1 |
\chapter{\label{chapt:methodology}METHODOLOGY} |
2 |
|
3 |
\section{\label{methodSection:rigidBodyIntegrators}Integrators for Rigid Body Motion in Molecular Dynamics} |
4 |
|
5 |
In order to mimic the experiments, which are usually performed under |
6 |
constant temperature and/or pressure, extended Hamiltonian system |
7 |
methods have been developed to generate statistical ensembles, such |
8 |
as canonical ensemble and isobaric-isothermal ensemble \textit{etc}. |
9 |
In addition to the standard ensemble, specific ensembles have been |
10 |
developed to account for the anisotropy between the lateral and |
11 |
normal directions of membranes. The $NPAT$ ensemble, in which the |
12 |
normal pressure and the lateral surface area of the membrane are |
13 |
kept constant, and the $NP\gamma T$ ensemble, in which the normal |
14 |
pressure and the lateral surface tension are kept constant were |
15 |
proposed to address this issue. |
16 |
|
17 |
Integration schemes for rotational motion of the rigid molecules in |
18 |
microcanonical ensemble have been extensively studied in the last |
19 |
two decades. Matubayasi and Nakahara developed a time-reversible |
20 |
integrator for rigid bodies in quaternion representation. Although |
21 |
it is not symplectic, this integrator still demonstrates a better |
22 |
long-time energy conservation than traditional methods because of |
23 |
the time-reversible nature. Extending Trotter-Suzuki to general |
24 |
system with a flat phase space, Miller and his colleagues devised an |
25 |
novel symplectic, time-reversible and volume-preserving integrator |
26 |
in quaternion representation, which was shown to be superior to the |
27 |
time-reversible integrator of Matubayasi and Nakahara. However, all |
28 |
of the integrators in quaternion representation suffer from the |
29 |
computational penalty of constructing a rotation matrix from |
30 |
quaternions to evolve coordinates and velocities at every time step. |
31 |
An alternative integration scheme utilizing rotation matrix directly |
32 |
proposed by Dullweber, Leimkuhler and McLachlan (DLM) also preserved |
33 |
the same structural properties of the Hamiltonian flow. In this |
34 |
section, the integration scheme of DLM method will be reviewed and |
35 |
extended to other ensembles. |
36 |
|
37 |
\subsection{\label{methodSection:DLM}Integrating the Equations of Motion: the |
38 |
DLM method} |
39 |
|
40 |
The DLM method uses a Trotter factorization of the orientational |
41 |
propagator. This has three effects: |
42 |
\begin{enumerate} |
43 |
\item the integrator is area-preserving in phase space (i.e. it is |
44 |
{\it symplectic}), |
45 |
\item the integrator is time-{\it reversible}, making it suitable for Hybrid |
46 |
Monte Carlo applications, and |
47 |
\item the error for a single time step is of order $\mathcal{O}\left(h^4\right)$ |
48 |
for timesteps of length $h$. |
49 |
\end{enumerate} |
50 |
|
51 |
The integration of the equations of motion is carried out in a |
52 |
velocity-Verlet style 2-part algorithm, where $h= \delta t$: |
53 |
|
54 |
{\tt moveA:} |
55 |
\begin{align*} |
56 |
{\bf v}\left(t + h / 2\right) &\leftarrow {\bf v}(t) |
57 |
+ \frac{h}{2} \left( {\bf f}(t) / m \right), \\ |
58 |
% |
59 |
{\bf r}(t + h) &\leftarrow {\bf r}(t) |
60 |
+ h {\bf v}\left(t + h / 2 \right), \\ |
61 |
% |
62 |
{\bf j}\left(t + h / 2 \right) &\leftarrow {\bf j}(t) |
63 |
+ \frac{h}{2} {\bf \tau}^b(t), \\ |
64 |
% |
65 |
\mathsf{A}(t + h) &\leftarrow \mathrm{rotate}\left( h {\bf j} |
66 |
(t + h / 2) \cdot \overleftrightarrow{\mathsf{I}}^{-1} \right). |
67 |
\end{align*} |
68 |
|
69 |
In this context, the $\mathrm{rotate}$ function is the reversible |
70 |
product of the three body-fixed rotations, |
71 |
\begin{equation} |
72 |
\mathrm{rotate}({\bf a}) = \mathsf{G}_x(a_x / 2) \cdot |
73 |
\mathsf{G}_y(a_y / 2) \cdot \mathsf{G}_z(a_z) \cdot \mathsf{G}_y(a_y |
74 |
/ 2) \cdot \mathsf{G}_x(a_x /2), |
75 |
\end{equation} |
76 |
where each rotational propagator, $\mathsf{G}_\alpha(\theta)$, |
77 |
rotates both the rotation matrix ($\mathsf{A}$) and the body-fixed |
78 |
angular momentum (${\bf j}$) by an angle $\theta$ around body-fixed |
79 |
axis $\alpha$, |
80 |
\begin{equation} |
81 |
\mathsf{G}_\alpha( \theta ) = \left\{ |
82 |
\begin{array}{lcl} |
83 |
\mathsf{A}(t) & \leftarrow & \mathsf{A}(0) \cdot \mathsf{R}_\alpha(\theta)^T, \\ |
84 |
{\bf j}(t) & \leftarrow & \mathsf{R}_\alpha(\theta) \cdot {\bf |
85 |
j}(0). |
86 |
\end{array} |
87 |
\right. |
88 |
\end{equation} |
89 |
$\mathsf{R}_\alpha$ is a quadratic approximation to the single-axis |
90 |
rotation matrix. For example, in the small-angle limit, the |
91 |
rotation matrix around the body-fixed x-axis can be approximated as |
92 |
\begin{equation} |
93 |
\mathsf{R}_x(\theta) \approx \left( |
94 |
\begin{array}{ccc} |
95 |
1 & 0 & 0 \\ |
96 |
0 & \frac{1-\theta^2 / 4}{1 + \theta^2 / 4} & -\frac{\theta}{1+ |
97 |
\theta^2 / 4} \\ |
98 |
0 & \frac{\theta}{1+ \theta^2 / 4} & \frac{1-\theta^2 / 4}{1 + |
99 |
\theta^2 / 4} |
100 |
\end{array} |
101 |
\right). |
102 |
\end{equation} |
103 |
All other rotations follow in a straightforward manner. |
104 |
|
105 |
After the first part of the propagation, the forces and body-fixed |
106 |
torques are calculated at the new positions and orientations |
107 |
|
108 |
{\tt doForces:} |
109 |
\begin{align*} |
110 |
{\bf f}(t + h) &\leftarrow |
111 |
- \left(\frac{\partial V}{\partial {\bf r}}\right)_{{\bf r}(t + h)}, \\ |
112 |
% |
113 |
{\bf \tau}^{s}(t + h) &\leftarrow {\bf u}(t + h) |
114 |
\times \frac{\partial V}{\partial {\bf u}}, \\ |
115 |
% |
116 |
{\bf \tau}^{b}(t + h) &\leftarrow \mathsf{A}(t + h) |
117 |
\cdot {\bf \tau}^s(t + h). |
118 |
\end{align*} |
119 |
|
120 |
{\sc oopse} automatically updates ${\bf u}$ when the rotation matrix |
121 |
$\mathsf{A}$ is calculated in {\tt moveA}. Once the forces and |
122 |
torques have been obtained at the new time step, the velocities can |
123 |
be advanced to the same time value. |
124 |
|
125 |
{\tt moveB:} |
126 |
\begin{align*} |
127 |
{\bf v}\left(t + h \right) &\leftarrow {\bf v}\left(t + h / 2 |
128 |
\right) |
129 |
+ \frac{h}{2} \left( {\bf f}(t + h) / m \right), \\ |
130 |
% |
131 |
{\bf j}\left(t + h \right) &\leftarrow {\bf j}\left(t + h / 2 |
132 |
\right) |
133 |
+ \frac{h}{2} {\bf \tau}^b(t + h) . |
134 |
\end{align*} |
135 |
|
136 |
The matrix rotations used in the DLM method end up being more costly |
137 |
computationally than the simpler arithmetic quaternion propagation. |
138 |
With the same time step, a 1000-molecule water simulation shows an |
139 |
average 7\% increase in computation time using the DLM method in |
140 |
place of quaternions. This cost is more than justified when |
141 |
comparing the energy conservation of the two methods as illustrated |
142 |
in Fig.~\ref{methodFig:timestep}. |
143 |
|
144 |
\begin{figure} |
145 |
\centering |
146 |
\includegraphics[width=\linewidth]{timeStep.eps} |
147 |
\caption[Energy conservation for quaternion versus DLM |
148 |
dynamics]{Energy conservation using quaternion based integration |
149 |
versus the method proposed by Dullweber \emph{et al.} with |
150 |
increasing time step. For each time step, the dotted line is total |
151 |
energy using the DLM integrator, and the solid line comes from the |
152 |
quaternion integrator. The larger time step plots are shifted up |
153 |
from the true energy baseline for clarity.} |
154 |
\label{methodFig:timestep} |
155 |
\end{figure} |
156 |
|
157 |
In Fig.~\ref{methodFig:timestep}, the resulting energy drift at |
158 |
various time steps for both the DLM and quaternion integration |
159 |
schemes is compared. All of the 1000 molecule water simulations |
160 |
