ViewVC Help
View File | Revision Log | Show Annotations | View Changeset | Root Listing
root/group/trunk/oopsePaper/EmpericalEnergy.tex
Revision: 961
Committed: Mon Jan 19 17:24:52 2004 UTC (20 years, 5 months ago) by chrisfen
Content type: application/x-tex
File size: 21051 byte(s)
Log Message:
Adjusted the SSD part of the paper...

File Contents

# User Rev Content
1 mmeineke 806
2 mmeineke 899 \section{\label{sec:empericalEnergy}The Emperical Energy Functions}
3 mmeineke 806
4 mmeineke 915 \subsection{\label{sec:atomsMolecules}Atoms, Molecules and Rigid Bodies}
5 mmeineke 806
6 gezelter 818 The basic unit of an {\sc oopse} simulation is the atom. The parameters
7 mmeineke 806 describing the atom are generalized to make the atom as flexible a
8     representation as possible. They may represent specific atoms of an
9     element, or be used for collections of atoms such as a methyl
10     group. The atoms are also capable of having a directional component
11     associated with them, often in the form of a dipole. Charges on atoms
12 mmeineke 930 are not currently suported by {\sc oopse}.
13 mmeineke 806
14     The second most basic building block of a simulation is the
15 chrisfen 925 molecule. The molecule is a way for {\sc oopse} to keep track of the
16     atoms in a simulation in logical manner. This particular unit will
17     store the identities of all the atoms associated with itself and is
18     responsible for the evaluation of its own bonded interaction
19     (i.e.~bonds, bends, and torsions).
20 mmeineke 899
21 chrisfen 935 As stated previously, one of the features that sets {\sc oopse} apart
22 mmeineke 915 from most of the current molecular simulation packages is the ability
23     to handle rigid body dynamics. Rigid bodies are non-spherical
24     particles or collections of particles that have a constant internal
25     potential and move collectively.\cite{Goldstein01} They are not
26 chrisfen 935 included in most simulation packages because of the requirement to
27     propagate the orientational degrees of freedom. Until recently,
28     integrators which propagate orientational motion have been lacking.
29 mmeineke 915
30     Moving a rigid body involves determination of both the force and
31     torque applied by the surroundings, which directly affect the
32 chrisfen 935 translational and rotational motion in turn. In order to accumulate
33     the total force on a rigid body, the external forces and torques must
34     first be calculated for all the internal particles. The total force on
35     the rigid body is simply the sum of these external forces.
36     Accumulation of the total torque on the rigid body is more complex
37     than the force in that it is the torque applied on the center of mass
38     that dictates rotational motion. The torque on rigid body {\it i} is
39 chrisfen 925 \begin{equation}
40 chrisfen 935 \boldsymbol{\tau}_i=
41     \sum_{a}(\mathbf{r}_{ia}-\mathbf{r}_i)\times \mathbf{f}_{ia}
42     + \boldsymbol{\tau}_{ia},
43 chrisfen 925 \label{eq:torqueAccumulate}
44     \end{equation}
45 chrisfen 935 where $\boldsymbol{\tau}_i$ and $\mathbf{r}_i$ are the torque on and
46     position of the center of mass respectively, while $\mathbf{f}_{ia}$,
47     $\mathbf{r}_{ia}$, and $\boldsymbol{\tau}_{ia}$ are the force on,
48     position of, and torque on the component particles of the rigid body.
49 mmeineke 915
50 chrisfen 935 The summation of the total torque is done in the body fixed axis of
51 mmeineke 915 the rigid body. In order to move between the space fixed and body
52 chrisfen 925 fixed coordinate axes, parameters describing the orientation must be
53     maintained for each rigid body. At a minimum, the rotation matrix
54 chrisfen 935 (\textbf{A}) can be described by the three Euler angles ($\phi,
55     \theta,$ and $\psi$), where the elements of \textbf{A} are composed of
56 chrisfen 925 trigonometric operations involving $\phi, \theta,$ and
57 chrisfen 935 $\psi$.\cite{Goldstein01} In order to avoid numerical instabilities
58 chrisfen 925 inherent in using the Euler angles, the four parameter ``quaternion''
59 chrisfen 935 scheme is often used. The elements of \textbf{A} can be expressed as
60     arithmetic operations involving the four quaternions ($q_0, q_1, q_2,$
61     and $q_3$).\cite{allen87:csl} Use of quaternions also leads to
62     performance enhancements, particularly for very small
63 chrisfen 925 systems.\cite{Evans77}
64 mmeineke 915
65 chrisfen 935 {\sc oopse} utilizes a relatively new scheme that propagates the
66     entire nine parameter rotation matrix internally. Further discussion
67     on this choice can be found in Sec.~\ref{sec:integrate}.
