1: %\documentclass{epl}
2: \documentstyle[pre,aps,tighten,epsf,floats]{revtex}
3: \begin{document}
4: %\usepackage{graphics}
5: %\usepackage{psfrag}
6: \title{Multiple current reversals in forced inhomogeneous ratchets}
7: \author{\large Debasis Dan,$^{1,*}$ Mangal C. Mahato,$^{2}$ andA. M. Jayannavar$^{1,\dagger }$ \vspace{0.2in}}\address{\large $^1$ Institute of Physics, Sachivalaya Marg,
8: Bhubaneswar 751005, India}
9: \address{\large $^2$ Department of Physics, Guru Ghasidas University, Bilaspur 495009, India}\maketitle
10: \maketitle
11: \begin{abstract}
12:
13: Transport properties of overdamped Brownian paricles in a rocked
14: thermal ratchet with space dependent
15: friction coefficient is studied. By tuning the parameters, the
16: direction of current exhibit multiple reversals, both as a function
17: of the thermal noise strength as well as the amplitude of rocking
18: force. Current reversals also occur under
19: deterministic conditions and exhibits intriguing structure. All these
20: results arise due to mutual interplay between potential asymmetry,
21: noise, driving frequency and inhomogeneous friction.
22:
23: \end{abstract}
24: \vspace{1.0in}
25: %\newpage
26: Fluctuation induced transport in ratchet systems has been an
27: active field of research over the last decade. In these systems in the
28: absence of any net microscopic forces, asymmetric potential can be
29: used to induce a unidirectional particle flow when subjected to an external
30: nonthermal fluctuations. These studies have been motivated in
31: part by the attempt
32: to understand the mechanism of movement of protein motors in biological
33: systems~\cite{Motor}. To this effect several physical models have been proposed under the
34: name of rocking ratchets~\cite{Rock,Rock2,Main}, flashing
35: ratchets~\cite{Flash,Flash2},
36: diffusion ratchets~\cite{Diffu}, correlation
37: ratchets~\cite{Corr}, etc. In all these studies the potential is taken
38: to be asymmetric
39: in space. It has also been shown that one can obtain unidirectional
40: current
41: in the presence of spatially symmetric potentials. For these nonequilibrium
42: systems external random force should be time asymmetric~\cite{Ma} or the presence
43: of space dependent mobility is
44: required~\cite{IntMod,Mill,Jay,But,Kam,Spc_mob,Dan3}. By suitably tuning the system
45: parameters such as temperature, friction coefficient, mass, etc, one can
46: even change the direction of the current. Indeed, the study of current
47: reversal phenomena has given rise to a research activity on its own.
48: The motivation being the possibility of new particle separation
49: devices superior to existing methods such as electrophoretic
50: method for particles of micrometer scale~\cite{Kep}.
51:
52: Bartussek \textit{et. al.}~\cite{Main} showed the occurrence of
53: current reversal in
54: rocked thermal ratchet with both amplitude of rocking force as well as
55: the temperature of thermal bath. They attributed this current reversal to the
56: ``mutual interplay between noise and finite-frequency driving''.
57: Multiple current reversals
58: have also been shown in the deterministic limit of these ratchets when
59: the inertial term is taken into account~\cite{Rock2,Under}. However
60: these multiple current reversals in inertial ratchets are not robust
61: in the presence of noise. Beside rocking ratchets
62: current reversals have also been observed in flashing
63: ratchets~\cite{Reim,Bier,Flash2}. We in this work report that
64: that multiple reversals can be achieved even in rocked \textit{overdamped} ratchet
65: in the presence of space dependent mobility, as a function of the noise strength
66: and amplitude of the rocking force. More over these systems
67: show current reversals when rocked adiabatically.
68: In the deterministic overdamped case we get current reversal as a function
69: of the amplitude of rocking force. Most of our results are attributed to the
70: presence of space dependent mobility.
71: We have studied the same system as due to Bartussek \textit{et
72: al}~\cite{Main}, except the presence
73: of space dependent friction term. In earlier work the spatial asymmetry of
74: the potential is responsible for unidirectional currents and their reversal
75: as function of frequency. As mentioned previouly one can get
76: unidirectional currents in the presence of symmetric potentials, but
77: for this space dependent friction is
78: required~\cite{IntMod,Mill,Jay,But,Kam,Spc_mob,Dan3}. In these systems transport
79: direction is given by the phase shift between mobility and periodic
80: potential. This phase shift induces spatial symmetry breaking as required for
81: the directed motion.