started with the same configuration, and the only difference was the |
161 |
method for handling rotational motion. At time steps of 0.1 and 0.5 |
162 |
fs, both methods for propagating molecule rotation conserve energy |
163 |
fairly well, with the quaternion method showing a slight energy |
164 |
drift over time in the 0.5 fs time step simulation. At time steps of |
165 |
1 and 2 fs, the energy conservation benefits of the DLM method are |
166 |
clearly demonstrated. Thus, while maintaining the same degree of |
167 |
energy conservation, one can take considerably longer time steps, |
168 |
leading to an overall reduction in computation time. |
169 |
|
170 |
\subsection{\label{methodSection:NVT}Nos\'{e}-Hoover Thermostatting} |
171 |
|
172 |
The Nos\'e-Hoover equations of motion are given by\cite{Hoover1985} |
173 |
\begin{eqnarray} |
174 |
\dot{{\bf r}} & = & {\bf v}, \\ |
175 |
\dot{{\bf v}} & = & \frac{{\bf f}}{m} - \chi {\bf v} ,\\ |
176 |
\dot{\mathsf{A}} & = & \mathsf{A} \cdot |
177 |
\mbox{ skew}\left(\overleftrightarrow{\mathsf{I}}^{-1} \cdot {\bf j}\right), \\ |
178 |
\dot{{\bf j}} & = & {\bf j} \times \left( |
179 |
\overleftrightarrow{\mathsf{I}}^{-1} \cdot {\bf j} \right) - \mbox{ |
180 |
rot}\left(\mathsf{A}^{T} \cdot \frac{\partial V}{\partial |
181 |
\mathsf{A}} \right) - \chi {\bf j}. \label{eq:nosehoovereom} |
182 |
\end{eqnarray} |
183 |
|
184 |
$\chi$ is an ``extra'' variable included in the extended system, and |
185 |
it is propagated using the first order equation of motion |
186 |
\begin{equation} |
187 |
\dot{\chi} = \frac{1}{\tau_{T}^2} \left( |
188 |
\frac{T}{T_{\mathrm{target}}} - 1 \right). \label{eq:nosehooverext} |
189 |
\end{equation} |
190 |
|
191 |
The instantaneous temperature $T$ is proportional to the total |
192 |
kinetic energy (both translational and orientational) and is given |
193 |
by |
194 |
\begin{equation} |
195 |
T = \frac{2 K}{f k_B} |
196 |
\end{equation} |
197 |
Here, $f$ is the total number of degrees of freedom in the system, |
198 |
\begin{equation} |
199 |
f = 3 N + 3 N_{\mathrm{orient}} - N_{\mathrm{constraints}}, |
200 |
\end{equation} |
201 |
and $K$ is the total kinetic energy, |
202 |
\begin{equation} |
203 |
K = \sum_{i=1}^{N} \frac{1}{2} m_i {\bf v}_i^T \cdot {\bf v}_i + |
204 |
\sum_{i=1}^{N_{\mathrm{orient}}} \frac{1}{2} {\bf j}_i^T \cdot |
205 |
\overleftrightarrow{\mathsf{I}}_i^{-1} \cdot {\bf j}_i. |
206 |
\end{equation} |
207 |
|
208 |
In eq.(\ref{eq:nosehooverext}), $\tau_T$ is the time constant for |
209 |
relaxation of the temperature to the target value. To set values |
210 |
for $\tau_T$ or $T_{\mathrm{target}}$ in a simulation, one would use |
211 |
the {\tt tauThermostat} and {\tt targetTemperature} keywords in the |
212 |
{\tt .bass} file. The units for {\tt tauThermostat} are fs, and the |
213 |
units for the {\tt targetTemperature} are degrees K. The |
214 |
integration of the equations of motion is carried out in a |
215 |
velocity-Verlet style 2 part algorithm: |
216 |
|
217 |
{\tt moveA:} |
218 |
\begin{align*} |
219 |
T(t) &\leftarrow \left\{{\bf v}(t)\right\}, \left\{{\bf j}(t)\right\} ,\\ |
220 |
% |
221 |
{\bf v}\left(t + h / 2\right) &\leftarrow {\bf v}(t) |
222 |
+ \frac{h}{2} \left( \frac{{\bf f}(t)}{m} - {\bf v}(t) |
223 |
\chi(t)\right), \\ |
224 |
% |
225 |
{\bf r}(t + h) &\leftarrow {\bf r}(t) |
226 |
+ h {\bf v}\left(t + h / 2 \right) ,\\ |
227 |
% |
228 |
{\bf j}\left(t + h / 2 \right) &\leftarrow {\bf j}(t) |
229 |
+ \frac{h}{2} \left( {\bf \tau}^b(t) - {\bf j}(t) |
230 |
\chi(t) \right) ,\\ |
231 |
% |
232 |
\mathsf{A}(t + h) &\leftarrow \mathrm{rotate} |
233 |
\left(h * {\bf j}(t + h / 2) |
234 |
\overleftrightarrow{\mathsf{I}}^{-1} \right) ,\\ |
235 |
% |
236 |
\chi\left(t + h / 2 \right) &\leftarrow \chi(t) |
237 |
+ \frac{h}{2 \tau_T^2} \left( \frac{T(t)} |
238 |
{T_{\mathrm{target}}} - 1 \right) . |
239 |
\end{align*} |
240 |
|
241 |
Here $\mathrm{rotate}(h * {\bf j} |
242 |
\overleftrightarrow{\mathsf{I}}^{-1})$ is the same symplectic |
243 |
Trotter factorization of the three rotation operations that was |
244 |
discussed in the section on the DLM integrator. Note that this |
245 |
operation modifies both the rotation matrix $\mathsf{A}$ and the |
246 |
angular momentum ${\bf j}$. {\tt moveA} propagates velocities by a |
247 |
half time step, and positional degrees of freedom by a full time |
248 |
step. The new positions (and orientations) are then used to |
249 |
calculate a new set of forces and torques in exactly the same way |
250 |
they are calculated in the {\tt doForces} portion of the DLM |
251 |
integrator. |
252 |
|
253 |
Once the forces and torques have been obtained at the new time step, |
254 |
the temperature, velocities, and the extended system variable can be |
255 |
advanced to the same time value. |
256 |
|
257 |
{\tt moveB:} |
258 |
\begin{align*} |
259 |
T(t + h) &\leftarrow \left\{{\bf v}(t + h)\right\}, |
260 |
\left\{{\bf j}(t + h)\right\}, \\ |
261 |
% |
262 |
\chi\left(t + h \right) &\leftarrow \chi\left(t + h / |
263 |
2 \right) + \frac{h}{2 \tau_T^2} \left( \frac{T(t+h)} |
264 |
{T_{\mathrm{target}}} - 1 \right), \\ |
265 |
% |
266 |
{\bf v}\left(t + h \right) &\leftarrow {\bf v}\left(t |
267 |
+ h / 2 \right) + \frac{h}{2} \left( |
268 |
\frac{{\bf f}(t + h)}{m} - {\bf v}(t + h) |
269 |
\chi(t h)\right) ,\\ |
270 |
% |
271 |
{\bf j}\left(t + h \right) &\leftarrow {\bf j}\left(t |
272 |
+ h / 2 \right) + \frac{h}{2} |
273 |
\left( {\bf \tau}^b(t + h) - {\bf j}(t + h) |
274 |
\chi(t + h) \right) . |
275 |
\end{align*} |
276 |
|
277 |
Since ${\bf v}(t + h)$ and ${\bf j}(t + h)$ are required to |
278 |
caclculate $T(t + h)$ as well as $\chi(t + h)$, they indirectly |
279 |
depend on their own values at time $t + h$. {\tt moveB} is |
280 |
therefore done in an iterative fashion until $\chi(t + h)$ becomes |
281 |
self-consistent. The relative tolerance for the self-consistency |
282 |
check defaults to a value of $\mbox{10}^{-6}$, but {\sc oopse} will |
283 |
terminate the iteration after 4 loops even if the consistency check |
284 |
has not been satisfied. |
285 |
|
286 |
The Nos\'e-Hoover algorithm is known to conserve a Hamiltonian for |
287 |
the extended system that is, to within a constant, identical to the |
288 |
Helmholtz free energy,\cite{Melchionna1993} |
289 |
\begin{equation} |
290 |
H_{\mathrm{NVT}} = V + K + f k_B T_{\mathrm{target}} \left( |
291 |
\frac{\tau_{T}^2 \chi^2(t)}{2} + \int_{0}^{t} \chi(t^\prime) |
292 |
dt^\prime \right). |
293 |
\end{equation} |
294 |
Poor choices of $h$ or $\tau_T$ can result in non-conservation of |
295 |
$H_{\mathrm{NVT}}$, so the conserved quantity is maintained in the |
296 |
last column of the {\tt .stat} file to allow checks on the quality |
297 |
of the integration. |
298 |
|
299 |
\subsection{\label{methodSection:NPTi}Constant-pressure integration with |
300 |
isotropic box deformations (NPTi)} |
301 |
|
302 |
To carry out isobaric-isothermal ensemble calculations {\sc oopse} |
303 |
implements the Melchionna modifications to the |
304 |
Nos\'e-Hoover-Andersen equations of motion,\cite{Melchionna1993} |
305 |
|
306 |
\begin{eqnarray} |
307 |
\dot{{\bf r}} & = & {\bf v} + \eta \left( {\bf r} - {\bf R}_0 \right), \\ |
308 |
\dot{{\bf v}} & = & \frac{{\bf f}}{m} - (\eta + \chi) {\bf v}, \\ |
309 |
\dot{\mathsf{A}} & = & \mathsf{A} \cdot |
310 |
\mbox{ skew}\left(\overleftrightarrow{I}^{-1} \cdot {\bf j}\right),\\ |
311 |
\dot{{\bf j}} & = & {\bf j} \times \left( |
312 |
\overleftrightarrow{I}^{-1} \cdot {\bf j} \right) - \mbox{ |
313 |
rot}\left(\mathsf{A}^{T} \cdot \frac{\partial |
314 |
V}{\partial \mathsf{A}} \right) - \chi {\bf j}, \\ |
315 |
\dot{\chi} & = & \frac{1}{\tau_{T}^2} \left( |
316 |
\frac{T}{T_{\mathrm{target}}} - 1 \right) ,\\ |
317 |