68 chrisfen 925
69 mmeineke 915 \subsection{\label{sec:LJPot}The Lennard Jones Potential}
70    
71     The most basic force field implemented in OOPSE is the Lennard-Jones
72 mmeineke 930 potential. The Lennard-Jones potential. Which mimics the Van der Waals
73     interaction at long distances, and uses an emperical repulsion at
74     short distances. The Lennard-Jones potential is given by:
75 mmeineke 915 \begin{equation}
76     V_{\text{LJ}}(r_{ij}) =
77     4\epsilon_{ij} \biggl[
78     \biggl(\frac{\sigma_{ij}}{r_{ij}}\biggr)^{12}
79     - \biggl(\frac{\sigma_{ij}}{r_{ij}}\biggr)^{6}
80     \biggr]
81     \label{eq:lennardJonesPot}
82     \end{equation}
83 mmeineke 930 Where $r_{ij}$ is the distance between particle $i$ and $j$,
84     $\sigma_{ij}$ scales the length of the interaction, and
85     $\epsilon_{ij}$ scales the well depth of the potential.
86 mmeineke 915
87 mmeineke 930 Because this potential is calculated between all pairs, the force
88 mmeineke 915 evaluation can become computationally expensive for large systems. To
89 mmeineke 930 keep the pair evaluation to a manegable number, OOPSE employs a
90     cut-off radius.\cite{allen87:csl} The cutoff radius is set to be
91     $2.5\sigma_{ii}$, where $\sigma_{ii}$ is the largest Lennard-Jones length
92 mmeineke 915 parameter in the system. Truncating the calculation at
93     $r_{\text{cut}}$ introduces a discontinuity into the potential
94     energy. To offset this discontinuity, the energy value at
95     $r_{\text{cut}}$ is subtracted from the entire potential. This causes
96 mmeineke 930 the potential to go to zero at the cut-off radius.
97 mmeineke 915
98     Interactions between dissimilar particles requires the generation of
99     cross term parameters for $\sigma$ and $\epsilon$. These are
100     calculated through the Lorentz-Berthelot mixing
101     rules:\cite{allen87:csl}
102     \begin{equation}
103     \sigma_{ij} = \frac{1}{2}[\sigma_{ii} + \sigma_{jj}]
104     \label{eq:sigmaMix}
105     \end{equation}
106     and
107     \begin{equation}
108     \epsilon_{ij} = \sqrt{\epsilon_{ii} \epsilon_{jj}}
109     \label{eq:epsilonMix}
110     \end{equation}
111    
112    
113 mmeineke 899 \subsection{\label{sec:DUFF}Dipolar Unified-Atom Force Field}
114    
115 mmeineke 930 The Dipolar Unified-atom Force Field ({\sc duff}) was developed to
116     simulate lipid bilayers. The systems require a model capable of forming
117 mmeineke 899 bilayers, while still being sufficiently computationally efficient to
118     allow simulations of large systems ($\approx$100's of phospholipids,
119     $\approx$1000's of waters) for long times ($\approx$10's of
120     nanoseconds).
121    
122 mmeineke 930 With this goal in mind, {\sc duff} has no point charges. Charge
123     neutral distributions were replaced with dipoles, while most atoms and
124     groups of atoms were reduced to Lennard-Jones interaction sites. This
125 mmeineke 899 simplification cuts the length scale of long range interactions from
126     $\frac{1}{r}$ to $\frac{1}{r^3}$, allowing us to avoid the
127 mmeineke 930 computationally expensive Ewald sum. Instead, we can use
128     neighbor-lists, reaction field, and cutoff radii for the dipolar
129     interactions.
