonian for interior molecules) & \\ |
1912 |
|
\end{tabular} |
1913 |
|
\end{center} |
1914 |
|
|
2394 |
|
in the body-fixed frame) and ${\bf V}$ is a generalized velocity, |
2395 |
|
${\bf V} = |
2396 |
|
\left\{{\bf v},{\bf \omega}\right\}$. The right side of |
2397 |
< |
Eq.~\ref{LDGeneralizedForm} consists of three generalized forces: a |
2397 |
> |
Eq. \ref{LDGeneralizedForm} consists of three generalized forces: a |
2398 |
|
system force (${\bf F}_{s}$), a frictional or dissipative force (${\bf |
2399 |
|
F}_{f}$) and a stochastic force (${\bf F}_{r}$). While the evolution |
2400 |
|
of the system in Newtonian mechanics is typically done in the lab |
2616 |
|
\endhead |
2617 |
|
\hline |
2618 |
|
\endfoot |
2619 |
< |
{\tt viscosity} & centipoise & Sets the value of viscosity of the implicit |
2619 |
> |
{\tt viscosity} & poise & Sets the value of viscosity of the implicit |
2620 |
|
solvent \\ |
2621 |
|
{\tt targetTemp} & K & Sets the target temperature of the system. |
2622 |
|
This parameter must be specified to use Langevin dynamics. \\ |
2623 |
|
{\tt HydroPropFile} & string & Specifies the name of the resistance |
2624 |
|
tensor (usually a {\tt .diff} file) which is precalculated by {\tt |
2625 |
< |
Hydro}. This keyworkd is not necessary if the simulation contains only |
2625 |
> |
Hydro}. This keyword is not necessary if the simulation contains only |
2626 |
|
simple bodies (spheres and ellipsoids). \\ |
2627 |
|
{\tt beadSize} & $\mbox{\AA}$ & Sets the diameter of the beads to use |
2628 |
|
when the {\tt RoughShell} model is used to approximate the resistance |
2629 |
|
tensor. |
2630 |
|
\label{table:ldParameters} |
2631 |
|
\end{longtable} |
2632 |
+ |
|
2633 |
+ |
\section{Langevin Hull Dynamics (LHull)} |
2634 |
+ |
|
2635 |
+ |
The Langevin Hull uses an external bath at a fixed constant pressure |
2636 |
+ |
($P$) and temperature ($T$) with an effective solvent viscosity |
2637 |
+ |
($\eta$). This bath interacts only with the objects on the exterior |
2638 |
+ |
hull of the system. Defining the hull of the atoms in a simulation is |
2639 |
+ |
done in a manner similar to the approach of Kohanoff, Caro and |
2640 |
+ |
Finnis.\cite{Kohanoff:2005qm} That is, any instantaneous configuration |
2641 |
+ |
of the atoms in the system is considered as a point cloud in three |
2642 |
+ |
dimensional space. Delaunay triangulation is used to find all facets |
2643 |
+ |
between coplanar |
2644 |
+ |
neighbors.\cite{delaunay,springerlink:10.1007/BF00977785} In highly |
2645 |
+ |
symmetric point clouds, facets can contain many atoms, but in all but |
2646 |
+ |
the most symmetric of cases, the facets are simple triangles in |
2647 |
+ |
3-space which contain exactly three atoms. |
2648 |
|
|
2649 |
+ |
The convex hull is the set of facets that have {\it no concave |
2650 |
+ |
corners} at an atomic site.\cite{Barber96,EDELSBRUNNER:1994oq} This |
2651 |
+ |
eliminates all facets on the interior of the point cloud, leaving only |
2652 |
+ |
those exposed to the bath. Sites on the convex hull are dynamic; as |
2653 |
+ |
molecules re-enter the cluster, all interactions between atoms on that |
2654 |
+ |
molecule and the external bath are removed. Since the edge is |
2655 |
+ |
determined dynamically as the simulation progresses, no {\it a priori} |
2656 |
+ |