\dot{\eta} & = & \frac{1}{\tau_{B}^2 f k_B T_{\mathrm{target}}} V |
318 |
\left( P - |
319 |
P_{\mathrm{target}} \right), \\ |
320 |
\dot{\mathcal{V}} & = & 3 \mathcal{V} \eta . \label{eq:melchionna1} |
321 |
\end{eqnarray} |
322 |
|
323 |
$\chi$ and $\eta$ are the ``extra'' degrees of freedom in the |
324 |
extended system. $\chi$ is a thermostat, and it has the same |
325 |
function as it does in the Nos\'e-Hoover NVT integrator. $\eta$ is |
326 |
a barostat which controls changes to the volume of the simulation |
327 |
box. ${\bf R}_0$ is the location of the center of mass for the |
328 |
entire system, and $\mathcal{V}$ is the volume of the simulation |
329 |
box. At any time, the volume can be calculated from the determinant |
330 |
of the matrix which describes the box shape: |
331 |
\begin{equation} |
332 |
\mathcal{V} = \det(\mathsf{H}). |
333 |
\end{equation} |
334 |
|
335 |
The NPTi integrator requires an instantaneous pressure. This |
336 |
quantity is calculated via the pressure tensor, |
337 |
\begin{equation} |
338 |
\overleftrightarrow{\mathsf{P}}(t) = \frac{1}{\mathcal{V}(t)} \left( |
339 |
\sum_{i=1}^{N} m_i {\bf v}_i(t) \otimes {\bf v}_i(t) \right) + |
340 |
\overleftrightarrow{\mathsf{W}}(t). |
341 |
\end{equation} |
342 |
The kinetic contribution to the pressure tensor utilizes the {\it |
343 |
outer} product of the velocities denoted by the $\otimes$ symbol. |
344 |
The stress tensor is calculated from another outer product of the |
345 |
inter-atomic separation vectors (${\bf r}_{ij} = {\bf r}_j - {\bf |
346 |
r}_i$) with the forces between the same two atoms, |
347 |
\begin{equation} |
348 |
\overleftrightarrow{\mathsf{W}}(t) = \sum_{i} \sum_{j>i} {\bf |
349 |
r}_{ij}(t) \otimes {\bf f}_{ij}(t). |
350 |
\end{equation} |
351 |
The instantaneous pressure is then simply obtained from the trace of |
352 |
the Pressure tensor, |
353 |
\begin{equation} |
354 |
P(t) = \frac{1}{3} \mathrm{Tr} \left( |
355 |
\overleftrightarrow{\mathsf{P}}(t). \right) |
356 |
\end{equation} |
357 |
|
358 |
In eq.(\ref{eq:melchionna1}), $\tau_B$ is the time constant for |
359 |
relaxation of the pressure to the target value. To set values for |
360 |
$\tau_B$ or $P_{\mathrm{target}}$ in a simulation, one would use the |
361 |
{\tt tauBarostat} and {\tt targetPressure} keywords in the {\tt |
362 |
.bass} file. The units for {\tt tauBarostat} are fs, and the units |
363 |
for the {\tt targetPressure} are atmospheres. Like in the NVT |
364 |
integrator, the integration of the equations of motion is carried |
365 |
out in a velocity-Verlet style 2 part algorithm: |
366 |
|
367 |
{\tt moveA:} |
368 |
\begin{align*} |
369 |
T(t) &\leftarrow \left\{{\bf v}(t)\right\}, \left\{{\bf j}(t)\right\} ,\\ |
370 |
% |
371 |
P(t) &\leftarrow \left\{{\bf r}(t)\right\}, \left\{{\bf v}(t)\right\} ,\\ |
372 |
% |
373 |
{\bf v}\left(t + h / 2\right) &\leftarrow {\bf v}(t) |
374 |
+ \frac{h}{2} \left( \frac{{\bf f}(t)}{m} - {\bf v}(t) |
375 |
\left(\chi(t) + \eta(t) \right) \right), \\ |
376 |
% |
377 |
{\bf j}\left(t + h / 2 \right) &\leftarrow {\bf j}(t) |
378 |
+ \frac{h}{2} \left( {\bf \tau}^b(t) - {\bf j}(t) |
379 |
\chi(t) \right), \\ |
380 |
% |
381 |
\mathsf{A}(t + h) &\leftarrow \mathrm{rotate}\left(h * |
382 |
{\bf j}(t + h / 2) \overleftrightarrow{\mathsf{I}}^{-1} |
383 |
\right) ,\\ |
384 |
% |
385 |
\chi\left(t + h / 2 \right) &\leftarrow \chi(t) + |
386 |
\frac{h}{2 \tau_T^2} \left( \frac{T(t)}{T_{\mathrm{target}}} - 1 |
387 |
\right) ,\\ |
388 |
% |
389 |
\eta(t + h / 2) &\leftarrow \eta(t) + \frac{h |
390 |
\mathcal{V}(t)}{2 N k_B T(t) \tau_B^2} \left( P(t) |
391 |
- P_{\mathrm{target}} \right), \\ |
392 |
% |
393 |
{\bf r}(t + h) &\leftarrow {\bf r}(t) + h |
394 |
\left\{ {\bf v}\left(t + h / 2 \right) |
395 |
+ \eta(t + h / 2)\left[ {\bf r}(t + h) |
396 |
- {\bf R}_0 \right] \right\} ,\\ |
397 |
% |
398 |
\mathsf{H}(t + h) &\leftarrow e^{-h \eta(t + h / 2)} |
399 |
\mathsf{H}(t). |
400 |
\end{align*} |
401 |
|
402 |
Most of these equations are identical to their counterparts in the |
403 |
NVT integrator, but the propagation of positions to time $t + h$ |
404 |
depends on the positions at the same time. {\sc oopse} carries out |
405 |
this step iteratively (with a limit of 5 passes through the |
406 |
iterative loop). Also, the simulation box $\mathsf{H}$ is scaled |
407 |
uniformly for one full time step by an exponential factor that |
408 |
depends on the value of $\eta$ at time $t + h / 2$. Reshaping the |
409 |
box uniformly also scales the volume of the box by |
410 |
\begin{equation} |
411 |
\mathcal{V}(t + h) \leftarrow e^{ - 3 h \eta(t + h /2)}. |
412 |
\mathcal{V}(t) |
413 |
\end{equation} |
414 |
|
415 |
The {\tt doForces} step for the NPTi integrator is exactly the same |
416 |
as in both the DLM and NVT integrators. Once the forces and torques |
417 |
have been obtained at the new time step, the velocities can be |
418 |
advanced to the same time value. |
419 |
|
420 |
{\tt moveB:} |
421 |
\begin{align*} |
422 |
T(t + h) &\leftarrow \left\{{\bf v}(t + h)\right\}, |
423 |
\left\{{\bf j}(t + h)\right\} ,\\ |
424 |
% |
425 |
P(t + h) &\leftarrow \left\{{\bf r}(t + h)\right\}, |
426 |
\left\{{\bf v}(t + h)\right\}, \\ |
427 |
% |
428 |
\chi\left(t + h \right) &\leftarrow \chi\left(t + h / |
429 |
2 \right) + \frac{h}{2 \tau_T^2} \left( \frac{T(t+h)} |
430 |
{T_{\mathrm{target}}} - 1 \right), \\ |
431 |
% |
432 |
\eta(t + h) &\leftarrow \eta(t + h / 2) + |
433 |
\frac{h \mathcal{V}(t + h)}{2 N k_B T(t + h) |
434 |
\tau_B^2} \left( P(t + h) - P_{\mathrm{target}} \right), \\ |
435 |
% |
436 |
{\bf v}\left(t + h \right) &\leftarrow {\bf v}\left(t |
437 |
+ h / 2 \right) + \frac{h}{2} \left( |
438 |
\frac{{\bf f}(t + h)}{m} - {\bf v}(t + h) |
439 |
(\chi(t + h) + \eta(t + h)) \right) ,\\ |
440 |
% |
441 |
{\bf j}\left(t + h \right) &\leftarrow {\bf j}\left(t |
442 |
+ h / 2 \right) + \frac{h}{2} \left( {\bf |
443 |
\tau}^b(t + h) - {\bf j}(t + h) |
444 |
\chi(t + h) \right) . |
445 |
\end{align*} |
446 |
|
447 |
Once again, since ${\bf v}(t + h)$ and ${\bf j}(t + h)$ are required |
448 |
to caclculate $T(t + h)$, $P(t + h)$, $\chi(t + h)$, and $\eta(t + |
449 |
h)$, they indirectly depend on their own values at time $t + h$. |
450 |
{\tt moveB} is therefore done in an iterative fashion until $\chi(t |
451 |
+ h)$ and $\eta(t + h)$ become self-consistent. The relative |
452 |
tolerance for the self-consistency check defaults to a value of |
453 |
$\mbox{10}^{-6}$, but {\sc oopse} will terminate the iteration after |
454 |
4 loops even if the consistency check has not been satisfied. |
455 |
|
456 |
The Melchionna modification of the Nos\'e-Hoover-Andersen algorithm |
457 |
is known to conserve a Hamiltonian for the extended system that is, |
458 |
to within a constant, identical to the Gibbs free energy, |
459 |
\begin{equation} |
460 |
H_{\mathrm{NPTi}} = V + K + f k_B T_{\mathrm{target}} \left( |
461 |
\frac{\tau_{T}^2 \chi^2(t)}{2} + \int_{0}^{t} \chi(t^\prime) |
462 |
dt^\prime \right) + P_{\mathrm{target}} \mathcal{V}(t). |
463 |
\end{equation} |
464 |
Poor choices of $\delta t$, $\tau_T$, or $\tau_B$ can result in |
465 |
non-conservation of $H_{\mathrm{NPTi}}$, so the conserved quantity |
466 |
is maintained in the last column of the {\tt .stat} file to allow |
467 |
checks on the quality of the integration. It is also known that |
468 |
this algorithm samples the equilibrium distribution for the enthalpy |
469 |
(including contributions for the thermostat and barostat), |
470 |
\begin{equation} |
471 |
H_{\mathrm{NPTi}} = V + K + \frac{f k_B T_{\mathrm{target}}}{2} |
472 |
\left( \chi^2 \tau_T^2 + \eta^2 \tau_B^2 \right) + |
473 |
P_{\mathrm{target}} \mathcal{V}(t). |
474 |
\end{equation} |
475 |
|
476 |
Bond constraints are applied at the end of both the {\tt moveA} and |