130 mmeineke 899
131 mmeineke 930 As an example, lipid head-groups in {\sc duff} are represented as
132     point dipole interaction sites. By placing a dipole of 20.6~Debye at
133     the head group center of mass, our model mimics the head group of
134     phosphatidylcholine.\cite{Cevc87} Additionally, a Lennard-Jones site
135     is located at the pseudoatom's center of mass. The model is
136     illustrated by the dark grey atom in Fig.~\ref{fig:lipidModel}. The water model we use to complement the dipoles of the lipids is out
137     repaarameterization of the soft sticky dipole (SSD) model of Ichiye
138     \emph{et al.}\cite{liu96:new_model}
139 mmeineke 899
140     \begin{figure}
141 mmeineke 930 \epsfxsize=\linewidth
142 mmeineke 918 \epsfbox{lipidModel.eps}
143 mmeineke 899 \caption{A representation of the lipid model. $\phi$ is the torsion angle, $\theta$ %
144 mmeineke 930 is the bend angle, $\mu$ is the dipole moment of the head group, and n
145     is the chain length.}
146 mmeineke 899 \label{fig:lipidModel}
147     \end{figure}
148    
149     Turning to the tails of the phospholipids, we have used a set of
150     scalable parameters to model the alkyl groups with Lennard-Jones
151     sites. For this, we have used the TraPPE force field of Siepmann
152     \emph{et al}.\cite{Siepmann1998} TraPPE is a unified-atom
153     representation of n-alkanes, which is parametrized against phase
154     equilibria using Gibbs Monte Carlo simulation
155     techniques.\cite{Siepmann1998} One of the advantages of TraPPE is that
156     it generalizes the types of atoms in an alkyl chain to keep the number
157     of pseudoatoms to a minimum; the parameters for an atom such as
158     $\text{CH}_2$ do not change depending on what species are bonded to
159     it.
160    
161     TraPPE also constrains of all bonds to be of fixed length. Typically,
162     bond vibrations are the fastest motions in a molecular dynamic
163     simulation. Small time steps between force evaluations must be used to
164     ensure adequate sampling of the bond potential conservation of
165     energy. By constraining the bond lengths, larger time steps may be
166     used when integrating the equations of motion.
167    
168    
169     \subsubsection{\label{subSec:energyFunctions}{\sc duff} Energy Functions}
170    
171     The total energy of function in {\sc duff} is given by the following:
172     \begin{equation}
173     V_{\text{Total}} = \sum^{N}_{I=1} V^{I}_{\text{Internal}}
174 mmeineke 930 + \sum^{N}_{I=1} \sum_{J>I} V^{IJ}_{\text{Cross}}
175 mmeineke 899 \label{eq:totalPotential}
176     \end{equation}
177     Where $V^{I}_{\text{Internal}}$ is the internal potential of a molecule:
178     \begin{equation}
179     V^{I}_{\text{Internal}} =
180     \sum_{\theta_{ijk} \in I} V_{\text{bend}}(\theta_{ijk})
181 mmeineke 930 + \sum_{\phi_{ijkl} \in I} V_{\text{torsion}}(\phi_{ijkl})
182 mmeineke 899 + \sum_{i \in I} \sum_{(j>i+4) \in I}
183     \biggl[ V_{\text{LJ}}(r_{ij}) + V_{\text{dipole}}
184     (\mathbf{r}_{ij},\boldsymbol{\Omega}_{i},\boldsymbol{\Omega}_{j})
185     \biggr]
186     \label{eq:internalPotential}
187     \end{equation}
188     Here $V_{\text{bend}}$ is the bend potential for all 1, 3 bonded pairs
189 mmeineke 930 within the molecule, and $V_{\text{torsion}}$ is the torsion potential
190     for all 1, 4 bonded pairs. The pairwise portions of the internal
191     potential are excluded for pairs that are closer than three bonds,
192     i.e.~atom pairs farther away than a torsion are included in the
193     pair-wise loop.