geometry is defined. The pressure and temperature bath interacts only |
2657 |
+ |
with the atoms on the edge and not with atoms interior to the |
2658 |
+ |
simulation. |
2659 |
+ |
|
2660 |
+ |
Atomic sites in the interior of the simulation move under standard |
2661 |
+ |
Newtonian dynamics, |
2662 |
+ |
\begin{equation} |
2663 |
+ |
m_i \dot{\mathbf v}_i(t)=-{\mathbf \nabla}_i U, |
2664 |
+ |
\label{eq:Newton} |
2665 |
+ |
\end{equation} |
2666 |
+ |
where $m_i$ is the mass of site $i$, ${\mathbf v}_i(t)$ is the |
2667 |
+ |
instantaneous velocity of site $i$ at time $t$, and $U$ is the total |
2668 |
+ |
potential energy. For atoms on the exterior of the cluster |
2669 |
+ |
(i.e. those that occupy one of the vertices of the convex hull), the |
2670 |
+ |
equation of motion is modified with an external force, ${\mathbf |
2671 |
+ |
F}_i^{\mathrm ext}$: |
2672 |
+ |
\begin{equation} |
2673 |
+ |
m_i \dot{\mathbf v}_i(t)=-{\mathbf \nabla}_i U + {\mathbf F}_i^{\mathrm ext}. |
2674 |
+ |
\end{equation} |
2675 |
+ |
|
2676 |
+ |
The external bath interacts indirectly with the atomic sites through |
2677 |
+ |
the intermediary of the hull facets. Since each vertex (or atom) |
2678 |
+ |
provides one corner of a triangular facet, the force on the facets are |
2679 |
+ |
divided equally to each vertex. However, each vertex can participate |
2680 |
+ |
in multiple facets, so the resultant force is a sum over all facets |
2681 |
+ |
$f$ containing vertex $i$: |
2682 |
+ |
\begin{equation} |
2683 |
+ |
{\mathbf F}_{i}^{\mathrm ext} = \sum_{\begin{array}{c}\mathrm{facets\ |
2684 |
+ |
} f \\ \mathrm{containing\ } i\end{array}} \frac{1}{3}\ {\mathbf |
2685 |
+ |
F}_f^{\mathrm ext} |
2686 |
+ |
\end{equation} |
2687 |
+ |
|
2688 |
+ |
The external pressure bath applies a force to the facets of the convex |
2689 |
+ |
hull in direct proportion to the area of the facet, while the thermal |
2690 |
+ |
coupling depends on the solvent temperature, viscosity and the size |
2691 |
+ |
and shape of each facet. The thermal interactions are expressed as a |
2692 |
+ |
standard Langevin description of the forces, |
2693 |
+ |
\begin{equation} |
2694 |
+ |
\begin{array}{rclclcl} |
2695 |
+ |
{\mathbf F}_f^{\text{ext}} & = & \text{external pressure} & + & \text{drag force} & + & \text{random force} \\ |
2696 |
+ |
& = & -\hat{n}_f P A_f & - & \Xi_f(t) {\mathbf v}_f(t) & + & {\mathbf R}_f(t) |
2697 |
+ |
\end{array} |
2698 |
+ |
\end{equation} |
2699 |
+ |
Here, $A_f$ and $\hat{n}_f$ are the area and (outward-facing) normal |
2700 |
+ |
vectors for facet $f$, respectively. ${\mathbf v}_f(t)$ is the |
2701 |
+ |
velocity of the facet centroid, |
2702 |
+ |
\begin{equation} |
2703 |
+ |
{\mathbf v}_f(t) = \frac{1}{3} \sum_{i=1}^{3} {\mathbf v}_i, |
2704 |
+ |
\end{equation} |
2705 |
+ |
and $\Xi_f(t)$ is an approximate ($3 \times 3$) resistance tensor that |
2706 |
+ |
depends on the geometry and surface area of facet $f$ and the |
2707 |
+ |
viscosity of the bath. The resistance tensor is related to the |
2708 |
+ |
fluctuations of the random force, $\mathbf{R}(t)$, by the |
2709 |
+ |
fluctuation-dissipation theorem, |
2710 |
+ |
\begin{eqnarray} |
2711 |
+ |
\left< {\mathbf R}_f(t) \right> & = & 0 \\ |
2712 |
+ |
\left<{\mathbf R}_f(t) {\mathbf R}_f^T(t^\prime)\right> & = & 2 k_B T\ |
2713 |
+ |