477 |
{\tt moveB} portions of the algorithm. |
478 |
|
479 |
\subsection{\label{methodSection:NPTf}Constant-pressure integration with a |
480 |
flexible box (NPTf)} |
481 |
|
482 |
There is a relatively simple generalization of the |
483 |
Nos\'e-Hoover-Andersen method to include changes in the simulation |
484 |
box {\it shape} as well as in the volume of the box. This method |
485 |
utilizes the full $3 \times 3$ pressure tensor and introduces a |
486 |
tensor of extended variables ($\overleftrightarrow{\eta}$) to |
487 |
control changes to the box shape. The equations of motion for this |
488 |
method are |
489 |
\begin{eqnarray} |
490 |
\dot{{\bf r}} & = & {\bf v} + \overleftrightarrow{\eta} \cdot \left( {\bf r} - {\bf R}_0 \right), \\ |
491 |
\dot{{\bf v}} & = & \frac{{\bf f}}{m} - (\overleftrightarrow{\eta} + |
492 |
\chi \cdot \mathsf{1}) {\bf v}, \\ |
493 |
\dot{\mathsf{A}} & = & \mathsf{A} \cdot |
494 |
\mbox{ skew}\left(\overleftrightarrow{I}^{-1} \cdot {\bf j}\right) ,\\ |
495 |
\dot{{\bf j}} & = & {\bf j} \times \left( |
496 |
\overleftrightarrow{I}^{-1} \cdot {\bf j} \right) - \mbox{ |
497 |
rot}\left(\mathsf{A}^{T} \cdot \frac{\partial |
498 |
V}{\partial \mathsf{A}} \right) - \chi {\bf j} ,\\ |
499 |
\dot{\chi} & = & \frac{1}{\tau_{T}^2} \left( |
500 |
\frac{T}{T_{\mathrm{target}}} - 1 \right) ,\\ |
501 |
\dot{\overleftrightarrow{\eta}} & = & \frac{1}{\tau_{B}^2 f k_B |
502 |
T_{\mathrm{target}}} V \left( \overleftrightarrow{\mathsf{P}} - P_{\mathrm{target}}\mathsf{1} \right) ,\\ |
503 |
\dot{\mathsf{H}} & = & \overleftrightarrow{\eta} \cdot \mathsf{H} . |
504 |
\label{eq:melchionna2} |
505 |
\end{eqnarray} |
506 |
|
507 |
Here, $\mathsf{1}$ is the unit matrix and |
508 |
$\overleftrightarrow{\mathsf{P}}$ is the pressure tensor. Again, |
509 |
the volume, $\mathcal{V} = \det \mathsf{H}$. |
510 |
|
511 |
The propagation of the equations of motion is nearly identical to |
512 |
the NPTi integration: |
513 |
|
514 |
{\tt moveA:} |
515 |
\begin{align*} |
516 |
T(t) &\leftarrow \left\{{\bf v}(t)\right\}, \left\{{\bf j}(t)\right\} ,\\ |
517 |
% |
518 |
\overleftrightarrow{\mathsf{P}}(t) &\leftarrow \left\{{\bf |
519 |
r}(t)\right\}, |
520 |
\left\{{\bf v}(t)\right\} ,\\ |
521 |
% |
522 |
{\bf v}\left(t + h / 2\right) &\leftarrow {\bf v}(t) |
523 |
+ \frac{h}{2} \left( \frac{{\bf f}(t)}{m} - |
524 |
\left(\chi(t)\mathsf{1} + \overleftrightarrow{\eta}(t) \right) \cdot |
525 |
{\bf v}(t) \right), \\ |
526 |
% |
527 |
{\bf j}\left(t + h / 2 \right) &\leftarrow {\bf j}(t) |
528 |
+ \frac{h}{2} \left( {\bf \tau}^b(t) - {\bf j}(t) |
529 |
\chi(t) \right), \\ |
530 |
% |
531 |
\mathsf{A}(t + h) &\leftarrow \mathrm{rotate}\left(h * |
532 |
{\bf j}(t + h / 2) \overleftrightarrow{\mathsf{I}}^{-1} |
533 |
\right), \\ |
534 |
% |
535 |
\chi\left(t + h / 2 \right) &\leftarrow \chi(t) + |
536 |
\frac{h}{2 \tau_T^2} \left( \frac{T(t)}{T_{\mathrm{target}}} |
537 |
- 1 \right), \\ |
538 |
% |
539 |
\overleftrightarrow{\eta}(t + h / 2) &\leftarrow |
540 |
\overleftrightarrow{\eta}(t) + \frac{h \mathcal{V}(t)}{2 N k_B |
541 |
T(t) \tau_B^2} \left( \overleftrightarrow{\mathsf{P}}(t) |
542 |
- P_{\mathrm{target}}\mathsf{1} \right), \\ |
543 |
% |
544 |
{\bf r}(t + h) &\leftarrow {\bf r}(t) + h \left\{ {\bf v} |
545 |
\left(t + h / 2 \right) + \overleftrightarrow{\eta}(t + |
546 |
h / 2) \cdot \left[ {\bf r}(t + h) |
547 |
- {\bf R}_0 \right] \right\}, \\ |
548 |
% |
549 |
\mathsf{H}(t + h) &\leftarrow \mathsf{H}(t) \cdot e^{-h |
550 |
\overleftrightarrow{\eta}(t + h / 2)} . |
551 |
\end{align*} |
552 |
{\sc oopse} uses a power series expansion truncated at second order |
553 |
for the exponential operation which scales the simulation box. |
554 |
|
555 |
The {\tt moveB} portion of the algorithm is largely unchanged from |
556 |
the NPTi integrator: |
557 |
|
558 |
{\tt moveB:} |
559 |
\begin{align*} |
560 |
T(t + h) &\leftarrow \left\{{\bf v}(t + h)\right\}, |
561 |
\left\{{\bf j}(t + h)\right\}, \\ |
562 |
% |
563 |
\overleftrightarrow{\mathsf{P}}(t + h) &\leftarrow \left\{{\bf r} |
564 |
(t + h)\right\}, \left\{{\bf v}(t |
565 |
+ h)\right\}, \left\{{\bf f}(t + h)\right\} ,\\ |
566 |
% |
567 |
\chi\left(t + h \right) &\leftarrow \chi\left(t + h / |
568 |
2 \right) + \frac{h}{2 \tau_T^2} \left( \frac{T(t+ |
569 |
h)}{T_{\mathrm{target}}} - 1 \right), \\ |
570 |
% |
571 |
\overleftrightarrow{\eta}(t + h) &\leftarrow |
572 |
\overleftrightarrow{\eta}(t + h / 2) + |
573 |
\frac{h \mathcal{V}(t + h)}{2 N k_B T(t + h) |
574 |
\tau_B^2} \left( \overleftrightarrow{P}(t + h) |
575 |
- P_{\mathrm{target}}\mathsf{1} \right) ,\\ |
576 |
% |
577 |
{\bf v}\left(t + h \right) &\leftarrow {\bf v}\left(t |
578 |
+ h / 2 \right) + \frac{h}{2} \left( |
579 |
\frac{{\bf f}(t + h)}{m} - |
580 |
(\chi(t + h)\mathsf{1} + \overleftrightarrow{\eta}(t |
581 |
+ h)) \right) \cdot {\bf v}(t + h), \\ |
582 |
% |
583 |
{\bf j}\left(t + h \right) &\leftarrow {\bf j}\left(t |
584 |
+ h / 2 \right) + \frac{h}{2} \left( {\bf \tau}^b(t |
585 |
+ h) - {\bf j}(t + h) \chi(t + h) \right) . |
586 |
\end{align*} |
587 |
|
588 |
The iterative schemes for both {\tt moveA} and {\tt moveB} are |
589 |
identical to those described for the NPTi integrator. |
590 |
|
591 |
The NPTf integrator is known to conserve the following Hamiltonian: |
592 |
\begin{equation} |
593 |
H_{\mathrm{NPTf}} = V + K + f k_B T_{\mathrm{target}} \left( |
594 |
\frac{\tau_{T}^2 \chi^2(t)}{2} + \int_{0}^{t} \chi(t^\prime) |
595 |
dt^\prime \right) + P_{\mathrm{target}} \mathcal{V}(t) + \frac{f k_B |
596 |
T_{\mathrm{target}}}{2} |
597 |
\mathrm{Tr}\left[\overleftrightarrow{\eta}(t)\right]^2 \tau_B^2. |
598 |
\end{equation} |
599 |
|
600 |
This integrator must be used with care, particularly in liquid |
601 |
simulations. Liquids have very small restoring forces in the |
602 |
off-diagonal directions, and the simulation box can very quickly |
603 |
form elongated and sheared geometries which become smaller than the |
604 |
electrostatic or Lennard-Jones cutoff radii. The NPTf integrator |
605 |
finds most use in simulating crystals or liquid crystals which |
606 |
assume non-orthorhombic geometries. |
607 |
|
608 |
\subsection{\label{methodSection:otherSpecialEnsembles}Other Special Ensembles} |
609 |
|
610 |
\subsubsection{\label{methodSection:NPAT}\textbf{NPAT Ensemble}} |
611 |
|
612 |
A comprehensive understanding of structure¨Cfunction relations of |
613 |
biological membrane system ultimately relies on structure and |
614 |
dynamics of lipid bilayer, which are strongly affected by the |
615 |
interfacial interaction between lipid molecules and surrounding |
616 |
media. One quantity to describe the interfacial interaction is so |
617 |
called the average surface area per lipid. Constat area and constant |
618 |
lateral pressure simulation can be achieved by extending the |
619 |
standard NPT ensemble with a different pressure control strategy |
620 |
|
621 |
\begin{equation} |
622 |
\dot {\overleftrightarrow{\eta}} _{\alpha \beta}=\left\{\begin{array}{ll} |
623 |
\frac{{V(P_{\alpha \beta } - P_{{\rm{target}}} )}}{{\tau_{\rm{B}}^{\rm{2}} fk_B T_{{\rm{target}}} }} |
624 |
& \mbox{if $ \alpha = \beta = z$}\\ |
625 |
0 & \mbox{otherwise}\\ |
626 |
\end{array} |
627 |
\right. |
628 |
\end{equation} |
629 |
|
630 |
Note that the iterative schemes for NPAT are identical to those |
631 |
described for the NPTi integrator. |
632 |
|
633 |
\subsubsection{\label{methodSection:NPrT}\textbf{NP$\gamma$T Ensemble}} |
634 |
|
635 |
Theoretically, the surface tension $\gamma$ of a stress free |
636 |
membrane system should be zero since its surface free energy $G$ is |
637 |
minimum with respect to surface area $A$ |
638 |
\[ |
639 |
\gamma = \frac{{\partial G}}{{\partial A}}. |
640 |
\] |
641 |