194 mmeineke 899
195    
196     The bend potential of a molecule is represented by the following function:
197     \begin{equation}
198     V_{\theta_{ijk}} = k_{\theta}( \theta_{ijk} - \theta_0 )^2 \label{eq:bendPot}
199     \end{equation}
200     Where $\theta_{ijk}$ is the angle defined by atoms $i$, $j$, and $k$
201     (see Fig.~\ref{fig:lipidModel}), and $\theta_0$ is the equilibrium
202     bond angle. $k_{\theta}$ is the force constant which determines the
203     strength of the harmonic bend. The parameters for $k_{\theta}$ and
204     $\theta_0$ are based off of those in TraPPE.\cite{Siepmann1998}
205    
206     The torsion potential and parameters are also taken from TraPPE. It is
207     of the form:
208     \begin{equation}
209 mmeineke 930 V_{\text{torsion}}(\phi) = c_1[1 + \cos \phi]
210 mmeineke 899 + c_2[1 + \cos(2\phi)]
211     + c_3[1 + \cos(3\phi)]
212     \label{eq:origTorsionPot}
213     \end{equation}
214     Here $\phi_{ijkl}$ is the angle defined by four bonded neighbors $i$,
215 mmeineke 930 $j$, $k$, and $l$ (again, see Fig.~\ref{fig:lipidModel}). For
216     computaional efficency, the torsion potential has been recast after
217     the method of CHARMM\cite{charmm1983} whereby the angle series is
218     converted to a power series of the form:
219 mmeineke 899 \begin{equation}
220     V_{\text{torsion}}(\phi_{ijkl}) =
221     k_3 \cos^3 \phi + k_2 \cos^2 \phi + k_1 \cos \phi + k_0
222     \label{eq:torsionPot}
223     \end{equation}
224     Where:
225     \begin{align*}
226     k_0 &= c_1 + c_3 \\
227     k_1 &= c_1 - 3c_3 \\
228     k_2 &= 2 c_2 \\
229     k_3 &= 4c_3
230     \end{align*}
231     By recasting the equation to a power series, repeated trigonometric
232     evaluations are avoided during the calculation of the potential.
233    
234    
235 mmeineke 930 The cross portion of the total potential, $V^{IJ}_{\text{Cross}}$, is
236     as follows:
237     \begin{equation}
238     V^{IJ}_{\text{Cross}} =
239     \sum_{i \in I} \sum_{j \in J}
240     \biggl[ V_{\text{LJ}}(r_{ij}) + V_{\text{dipole}}
241     (\mathbf{r}_{ij},\boldsymbol{\Omega}_{i},\boldsymbol{\Omega}_{j})
242     + V_{\text{sticky}}
243     (\mathbf{r}_{ij},\boldsymbol{\Omega}_{i},\boldsymbol{\Omega}_{j})
244     \biggr]
245     \label{eq:crossPotentail}
246     \end{equation}
247     Where $V_{\text{LJ}}$ is the Lennard Jones potential,
248     $V_{\text{dipole}}$ is the dipole dipole potential, and
249     $V_{\text{sticky}}$ is the sticky potential defined by the SSD
250     model. Note that not all atom types include all interactions.
251 mmeineke 915
252 mmeineke 899 The dipole-dipole potential has the following form:
253     \begin{equation}
254     V_{\text{dipole}}(\mathbf{r}_{ij},\boldsymbol{\Omega}_{i},
255 mmeineke 930 \boldsymbol{\Omega}_{j}) = \frac{|\mu_i||\mu_j|}{4\pi\epsilon_{0}r_{ij}^{3}} \biggl[
256     \boldsymbol{\hat{u}}_{i} \cdot \boldsymbol{\hat{u}}_{j}
257 mmeineke 899 -
258 mmeineke 930 \frac{3(\boldsymbol{\hat{u}}_i \cdot \mathbf{r}_{ij}) %
259     (\boldsymbol{\hat{u}}_j \cdot \mathbf{r}_{ij}) }
260     {r^{2}_{ij}} \biggr]
261 mmeineke 899 \label{eq:dipolePot}
262     \end{equation}
263     Here $\mathbf{r}_{ij}$ is the vector starting at atom $i$ pointing
264     towards $j$, and $\boldsymbol{\Omega}_i$ and $\boldsymbol{\Omega}_j$
265 mmeineke 930 are the Euler angles of atom $i$ and $j$ respectively. $|\mu_i|$ is
266     the magnitude of the dipole moment of atom $i$ and
267     $\boldsymbol{\hat{u}}_i$ is the standard unit orientation vector of
268     $\boldsymbol{\Omega}_i$.