\Xi_f(t)\delta(t-t^\prime). |
2714 |
+ |
\label{eq:randomForce} |
2715 |
+ |
\end{eqnarray} |
2716 |
+ |
|
2717 |
+ |
Once the resistance tensor is known for a given facet, a stochastic |
2718 |
+ |
vector that has the properties in Eq. (\ref{eq:randomForce}) can be |
2719 |
+ |
calculated efficiently by carrying out a Cholesky decomposition to |
2720 |
+ |
obtain the square root matrix of the resistance tensor, |
2721 |
+ |
\begin{equation} |
2722 |
+ |
\Xi_f = {\bf S} {\bf S}^{T}, |
2723 |
+ |
\label{eq:Cholesky} |
2724 |
+ |
\end{equation} |
2725 |
+ |
where ${\bf S}$ is a lower triangular matrix.\cite{Schlick2002} A |
2726 |
+ |
vector with the statistics required for the random force can then be |
2727 |
+ |
obtained by multiplying ${\bf S}$ onto a random 3-vector ${\bf Z}$ which |
2728 |
+ |
has elements chosen from a Gaussian distribution, such that: |
2729 |
+ |
\begin{equation} |
2730 |
+ |
\langle {\bf Z}_i \rangle = 0, \hspace{1in} \langle {\bf Z}_i \cdot |
2731 |
+ |
{\bf Z}_j \rangle = \frac{2 k_B T}{\delta t} \delta_{ij}, |
2732 |
+ |
\end{equation} |
2733 |
+ |
where $\delta t$ is the timestep in use during the simulation. The |
2734 |
+ |
random force, ${\bf R}_{f} = {\bf S} {\bf Z}$, can be shown to |
2735 |
+ |
have the correct properties required by Eq. (\ref{eq:randomForce}). |
2736 |
+ |
|
2737 |
+ |
Our treatment of the resistance tensor is approximate. $\Xi_f$ for a |
2738 |
+ |
rigid triangular plate would normally be treated as a $6 \times 6$ |
2739 |
+ |
tensor that includes translational and rotational drag as well as |
2740 |
+ |
translational-rotational coupling. The computation of resistance |
2741 |
+ |
tensors for rigid bodies has been detailed |
2742 |
+ |
elsewhere,\cite{JoseGarciadelaTorre02012000,Garcia-de-la-Torre:2001wd,GarciadelaTorreJ2002,Sun:2008fk} |
2743 |
+ |
but the standard approach involving bead approximations would be |
2744 |
+ |
prohibitively expensive if it were recomputed at each step in a |
2745 |
+ |
molecular dynamics simulation. |
2746 |
+ |
|
2747 |
+ |
Instead, we are utilizing an approximate resistance tensor obtained by |
2748 |
+ |
first constructing the Oseen tensor for the interaction of the |
2749 |
+ |
centroid of the facet ($f$) with each of the subfacets $\ell=1,2,3$, |
2750 |
+ |
\begin{equation} |
2751 |
+ |
T_{\ell f}=\frac{A_\ell}{8\pi\eta R_{\ell f}}\left(I + |
2752 |
+ |
\frac{\mathbf{R}_{\ell f}\mathbf{R}_{\ell f}^T}{R_{\ell f}^2}\right) |
2753 |
+ |
\end{equation} |
2754 |
+ |
Here, $A_\ell$ is the area of subfacet $\ell$ which is a triangle |
2755 |
+ |
containing two of the vertices of the facet along with the centroid. |
2756 |
+ |
$\mathbf{R}_{\ell f}$ is the vector between the centroid of facet $f$ |
2757 |
+ |
and the centroid of sub-facet $\ell$, and $I$ is the ($3 \times 3$) |
2758 |
+ |
identity matrix. $\eta$ is the viscosity of the external bath. |
2759 |
+ |
|
2760 |
+ |
The tensors for each of the sub-facets are added together, and the |
2761 |
+ |
resulting matrix is inverted to give a $3 \times 3$ resistance tensor |
2762 |
+ |
for translations of the triangular facet, |
2763 |
+ |
\begin{equation} |
2764 |
+ |
\Xi_f(t) =\left[\sum_{i=1}^3 T_{if}\right]^{-1}. |
2765 |
+ |
\end{equation} |
2766 |
+ |
Note that this treatment ignores rotations (and |
2767 |
+ |
translational-rotational coupling) of the facet. In compact systems, |