However, a surface tension of zero is not appropriate for relatively |
642 |
small patches of membrane. In order to eliminate the edge effect of |
643 |
the membrane simulation, a special ensemble, NP$\gamma$T, is |
644 |
proposed to maintain the lateral surface tension and normal |
645 |
pressure. The equation of motion for cell size control tensor, |
646 |
$\eta$, in $NP\gamma T$ is |
647 |
\begin{equation} |
648 |
\dot {\overleftrightarrow{\eta}} _{\alpha \beta}=\left\{\begin{array}{ll} |
649 |
- A_{xy} (\gamma _\alpha - \gamma _{{\rm{target}}} ) & \mbox{$\alpha = \beta = x$ or $\alpha = \beta = y$}\\ |
650 |
\frac{{V(P_{\alpha \beta } - P_{{\rm{target}}} )}}{{\tau _{\rm{B}}^{\rm{2}} fk_B T_{{\rm{target}}}}} & \mbox{$\alpha = \beta = z$} \\ |
651 |
0 & \mbox{$\alpha \ne \beta$} \\ |
652 |
\end{array} |
653 |
\right. |
654 |
\end{equation} |
655 |
where $ \gamma _{{\rm{target}}}$ is the external surface tension and |
656 |
the instantaneous surface tensor $\gamma _\alpha$ is given by |
657 |
\begin{equation} |
658 |
\gamma _\alpha = - h_z( \overleftrightarrow{P} _{\alpha \alpha } |
659 |
- P_{{\rm{target}}} ) |
660 |
\label{methodEquation:instantaneousSurfaceTensor} |
661 |
\end{equation} |
662 |
|
663 |
There is one additional extended system integrator (NPTxyz), in |
664 |
which each attempt to preserve the target pressure along the box |
665 |
walls perpendicular to that particular axis. The lengths of the box |
666 |
axes are allowed to fluctuate independently, but the angle between |
667 |
the box axes does not change. It should be noted that the NPTxyz |
668 |
integrator is a special case of $NP\gamma T$ if the surface tension |
669 |
$\gamma$ is set to zero. |
670 |
|
671 |
\section{\label{methodSection:zcons}Z-Constraint Method} |
672 |
|
673 |
Based on the fluctuation-dissipation theorem, a force |
674 |
auto-correlation method was developed by Roux and Karplus to |
675 |
investigate the dynamics of ions inside ion channels\cite{Roux1991}. |
676 |
The time-dependent friction coefficient can be calculated from the |
677 |
deviation of the instantaneous force from its mean force. |
678 |
\begin{equation} |
679 |
\xi(z,t)=\langle\delta F(z,t)\delta F(z,0)\rangle/k_{B}T, |
680 |
\end{equation} |
681 |
where% |
682 |
\begin{equation} |
683 |
\delta F(z,t)=F(z,t)-\langle F(z,t)\rangle. |
684 |
\end{equation} |
685 |
|
686 |
If the time-dependent friction decays rapidly, the static friction |
687 |
coefficient can be approximated by |
688 |
\begin{equation} |
689 |
\xi_{\text{static}}(z)=\int_{0}^{\infty}\langle\delta F(z,t)\delta |
690 |
F(z,0)\rangle dt. |
691 |
\end{equation} |
692 |
Allowing diffusion constant to then be calculated through the |
693 |
Einstein relation:\cite{Marrink1994} |
694 |
\begin{equation} |
695 |
D(z)=\frac{k_{B}T}{\xi_{\text{static}}(z)}=\frac{(k_{B}T)^{2}}{\int_{0}^{\infty |
696 |
}\langle\delta F(z,t)\delta F(z,0)\rangle dt}.% |
697 |
\end{equation} |
698 |
|
699 |
The Z-Constraint method, which fixes the z coordinates of the |
700 |
molecules with respect to the center of the mass of the system, has |
701 |
been a method suggested to obtain the forces required for the force |
702 |
auto-correlation calculation.\cite{Marrink1994} However, simply |
703 |
resetting the coordinate will move the center of the mass of the |
704 |
whole system. To avoid this problem, we reset the forces of |
705 |
z-constrained molecules as well as subtract the total constraint |
706 |
forces from the rest of the system after the force calculation at |
707 |
each time step instead of resetting the coordinate. |
708 |
|
709 |
After the force calculation, define $G_\alpha$ as |
710 |
\begin{equation} |
711 |
G_{\alpha} = \sum_i F_{\alpha i}, \label{oopseEq:zc1} |
712 |
\end{equation} |
713 |
where $F_{\alpha i}$ is the force in the z direction of atom $i$ in |
714 |
z-constrained molecule $\alpha$. The forces of the z constrained |
715 |
molecule are then set to: |
716 |
\begin{equation} |
717 |
F_{\alpha i} = F_{\alpha i} - |
718 |
\frac{m_{\alpha i} G_{\alpha}}{\sum_i m_{\alpha i}}. |
719 |
\end{equation} |
720 |
Here, $m_{\alpha i}$ is the mass of atom $i$ in the z-constrained |
721 |
molecule. Having rescaled the forces, the velocities must also be |
722 |
rescaled to subtract out any center of mass velocity in the z |
723 |
direction. |
724 |
\begin{equation} |
725 |
v_{\alpha i} = v_{\alpha i} - |
726 |
\frac{\sum_i m_{\alpha i} v_{\alpha i}}{\sum_i m_{\alpha i}}, |
727 |
\end{equation} |
728 |
where $v_{\alpha i}$ is the velocity of atom $i$ in the z direction. |
729 |
Lastly, all of the accumulated z constrained forces must be |
730 |
subtracted from the system to keep the system center of mass from |
731 |
drifting. |
732 |
\begin{equation} |
733 |
F_{\beta i} = F_{\beta i} - \frac{m_{\beta i} \sum_{\alpha} |
734 |
G_{\alpha}} |
735 |
{\sum_{\beta}\sum_i m_{\beta i}}, |
736 |
\end{equation} |
737 |
where $\beta$ are all of the unconstrained molecules in the system. |
738 |
Similarly, the velocities of the unconstrained molecules must also |
739 |
be scaled. |
740 |
\begin{equation} |
741 |
v_{\beta i} = v_{\beta i} + \sum_{\alpha} |
742 |
\frac{\sum_i m_{\alpha i} v_{\alpha i}}{\sum_i m_{\alpha i}}. |
743 |
\end{equation} |
744 |
|
745 |
At the very beginning of the simulation, the molecules may not be at |
746 |
their constrained positions. To move a z-constrained molecule to its |
747 |
specified position, a simple harmonic potential is used |
748 |
\begin{equation} |
749 |
U(t)=\frac{1}{2}k_{\text{Harmonic}}(z(t)-z_{\text{cons}})^{2},% |
750 |
\end{equation} |
751 |
where $k_{\text{Harmonic}}$ is the harmonic force constant, $z(t)$ |
752 |
is the current $z$ coordinate of the center of mass of the |
753 |
constrained molecule, and $z_{\text{cons}}$ is the constrained |
754 |
position. The harmonic force operating on the z-constrained molecule |
755 |
at time $t$ can be calculated by |
756 |
\begin{equation} |
757 |
F_{z_{\text{Harmonic}}}(t)=-\frac{\partial U(t)}{\partial z(t)}= |
758 |
-k_{\text{Harmonic}}(z(t)-z_{\text{cons}}). |
759 |
\end{equation} |
760 |
|
761 |
|
762 |
\section{\label{methodSection:langevin}Integrators for Langevin Dynamics of Rigid Bodies} |
763 |
|
764 |
%\subsection{\label{methodSection:temperature}Temperature Control} |
765 |
|
766 |
%\subsection{\label{methodSection:pressureControl}Pressure Control} |
767 |
|
768 |
%\section{\label{methodSection:hydrodynamics}Hydrodynamics} |
769 |
|
770 |
%applications of langevin dynamics |
771 |
As an excellent alternative to newtonian dynamics, Langevin |
772 |
dynamics, which mimics a simple heat bath with stochastic and |
773 |
dissipative forces, has been applied in a variety of studies. The |
774 |
stochastic treatment of the solvent enables us to carry out |
775 |
substantially longer time simulation. Implicit solvent Langevin |
776 |
dynamics simulation of met-enkephalin not only outperforms explicit |
777 |
solvent simulation on computation efficiency, but also agrees very |
778 |
well with explicit solvent simulation on dynamics |
779 |
properties\cite{Shen2002}. Recently, applying Langevin dynamics with |
780 |
UNRES model, Liow and his coworkers suggest that protein folding |
781 |
pathways can be possibly exploited within a reasonable amount of |
782 |
time\cite{Liwo2005}. The stochastic nature of the Langevin dynamics |
783 |
also enhances the sampling of the system and increases the |
784 |
probability of crossing energy barrier\cite{Banerjee2004, Cui2003}. |
785 |
Combining Langevin dynamics with Kramers's theory, Klimov and |
786 |
Thirumalai identified the free-energy barrier by studying the |
787 |
viscosity dependence of the protein folding rates\cite{Klimov1997}. |
788 |
In order to account for solvent induced interactions missing from |