269 mmeineke 899
270    
271 chrisfen 925 \subsection{\label{sec:SSD}The {\sc DUFF} Water Models: SSD/E and SSD/RF}
272 mmeineke 899
273 chrisfen 925 In the interest of computational efficiency, the default solvent used
274 chrisfen 961 by {\sc oopse} is the extended Soft Sticky Dipole (SSD/E) water
275     model.\cite{Gezelter04} The original SSD was developed by Ichiye
276     \emph{et al.}\cite{Ichiye96} as a modified form of the hard-sphere
277     water model proposed by Bratko, Blum, and
278 mmeineke 899 Luzar.\cite{Bratko85,Bratko95} It consists of a single point dipole
279     with a Lennard-Jones core and a sticky potential that directs the
280     particles to assume the proper hydrogen bond orientation in the first
281     solvation shell. Thus, the interaction between two SSD water molecules
282     \emph{i} and \emph{j} is given by the potential
283     \begin{equation}
284 chrisfen 925 V_{ij} =
285     V_{ij}^{LJ} (r_{ij})\ + V_{ij}^{dp}
286     (\mathbf{r}_{ij},\boldsymbol{\Omega}_i,\boldsymbol{\Omega}_j)\ +
287     V_{ij}^{sp}
288     (\mathbf{r}_{ij},\boldsymbol{\Omega}_i,\boldsymbol{\Omega}_j),
289     \label{eq:ssdPot}
290 mmeineke 899 \end{equation}
291     where the $\mathbf{r}_{ij}$ is the position vector between molecules
292 chrisfen 925 \emph{i} and \emph{j} with magnitude equal to the distance $r_{ij}$, and
293 mmeineke 899 $\boldsymbol{\Omega}_i$ and $\boldsymbol{\Omega}_j$ represent the
294 chrisfen 925 orientations of the respective molecules. The Lennard-Jones and dipole
295     parts of the potential are given by equations \ref{eq:lennardJonesPot}
296     and \ref{eq:dipolePot} respectively. The sticky part is described by
297     the following,
298 mmeineke 899 \begin{equation}
299 chrisfen 925 u_{ij}^{sp}(\mathbf{r}_{ij},\boldsymbol{\Omega}_i,\boldsymbol{\Omega}_j)=
300     \frac{\nu_0}{2}[s(r_{ij})w(\mathbf{r}_{ij},
301     \boldsymbol{\Omega}_i,\boldsymbol{\Omega}_j) +
302     s^\prime(r_{ij})w^\prime(\mathbf{r}_{ij},
303     \boldsymbol{\Omega}_i,\boldsymbol{\Omega}_j)]\ ,
304     \label{eq:stickyPot}
305 mmeineke 899 \end{equation}
306 chrisfen 925 where $\nu_0$ is a strength parameter for the sticky potential, and
307     $s$ and $s^\prime$ are cubic switching functions which turn off the
308     sticky interaction beyond the first solvation shell. The $w$ function
309     can be thought of as an attractive potential with tetrahedral
310     geometry:
311 mmeineke 899 \begin{equation}
312 chrisfen 925 w({\bf r}_{ij},{\bf \Omega}_i,{\bf \Omega}_j)=
313     \sin\theta_{ij}\sin2\theta_{ij}\cos2\phi_{ij},
314     \label{eq:stickyW}
315 mmeineke 899 \end{equation}
316 chrisfen 925 while the $w^\prime$ function counters the normal aligned and
317     anti-aligned structures favored by point dipoles:
318 mmeineke 899 \begin{equation}
319 chrisfen 925 w^\prime({\bf r}_{ij},{\bf \Omega}_i,{\bf \Omega}_j)=
320     (\cos\theta_{ij}-0.6)^2(\cos\theta_{ij}+0.8)^2-w^0,
321     \label{eq:stickyWprime}
322 mmeineke 899 \end{equation}
323 chrisfen 925 It should be noted that $w$ is proportional to the sum of the $Y_3^2$
324     and $Y_3^{-2}$ spherical harmonics (a linear combination which
325     enhances the tetrahedral geometry for hydrogen bonded structures),
326     while $w^\prime$ is a purely empirical function. A more detailed
327     description of the functional parts and variables in this potential
328     can be found in the original SSD
329     articles.\cite{Ichiye96,Ichiye96b,Ichiye99,Ichiye03}
330 mmeineke 899
331 chrisfen 925 Since SSD is a single-point {\it dipolar} model, the force
332     calculations are simplified significantly relative to the standard
333     {\it charged} multi-point models. In the original Monte Carlo
334     simulations using this model, Ichiye {\it et al.} reported that using
335     SSD decreased computer time by a factor of 6-7 compared to other
336 chrisfen 961 models.\cite{Ichiye96} What is most impressive is that these savings
337 chrisfen 925 did not come at the expense of accurate depiction of the liquid state
338     properties. Indeed, SSD maintains reasonable agreement with the Soper