2768 |
+ |
the facets stay relatively fixed in orientation between |
2769 |
+ |
configurations, so this appears to be a reasonably good approximation. |
2770 |
+ |
|
2771 |
+ |
At each |
2772 |
+ |
molecular dynamics time step, the following process is carried out: |
2773 |
+ |
\begin{enumerate} |
2774 |
+ |
\item The standard inter-atomic forces ($\nabla_iU$) are computed. |
2775 |
+ |
\item Delaunay triangulation is carried out using the current atomic |
2776 |
+ |
configuration. |
2777 |
+ |
\item The convex hull is computed and facets are identified. |
2778 |
+ |
\item For each facet: |
2779 |
+ |
\begin{itemize} |
2780 |
+ |
\item[a.] The force from the pressure bath ($-\hat{n}_fPA_f$) is |
2781 |
+ |
computed. |
2782 |
+ |
\item[b.] The resistance tensor ($\Xi_f(t)$) is computed using the |
2783 |
+ |
viscosity ($\eta$) of the bath. |
2784 |
+ |
\item[c.] Facet drag ($-\Xi_f(t) \mathbf{v}_f(t)$) forces are |
2785 |
+ |
computed. |
2786 |
+ |
\item[d.] Random forces ($\mathbf{R}_f(t)$) are computed using the |
2787 |
+ |
resistance tensor and the temperature ($T$) of the bath. |
2788 |
+ |
\end{itemize} |
2789 |
+ |
\item The facet forces are divided equally among the vertex atoms. |
2790 |
+ |
\item Atomic positions and velocities are propagated. |
2791 |
+ |
\end{enumerate} |
2792 |
+ |
The Delaunay triangulation and computation of the convex hull are done |
2793 |
+ |
using calls to the qhull library.\cite{Qhull} There is a minimal |
2794 |
+ |
penalty for computing the convex hull and resistance tensors at each |
2795 |
+ |
step in the molecular dynamics simulation (roughly 0.02 $\times$ cost |
2796 |
+ |
of a single force evaluation), and the convex hull is remarkably easy |
2797 |
+ |
to parallelize on distributed memory machines. |
2798 |
+ |
|
2799 |
+ |
|
2800 |
+ |
\begin{longtable}[c]{GBF} |
2801 |
+ |
\caption{Meta-data Keywords: Required parameters for the Langevin Hull integrator} |
2802 |
+ |
\\ |
2803 |
+ |
{\bf keyword} & {\bf units} & {\bf use} \\ \hline |
2804 |
+ |
\endhead |
2805 |
+ |
\hline |
2806 |
+ |
\endfoot |
2807 |
+ |
{\tt viscosity} & poise & Sets the value of viscosity of the implicit |
2808 |
+ |
solven . \\ |
2809 |
+ |
{\tt targetTemp} & K & Sets the target temperature of the system. |
2810 |
+ |
This parameter must be specified to use Langevin Hull dynamics. \\ |
2811 |
+ |
{\tt targetPressure} & atm & Sets the target pressure of the system. |
2812 |
+ |
This parameter must be specified to use Langevin Hull dynamics. \\ |
2813 |
+ |
{\tt usePeriodicBoundaryConditions = false} & logical & Turns off periodic boundary conditions. |
2814 |
+ |
This parameter must be set to \tt false \\ |
2815 |
+ |
\label{table:lhullParameters} |
2816 |
+ |
\end{longtable} |
2817 |
+ |
|
2818 |
+ |
|
2819 |
|
\section{\label{sec:constraints}Constraint Methods} |
2820 |
|
|
2821 |
|
\subsection{\label{section:rattle}The {\sc rattle} Method for Bond |
3379 |
|
\end{center} |
3380 |
|
|
3381 |
|
For example, the phrase {\tt select mass > 16.0 and charge < -2} |
3382 |
< |
wouldselect StuntDoubles which have mass greater than 16.0 and charges |
3382 |
> |
would select StuntDoubles which have mass greater than 16.0 and charges |
3383 |
|
less than -2. |
3384 |
|
|
3385 |
|
\subsection{\label{section:within}Within expressions} |