789 |
implicit solvent model, Kaya incorporated desolvation free energy |
790 |
barrier into implicit coarse-grained solvent model in protein |
791 |
folding/unfolding study and discovered a higher free energy barrier |
792 |
between the native and denatured states. Because of its stability |
793 |
against noise, Langevin dynamics is very suitable for studying |
794 |
remagnetization processes in various |
795 |
systems\cite{Palacios1998,Berkov2002,Denisov2003}. For instance, the |
796 |
oscillation power spectrum of nanoparticles from Langevin dynamics |
797 |
simulation has the same peak frequencies for different wave |
798 |
vectors,which recovers the property of magnetic excitations in small |
799 |
finite structures\cite{Berkov2005a}. In an attempt to reduce the |
800 |
computational cost of simulation, multiple time stepping (MTS) |
801 |
methods have been introduced and have been of great interest to |
802 |
macromolecule and protein community\cite{Tuckerman1992}. Relying on |
803 |
the observation that forces between distant atoms generally |
804 |
demonstrate slower fluctuations than forces between close atoms, MTS |
805 |
method are generally implemented by evaluating the slowly |
806 |
fluctuating forces less frequently than the fast ones. |
807 |
Unfortunately, nonlinear instability resulting from increasing |
808 |
timestep in MTS simulation have became a critical obstruction |
809 |
preventing the long time simulation. Due to the coupling to the heat |
810 |
bath, Langevin dynamics has been shown to be able to damp out the |
811 |
resonance artifact more efficiently\cite{Sandu1999}. |
812 |
|
813 |
%review rigid body dynamics |
814 |
Rigid bodies are frequently involved in the modeling of different |
815 |
areas, from engineering, physics, to chemistry. For example, |
816 |
missiles and vehicle are usually modeled by rigid bodies. The |
817 |
movement of the objects in 3D gaming engine or other physics |
818 |
simulator is governed by the rigid body dynamics. In molecular |
819 |
simulation, rigid body is used to simplify the model in |
820 |
protein-protein docking study\cite{Gray2003}. |
821 |
|
822 |
It is very important to develop stable and efficient methods to |
823 |
integrate the equations of motion of orientational degrees of |
824 |
freedom. Euler angles are the nature choice to describe the |
825 |
rotational degrees of freedom. However, due to its singularity, the |
826 |
numerical integration of corresponding equations of motion is very |
827 |
inefficient and inaccurate. Although an alternative integrator using |
828 |
different sets of Euler angles can overcome this |
829 |
difficulty\cite{Ryckaert1977, Andersen1983}, the computational |
830 |
penalty and the lost of angular momentum conservation still remain. |
831 |
In 1977, a singularity free representation utilizing quaternions was |
832 |
developed by Evans\cite{Evans1977}. Unfortunately, this approach |
833 |
suffer from the nonseparable Hamiltonian resulted from quaternion |
834 |
representation, which prevents the symplectic algorithm to be |
835 |
utilized. Another different approach is to apply holonomic |
836 |
constraints to the atoms belonging to the rigid |
837 |
body\cite{Barojas1973}. Each atom moves independently under the |
838 |
normal forces deriving from potential energy and constraint forces |
839 |
which are used to guarantee the rigidness. However, due to their |
840 |
iterative nature, SHAKE and Rattle algorithm converge very slowly |
841 |
when the number of constraint increases. |
842 |
|
843 |
The break through in geometric literature suggests that, in order to |
844 |
develop a long-term integration scheme, one should preserve the |
845 |
geometric structure of the flow. Matubayasi and Nakahara developed a |
846 |
time-reversible integrator for rigid bodies in quaternion |
847 |
representation. Although it is not symplectic, this integrator still |
848 |
demonstrates a better long-time energy conservation than traditional |
849 |
methods because of the time-reversible nature. Extending |
850 |
Trotter-Suzuki to general system with a flat phase space, Miller and |
851 |
his colleagues devised an novel symplectic, time-reversible and |
852 |
volume-preserving integrator in quaternion representation. However, |
853 |
all of the integrators in quaternion representation suffer from the |
854 |
computational penalty of constructing a rotation matrix from |
855 |
quaternions to evolve coordinates and velocities at every time step. |
856 |
An alternative integration scheme utilizing rotation matrix directly |
857 |
is RSHAKE\cite{Kol1997}, in which a conjugate momentum to rotation |
858 |
matrix is introduced to re-formulate the Hamiltonian's equation and |
859 |
the Hamiltonian is evolved in a constraint manifold by iteratively |
860 |
satisfying the orthogonality constraint. However, RSHAKE is |
861 |
inefficient because of the iterative procedure. An extremely |
862 |
efficient integration scheme in rotation matrix representation, |
863 |
which also preserves the same structural properties of the |
864 |
Hamiltonian flow as Miller's integrator, is proposed by Dullweber, |
865 |
Leimkuhler and McLachlan (DLM)\cite{Dullweber1997}. |
866 |
|
867 |
%review langevin/browninan dynamics for arbitrarily shaped rigid body |
868 |
Combining Langevin or Brownian dynamics with rigid body dynamics, |
869 |
one can study the slow processes in biomolecular systems. Modeling |
870 |
the DNA as a chain of rigid spheres beads, which subject to harmonic |
871 |
potentials as well as excluded volume potentials, Mielke and his |
872 |
coworkers discover rapid superhelical stress generations from the |
873 |
stochastic simulation of twin supercoiling DNA with response to |
874 |
induced torques\cite{Mielke2004}. Membrane fusion is another key |
875 |
biological process which controls a variety of physiological |
876 |
functions, such as release of neurotransmitters \textit{etc}. A |
877 |
typical fusion event happens on the time scale of millisecond, which |
878 |
is impracticable to study using all atomistic model with newtonian |
879 |
mechanics. With the help of coarse-grained rigid body model and |
880 |
stochastic dynamics, the fusion pathways were exploited by many |
881 |
researchers\cite{Noguchi2001,Noguchi2002,Shillcock2005}. Due to the |
882 |
difficulty of numerical integration of anisotropy rotation, most of |
883 |
the rigid body models are simply modeled by sphere, cylinder, |
884 |
ellipsoid or other regular shapes in stochastic simulations. In an |
885 |
effort to account for the diffusion anisotropy of the arbitrary |
886 |
particles, Fernandes and de la Torre improved the original Brownian |
887 |
dynamics simulation algorithm\cite{Ermak1978,Allison1991} by |
888 |
incorporating a generalized $6\times6$ diffusion tensor and |
889 |
introducing a simple rotation evolution scheme consisting of three |
890 |
consecutive rotations\cite{Fernandes2002}. Unfortunately, unexpected |
891 |
error and bias are introduced into the system due to the arbitrary |
892 |
order of applying the noncommuting rotation |
893 |
operators\cite{Beard2003}. Based on the observation the momentum |
894 |
relaxation time is much less than the time step, one may ignore the |
895 |
inertia in Brownian dynamics. However, assumption of the zero |
896 |
average acceleration is not always true for cooperative motion which |
897 |
is common in protein motion. An inertial Brownian dynamics (IBD) was |
898 |
proposed to address this issue by adding an inertial correction |
899 |
term\cite{Beard2003}. As a complement to IBD which has a lower bound |
900 |
in time step because of the inertial relaxation time, long-time-step |
901 |