339 chrisfen 961 diffraction data for the structural features of liquid
340 chrisfen 925 water.\cite{Soper86,Ichiye96} Additionally, the dynamical properties
341     exhibited by SSD agree with experiment better than those of more
342     computationally expensive models (like TIP3P and
343     SPC/E).\cite{Ichiye99} The combination of speed and accurate depiction
344     of solvent properties makes SSD a very attractive model for the
345     simulation of large scale biochemical simulations.
346 mmeineke 899
347     Recent constant pressure simulations revealed issues in the original
348     SSD model that led to lower than expected densities at all target
349 chrisfen 925 pressures.\cite{Ichiye03,Gezelter04} The default model in {\sc oopse}
350 chrisfen 961 is therefore SSD/E, a density corrected derivative of SSD that
351     exhibits improved liquid structure and transport behavior. If the use
352     of a reaction field long-range interaction correction is desired, it
353     is recommended that the parameters be modified to those of the SSD/RF
354     model. Solvent parameters can be easily modified in an accompanying
355     {\sc BASS} file as illustrated in the scheme below. A table of the
356     parameter values and the drawbacks and benefits of the different
357     density corrected SSD models can be found in reference
358     \ref{Gezelter04}.
359 mmeineke 899
360 chrisfen 925 !!!Place a {\sc BASS} scheme showing SSD parameters around here!!!
361 mmeineke 899
362 mmeineke 930 \subsection{\label{sec:eam}Embedded Atom Method}
363 mmeineke 899
364 chuckv 931 Several other molecular dynamics packages\cite{dynamo86} exist which have the
365 mmeineke 918 capacity to simulate metallic systems, including some that have
366     parallel computational abilities\cite{plimpton93}. Potentials that
367     describe bonding transition metal
368     systems\cite{Finnis84,Ercolessi88,Chen90,Qi99,Ercolessi02} have a
369 chuckv 931 attractive interaction which models ``Embedding''
370     a positively charged metal ion in the electron density due to the
371 mmeineke 918 free valance ``sea'' of electrons created by the surrounding atoms in
372     the system. A mostly repulsive pairwise part of the potential
373     describes the interaction of the positively charged metal core ions
374     with one another. A particular potential description called the
375 chuckv 931 Embedded Atom Method\cite{Daw84,FBD86,johnson89,Lu97}({\sc eam}) that has
376 mmeineke 918 particularly wide adoption has been selected for inclusion in OOPSE. A
377 chuckv 931 good review of {\sc eam} and other metallic potential formulations was done
378 mmeineke 918 by Voter.\cite{voter}
379 mmeineke 915
380 mmeineke 918 The {\sc eam} potential has the form:
381     \begin{eqnarray}
382     V & = & \sum_{i} F_{i}\left[\rho_{i}\right] + \sum_{i} \sum_{j \neq i}
383     \phi_{ij}({\bf r}_{ij}) \\
384     \rho_{i} & = & \sum_{j \neq i} f_{j}({\bf r}_{ij})
385 chuckv 932 \end{eqnarray}S
386 mmeineke 918
387 chuckv 932 where $F_{i} $ is the embedding function that equates the energy required to embed a
388     positively-charged core ion $i$ into a linear superposition of
389 mmeineke 918 sperically averaged atomic electron densities given by
390 chuckv 931 $\rho_{i}$. $\phi_{ij}$ is a primarily repulsive pairwise interaction
391     between atoms $i$ and $j$. In the original formulation of
392     {\sc eam} cite{Daw84}, $\phi_{ij}$ was an entirely repulsive term, however
393 chuckv 933 in later refinements to EAM have shown that non-uniqueness between $F$
394 chuckv 931 and $\phi$ allow for more general forms for $\phi$.\cite{Daw89}
395     There is a cutoff distance, $r_{cut}$, which limits the
396 mmeineke 918 summations in the {\sc eam} equation to the few dozen atoms
397     surrounding atom $i$ for both the density $\rho$ and pairwise $\phi$
398 chuckv 931 interactions. Foiles et al. fit EAM potentials for fcc metals Cu, Ag, Au, Ni, Pd, Pt and alloys of these metals\cite{FDB86}. These potential fits are in the DYNAMO 86 format and are included with {\sc oopse}.
399 mmeineke 918
400 chuckv 931
401 mmeineke 915 \subsection{\label{Sec:pbc}Periodic Boundary Conditions}
402 mmeineke 937
403     \newcommand{\roundme}{\operatorname{round}}
404 mmeineke 915
405 mmeineke 930 \textit{Periodic boundary conditions} are widely used to simulate truly
406     macroscopic systems with a relatively small number of particles. The
407     simulation box is replicated throughout space to form an infinite
408     lattice. During the simulation, when a particle moves in the primary
409     cell, its image in other boxes move in exactly the same direction with
410     exactly the same orientation.Thus, as a particle leaves the primary
411     cell, one of its images will enter through the opposite face.If the
412     simulation box is large enough to avoid "feeling" the symmetries of
413     the periodic lattice, surface effects can be ignored. Cubic,
414     orthorhombic and parallelepiped are the available periodic cells In
415     OOPSE. We use a matrix to describe the property of the simulation
416     box. Therefore, both the size and shape of the simulation box can be
417     changed during the simulation. The transformation from box space
418     vector $\mathbf{s}$ to its corresponding real space vector
419     $\mathbf{r}$ is defined by
420     \begin{equation}
421     \mathbf{r}=\underline{\underline{H}}\cdot\mathbf{s}%
422     \end{equation}
423    
424    
425     where $H=(h_{x},h_{y},h_{z})$ is a transformation matrix made up of
426     the three box axis vectors. $h_{x},h_{y}$ and $h_{z}$ represent the
427     three sides of the simulation box respectively.
428    
429     To find the minimum image, we convert the real vector to its
430     corresponding vector in box space first, \bigskip%
431     \begin{equation}
432     \mathbf{s}=\underline{\underline{H}}^{-1}\cdot\mathbf{r}%
433     \end{equation}
434     And then, each element of $\mathbf{s}$ is wrapped to lie between -0.5 to 0.5,
435     \begin{equation}
436 mmeineke 937 s_{i}^{\prime}=s_{i}-\roundme(s_{i})
437 mmeineke 930 \end{equation}
438     where
439    
440     %
441    
442     \begin{equation}
443 mmeineke 937 \roundme(x)=\left\{
444     \begin{array}{cc}
445 mmeineke 930 \lfloor{x+0.5}\rfloor & \text{if \ }x\geqslant0\\
446     \lceil{x-0.5}\rceil & \text{otherwise}%
447     \end{array}
448     \right.
449     \end{equation}
450 mmeineke 937 For example, $\roundme(3.6)=4$, $\roundme(3.1)=3$, $\roundme(-3.6)=-4$,
451     $\roundme(-3.1)=-3$.
452 mmeineke 930
453     Finally, we obtain the minimum image coordinates by transforming back
454     to real space,%
455    
456     \begin{equation}
457     \mathbf{r}^{\prime}=\underline{\underline{H}}^{-1}\cdot\mathbf{s}^{\prime}%
458     \end{equation}
459