inertial dynamics (LTID) can be used to investigate the inertial |
902 |
behavior of the polymer segments in low friction |
903 |
regime\cite{Beard2003}. LTID can also deal with the rotational |
904 |
dynamics for nonskew bodies without translation-rotation coupling by |
905 |
separating the translation and rotation motion and taking advantage |
906 |
of the analytical solution of hydrodynamics properties. However, |
907 |
typical nonskew bodies like cylinder and ellipsoid are inadequate to |
908 |
represent most complex macromolecule assemblies. These intricate |
909 |
molecules have been represented by a set of beads and their |
910 |
hydrodynamics properties can be calculated using variant |
911 |
hydrodynamic interaction tensors. |
912 |
|
913 |
The goal of the present work is to develop a Langevin dynamics |
914 |
algorithm for arbitrary rigid particles by integrating the accurate |
915 |
estimation of friction tensor from hydrodynamics theory into the |
916 |
sophisticated rigid body dynamics. |
917 |
|
918 |
|
919 |
\subsection{Friction Tensor} |
920 |
|
921 |
For an arbitrary rigid body moves in a fluid, it may experience |
922 |
friction force $f_r$ or friction torque $\tau _r$ along the opposite |
923 |
direction of the velocity $v$ or angular velocity $\omega$ at |
924 |
arbitrary origin $P$, |
925 |
\begin{equation} |
926 |
\left( \begin{array}{l} |
927 |
f_r \\ |
928 |
\tau _r \\ |
929 |
\end{array} \right) = - \left( {\begin{array}{*{20}c} |
930 |
{\Xi _{P,t} } & {\Xi _{P,c}^T } \\ |
931 |
{\Xi _{P,c} } & {\Xi _{P,r} } \\ |
932 |
\end{array}} \right)\left( \begin{array}{l} |
933 |
\nu \\ |
934 |
\omega \\ |
935 |
\end{array} \right) |
936 |
\end{equation} |
937 |
where $\Xi _{P,t}t$ is the translation friction tensor, $\Xi _{P,r}$ |
938 |
is the rotational friction tensor and $\Xi _{P,c}$ is the |
939 |
translation-rotation coupling tensor. The procedure of calculating |
940 |
friction tensor using hydrodynamic tensor and comparison between |
941 |
bead model and shell model were elaborated by Carrasco \textit{et |
942 |
al}\cite{Carrasco1999}. An important property of the friction tensor |
943 |
is that the translational friction tensor is independent of origin |
944 |
while the rotational and coupling are sensitive to the choice of the |
945 |
origin \cite{Brenner1967}, which can be described by |
946 |
\begin{equation} |
947 |
\begin{array}{c} |
948 |
\Xi _{P,t} = \Xi _{O,t} = \Xi _t \\ |
949 |
\Xi _{P,c} = \Xi _{O,c} - r_{OP} \times \Xi _t \\ |
950 |
\Xi _{P,r} = \Xi _{O,r} - r_{OP} \times \Xi _t \times r_{OP} + \Xi _{O,c} \times r_{OP} - r_{OP} \times \Xi _{O,c}^T \\ |
951 |
\end{array} |
952 |
\end{equation} |
953 |
Where $O$ is another origin and $r_{OP}$ is the vector joining $O$ |
954 |
and $P$. It is also worthy of mention that both of translational and |
955 |
rotational frictional tensors are always symmetric. In contrast, |
956 |
coupling tensor is only symmetric at center of reaction: |
957 |
\begin{equation} |
958 |
\Xi _{R,c} = \Xi _{R,c}^T |
959 |
\end{equation} |
960 |
The proper location for applying friction force is the center of |
961 |
reaction, at which the trace of rotational resistance tensor reaches |
962 |
minimum. |
963 |
|
964 |
\subsection{Rigid body dynamics} |
965 |
|
966 |
The Hamiltonian of rigid body can be separated in terms of potential |
967 |
energy $V(r,A)$ and kinetic energy $T(p,\pi)$, |
968 |
\[ |
969 |
H = V(r,A) + T(v,\pi ) |
970 |
\] |
971 |
A second-order symplectic method is now obtained by the composition |
972 |
of the flow maps, |
973 |
\[ |
974 |
\varphi _{\Delta t} = \varphi _{\Delta t/2,V} \circ \varphi |
975 |
_{\Delta t,T} \circ \varphi _{\Delta t/2,V}. |
976 |
\] |
977 |
Moreover, $\varphi _{\Delta t/2,V}$ can be divided into two |
978 |
sub-flows which corresponding to force and torque respectively, |
979 |
\[ |
980 |
\varphi _{\Delta t/2,V} = \varphi _{\Delta t/2,F} \circ \varphi |
981 |
_{\Delta t/2,\tau }. |
982 |
\] |
983 |
Since the associated operators of $\varphi _{\Delta t/2,F} $ and |
984 |
$\circ \varphi _{\Delta t/2,\tau }$ are commuted, the composition |
985 |
order inside $\varphi _{\Delta t/2,V}$ does not matter. |
986 |
|
987 |
Furthermore, kinetic potential can be separated to translational |
988 |
kinetic term, $T^t (p)$, and rotational kinetic term, $T^r (\pi )$, |
989 |
\begin{equation} |
990 |
T(p,\pi ) =T^t (p) + T^r (\pi ). |
991 |
\end{equation} |
992 |
where $ T^t (p) = \frac{1}{2}v^T m v $ and $T^r (\pi )$ is defined |
993 |
by \ref{introEquation:rotationalKineticRB}. Therefore, the |
994 |
corresponding flow maps are given by |
995 |
\[ |
996 |
\varphi _{\Delta t,T} = \varphi _{\Delta t,T^t } \circ \varphi |
997 |
_{\Delta t,T^r }. |
998 |
\] |
999 |
The free rigid body is an example of Lie-Poisson system with |
1000 |
Hamiltonian function |
1001 |
\begin{equation} |
1002 |
T^r (\pi ) = T_1 ^r (\pi _1 ) + T_2^r (\pi _2 ) + T_3^r (\pi _3 ) |
1003 |
\label{introEquation:rotationalKineticRB} |
1004 |
\end{equation} |
1005 |
where $T_i^r (\pi _i ) = \frac{{\pi _i ^2 }}{{2I_i }}$ and |
1006 |
Lie-Poisson structure matrix, |
1007 |
\begin{equation} |
1008 |
J(\pi ) = \left( {\begin{array}{*{20}c} |
1009 |
0 & {\pi _3 } & { - \pi _2 } \\ |
1010 |
{ - \pi _3 } & 0 & {\pi _1 } \\ |
1011 |
{\pi _2 } & { - \pi _1 } & 0 \\ |
1012 |
\end{array}} \right) |
1013 |
\end{equation} |
1014 |
Thus, the dynamics of free rigid body is governed by |
1015 |
\begin{equation} |
1016 |
\frac{d}{{dt}}\pi = J(\pi )\nabla _\pi T^r (\pi ) |
1017 |
\end{equation} |
1018 |
One may notice that each $T_i^r$ in Equation |
1019 |
\ref{introEquation:rotationalKineticRB} can be solved exactly. For |
1020 |
instance, the equations of motion due to $T_1^r$ are given by |
1021 |
\begin{equation} |
1022 |
\frac{d}{{dt}}\pi = R_1 \pi ,\frac{d}{{dt}}A = AR_1 |
1023 |
\label{introEqaution:RBMotionSingleTerm} |
1024 |
\end{equation} |
1025 |
where |
1026 |
\[ R_1 = \left( {\begin{array}{*{20}c} |
1027 |
0 & 0 & 0 \\ |
1028 |
0 & 0 & {\pi _1 } \\ |
1029 |
0 & { - \pi _1 } & 0 \\ |
1030 |
\end{array}} \right). |
1031 |
\] |
1032 |
The solutions of Equation \ref{introEqaution:RBMotionSingleTerm} is |
1033 |
\[ |
1034 |
\pi (\Delta t) = e^{\Delta tR_1 } \pi (0),A(\Delta t) = |
1035 |
A(0)e^{\Delta tR_1 } |
1036 |
\] |
1037 |
with |
1038 |
\[ |
1039 |
e^{\Delta tR_1 } = \left( {\begin{array}{*{20}c} |
1040 |
0 & 0 & 0 \\ |
1041 |
0 & {\cos \theta _1 } & {\sin \theta _1 } \\ |
1042 |
0 & { - \sin \theta _1 } & {\cos \theta _1 } \\ |
1043 |
\end{array}} \right),\theta _1 = \frac{{\pi _1 }}{{I_1 }}\Delta t. |
1044 |
\] |
1045 |
To reduce the cost of computing expensive functions in $e^{\Delta |
1046 |
tR_1 }$, we can use Cayley transformation, |
1047 |
\[ |
1048 |
e^{\Delta tR_1 } \approx (1 - \Delta tR_1 )^{ - 1} (1 + \Delta tR_1 |
1049 |
) |
1050 |
\] |
1051 |
The flow maps for $T_2^r$ and $T_3^r$ can be found in the same |
1052 |
manner. |
1053 |
|
1054 |
In order to construct a second-order symplectic method, we split the |
1055 |
angular kinetic Hamiltonian function into five terms |
1056 |
\[ |
1057 |
T^r (\pi ) = \frac{1}{2}T_1 ^r (\pi _1 ) + \frac{1}{2}T_2^r (\pi _2 |
1058 |
) + T_3^r (\pi _3 ) + \frac{1}{2}T_2^r (\pi _2 ) + \frac{1}{2}T_1 ^r |
1059 |
(\pi _1 ) |
1060 |
\]. |
1061 |
Concatenating flows corresponding to these five terms, we can obtain |
1062 |
the flow map for free rigid body, |
1063 |
\[ |
1064 |
\varphi _{\Delta t,T^r } = \varphi _{\Delta t/2,\pi _1 } \circ |
1065 |
\varphi _{\Delta t/2,\pi _2 } \circ \varphi _{\Delta t,\pi _3 } |
1066 |
\circ \varphi _{\Delta t/2,\pi _2 } \circ \varphi _{\Delta t/2,\pi |
1067 |
_1 }. |
1068 |
\] |
1069 |
|
1070 |
The equations of motion corresponding to potential energy and |
1071 |
kinetic energy are listed in the below table, |
1072 |
\begin{center} |
1073 |
\begin{tabular}{|l|l|} |
1074 |
\hline |
1075 |
% after \\: \hline or \cline{col1-col2} \cline{col3-col4} ... |
1076 |
Potential & Kinetic \\ |
1077 |
$\frac{{dq}}{{dt}} = \frac{p}{m}$ & $\frac{d}{{dt}}q = p$ \\ |
1078 |
$\frac{d}{{dt}}p = - \frac{{\partial V}}{{\partial q}}$ & $ \frac{d}{{dt}}p = 0$ \\ |
1079 |
$\frac{d}{{dt}}Q = 0$ & $ \frac{d}{{dt}}Q = Qskew(I^{ - 1} j)$ \\ |
1080 |
$ \frac{d}{{dt}}\pi = \sum\limits_i {\left( {Q^T F_i (r,Q)} \right) \times X_i }$ & $\frac{d}{{dt}}\pi = \pi \times I^{ - 1} \pi$\\ |
1081 |
\hline |
1082 |
\end{tabular} |
1083 |
\end{center} |
1084 |
|
1085 |
Finally, we obtain the overall symplectic flow maps for free moving |
1086 |
rigid body |
1087 |
\begin{align*} |
1088 |
\varphi _{\Delta t} = &\varphi _{\Delta t/2,F} \circ \varphi _{\Delta t/2,\tau } \circ \\ |
1089 |
&\varphi _{\Delta t,T^t } \circ \varphi _{\Delta t/2,\pi _1 } \circ \varphi _{\Delta t/2,\pi _2 } \circ \varphi _{\Delta t,\pi _3 } \circ \varphi _{\Delta t/2,\pi _2 } \circ \varphi _{\Delta t/2,\pi _1 } \circ \\ |
1090 |
&\varphi _{\Delta t/2,\tau } \circ \varphi _{\Delta t/2,F} .\\ |
1091 |
\label{introEquation:overallRBFlowMaps} |
1092 |
\end{align*} |
1093 |
|
1094 |
\subsection{Langevin dynamics for rigid particles of arbitrary shape} |
1095 |
|
1096 |
Consider a Langevin equation of motions in generalized coordinates |
1097 |
\begin{equation} |
1098 |
M_i \dot V_i (t) = F_{s,i} (t) + F_{f,i(t)} + F_{r,i} (t) |
1099 |
\label{LDGeneralizedForm} |
1100 |
\end{equation} |
1101 |
where $M_i$ is a $6\times6$ generalized diagonal mass (include mass |
1102 |
and moment of inertial) matrix and $V_i$ is a generalized velocity, |
1103 |
$V_i = V_i(v_i,\omega _i)$. The right side of Eq. |
1104 |
(\ref{LDGeneralizedForm}) consists of three generalized forces in |
1105 |
lab-fixed frame, systematic force $F_{s,i}$, dissipative force |
1106 |
$F_{f,i}$ and stochastic force $F_{r,i}$. While the evolution of the |
1107 |
system in Newtownian mechanics typically refers to lab-fixed frame, |
1108 |
it is also convenient to handle the rotation of rigid body in |
1109 |
body-fixed frame. Thus the friction and random forces are calculated |
1110 |
in body-fixed frame and converted back to lab-fixed frame by: |
1111 |
\[ |
1112 |
\begin{array}{l} |
1113 |
F_{f,i}^l (t) = A^T F_{f,i}^b (t), \\ |
1114 |
F_{r,i}^l (t) = A^T F_{r,i}^b (t) \\ |
1115 |
\end{array}. |
1116 |
\] |
1117 |
Here, the body-fixed friction force $F_{r,i}^b$ is proportional to |
1118 |
the body-fixed velocity at center of resistance $v_{R,i}^b$ and |
1119 |
angular velocity $\omega _i$, |
1120 |
\begin{equation} |
1121 |
F_{r,i}^b (t) = \left( \begin{array}{l} |
1122 |
f_{r,i}^b (t) \\ |
1123 |
\tau _{r,i}^b (t) \\ |
1124 |
\end{array} \right) = - \left( {\begin{array}{*{20}c} |
1125 |
{\Xi _{R,t} } & {\Xi _{R,c}^T } \\ |
1126 |
{\Xi _{R,c} } & {\Xi _{R,r} } \\ |
1127 |
\end{array}} \right)\left( \begin{array}{l} |
1128 |
v_{R,i}^b (t) \\ |
1129 |
\omega _i (t) \\ |
1130 |
\end{array} \right), |
1131 |
\end{equation} |
1132 |
while the random force $F_{r,i}^l$ is a Gaussian stochastic variable |
1133 |
with zero mean and variance |
1134 |
\begin{equation} |
1135 |
\left\langle {F_{r,i}^l (t)(F_{r,i}^l (t'))^T } \right\rangle = |
1136 |
\left\langle {F_{r,i}^b (t)(F_{r,i}^b (t'))^T } \right\rangle = |
1137 |
2k_B T\Xi _R \delta (t - t'). |
1138 |
\end{equation} |
1139 |
The equation of motion for $v_i$ can be written as |
1140 |
\begin{equation} |
1141 |
m\dot v_i (t) = f_{t,i} (t) = f_{s,i} (t) + f_{f,i}^l (t) + |
1142 |
f_{r,i}^l (t) |
1143 |
\end{equation} |
1144 |
Since the frictional force is applied at the center of resistance |
1145 |
which generally does not coincide with the center of mass, an extra |
1146 |
torque is exerted at the center of mass. Thus, the net body-fixed |
1147 |
frictional torque at the center of mass, $\tau _{n,i}^b (t)$, is |
1148 |
given by |
1149 |
\begin{equation} |
1150 |
\tau _{r,i}^b = \tau _{r,i}^b +r_{MR} \times f_{r,i}^b |
1151 |
\end{equation} |
1152 |
where $r_{MR}$ is the vector from the center of mass to the center |
1153 |
of the resistance. Instead of integrating angular velocity in |
1154 |
lab-fixed frame, we consider the equation of motion of angular |
1155 |
momentum in body-fixed frame |
1156 |
\begin{equation} |
1157 |
\dot \pi _i (t) = \tau _{t,i} (t) = \tau _{s,i} (t) + \tau _{f,i}^b |
1158 |
(t) + \tau _{r,i}^b(t) |
1159 |
\end{equation} |
1160 |
|
1161 |
Embedding the friction terms into force and torque, one can |
1162 |
integrate the langevin equations of motion for rigid body of |
1163 |
arbitrary shape in a velocity-Verlet style 2-part algorithm, where |
1164 |
$h= \delta t$: |
1165 |
|
1166 |
{\tt part one:} |
1167 |
\begin{align*} |
1168 |
v_i (t + h/2) &\leftarrow v_i (t) + \frac{{hf_{t,i}^l (t)}}{{2m_i }} \\ |
1169 |
\pi _i (t + h/2) &\leftarrow \pi _i (t) + \frac{{h\tau _{t,i}^b (t)}}{2} \\ |
1170 |
r_i (t + h) &\leftarrow r_i (t) + hv_i (t + h/2) \\ |
1171 |
A_i (t + h) &\leftarrow rotate\left( {h\pi _i (t + h/2)I^{ - 1} } \right) \\ |
1172 |
\end{align*} |
1173 |
In this context, the $\mathrm{rotate}$ function is the reversible |
1174 |
product of five consecutive body-fixed rotations, |
1175 |
\begin{equation} |
1176 |
\mathrm{rotate}({\bf a}) = \mathsf{G}_x(a_x / 2) \cdot |
1177 |
\mathsf{G}_y(a_y / 2) \cdot \mathsf{G}_z(a_z) \cdot \mathsf{G}_y(a_y |
1178 |
/ 2) \cdot \mathsf{G}_x(a_x /2), |
1179 |
\end{equation} |
1180 |
where each rotational propagator, $\mathsf{G}_\alpha(\theta)$, |
1181 |
rotates both the rotation matrix ($\mathsf{A}$) and the body-fixed |
1182 |
angular momentum ($\pi$) by an angle $\theta$ around body-fixed axis |
1183 |
$\alpha$, |
1184 |
\begin{equation} |
1185 |
\mathsf{G}_\alpha( \theta ) = \left\{ |
1186 |
\begin{array}{lcl} |
1187 |
\mathsf{A}(t) & \leftarrow & \mathsf{A}(0) \cdot \mathsf{R}_\alpha(\theta)^T, \\ |
1188 |
{\bf j}(t) & \leftarrow & \mathsf{R}_\alpha(\theta) \cdot {\bf |
1189 |
j}(0). |
1190 |
\end{array} |
1191 |
\right. |
1192 |
\end{equation} |
1193 |
$\mathsf{R}_\alpha$ is a quadratic approximation to the single-axis |
1194 |
rotation matrix. For example, in the small-angle limit, the |
1195 |
rotation matrix around the body-fixed x-axis can be approximated as |
1196 |
\begin{equation} |
1197 |
\mathsf{R}_x(\theta) \approx \left( |
1198 |
\begin{array}{ccc} |
1199 |
1 & 0 & 0 \\ |
1200 |
0 & \frac{1-\theta^2 / 4}{1 + \theta^2 / 4} & -\frac{\theta}{1+ |
1201 |
\theta^2 / 4} \\ |
1202 |
0 & \frac{\theta}{1+ \theta^2 / 4} & \frac{1-\theta^2 / 4}{1 + |
1203 |
\theta^2 / 4} |
1204 |
\end{array} |
1205 |
\right). |
1206 |
\end{equation} |
1207 |
All other rotations follow in a straightforward manner. |
1208 |
|
1209 |
After the first part of the propagation, the friction and random |
1210 |
forces are generated at the center of resistance in body-fixed frame |
1211 |
and converted back into lab-fixed frame |
1212 |
\[ |
1213 |
f_{t,i}^l (t + h) = - \left( {\frac{{\partial V}}{{\partial r_i }}} |
1214 |
\right)_{r_i (t + h)} + A_i^T (t + h)[F_{f,i}^b (t + h) + F_{r,i}^b |
1215 |
(t + h)], |
1216 |
\] |
1217 |
while the system torque in lab-fixed frame is transformed into |
1218 |
body-fixed frame, |
1219 |
\[ |
1220 |
\tau _{t,i}^b (t + h) = A\tau _{s,i}^l (t) + \tau _{n,i}^b (t) + |
1221 |
\tau _{r,i}^b (t). |
1222 |
\] |
1223 |
Once the forces and torques have been obtained at the new time step, |
1224 |
the velocities can be advanced to the same time value. |
1225 |
|
1226 |
{\tt part two:} |
1227 |
\begin{align*} |
1228 |
v_i (t) &\leftarrow v_i (t + h/2) + \frac{{hf_{t,i}^l (t + h)}}{{2m_i }} \\ |
1229 |
\pi _i (t) &\leftarrow \pi _i (t + h/2) + \frac{{h\tau _{t,i}^b (t + h)}}{2} \\ |
1230 |
\end{align*} |
1231 |
|
1232 |
\subsection{Results and discussion} |