82: Appropriately choosing the phase shift leads to current reversal. We would like to
83: emphasize that space dependent friction does not alter the equilibrium
84: property of the system, however when the system is driven out of equilibrium
85: not trivial dynamical effects arise due to space dependent friction~\cite{Dan3,PLA,PRE}. Naturally
86: in our present system we expect additional effects arising due to combination
87: of spatial asymmetry and position dependence of friction coefficient.
88: It is to be noted that systems with space dependent friction are not uncommon.
89: Brownian motion in confined geometries show space dependent friction~\cite{Conf}.
90: Particles diffusing close to surface have space dependent
91: friction coefficient~\cite{Conf,Surf}. It is believed that molecular motor
92: proteins move
93: close along the periodic structure of microtubules and will therefore
94: experience a position dependent mobility~\cite{Spc_mob}. Frictional inhomogeneities are
95: common in super lattice structures and Josephson
96: junctions~\cite{Falco} also.
97:
98: We consider an overdamped Brownian particle moving in a
99: asymmetric potential V(x) with space dependent friction coefficient
100: $\eta$(x) under the influence of external force field $F$(t) at
101: temperature $T$.
102: Throughout our analysis we take the ratchet potential
103: V(x) = $-\frac{1}{2 \pi} (\sin (2\pi x) + \frac{\mu}{4} \sin (4 \pi
104: x))$. Here $\mu$ is the asymmetry parameter with values taken in the
105: range $0 <\mu <1 $, friction coefficient $\eta (x) = \eta_{0}(1-\lambda \sin
106: (2\pi x + \phi))$, $|\lambda| < 1$. $\phi$ determines the relative
107: phase shift between friction coefficient and potential .
108: The forcing term is
109: taken to be F(t) = A$\sin (w t + \theta)$, ($w = \frac{2
110: \pi}{\tau}$, where $\tau$ is period of force) . Without any
111: loss of generality $\theta$ is taken to be zero.
112: The correct Langevin equation for this system in the overdamped limit
113: is given by~\cite{IntMod,Pram,Sancho}
114:
115: %
116: \begin{equation}
117: \dot{x} = -\frac{(V'(x) - F(t))}{\eta (x)} - k_{B}T \frac{\eta '(x)}{(\eta
118: (x))^{2}} + \sqrt{\frac{k_{B}T}{\eta (x)}}\xi (t) ,
119: \label{Langvn}
120: \end{equation}
121: %
122: where $\xi(t)$ is a zero mean thermal Gaussian noise with correlation
123: $<\xi (t) \xi (t')> = 2 \delta (t-t')$
124: The equation of motion is equivalently described by the Fokker
125: Planck Equation ( FPE)~\cite{IntMod}
126:
127: %
128: \begin{equation}
129: \frac{\partial P(x,t)}{\partial t} = \frac{\partial}{\partial x}
130: \frac{1}{\eta (x)} [k_{B}T \frac{\partial}{\partial x} + (V'(x) -
131: F(t))] P(x,t) .
132: \label{FPE}
133: \end{equation}
134: %
135: where $P(x,t)$ is the probability density at position $x$ at time
136: $t$. Equation~(\ref{FPE}), in the form of a continuity equation
137:
138: \begin{equation}
139: \frac{\partial P(x,t)}{\partial t} = - \frac{\partial
140: J(x,t)}{\partial x} ,
141: \label{Conti}
142: \end{equation}
143:
144: where
145: \begin{equation}
146: J(x,t) = -\frac{1}{\eta(x)} [(V'(x)-F(t)) + k_{B}T
147: \frac{\partial}{\partial x}]P(x,t),
148: \label{Current}
149: \end{equation}
150: is the probability current. Since the potential and the driving force
151: have spatial and temporal periodicity respectively, therefore $J(x,t)
152: = J(x + 1,t + \tau)$, ~\cite{Risk,Main}. The average current $j$
153: in the system is given by
154: \begin{equation}
155: j = \lim_{t \rightarrow \infty} \frac{1}{\tau} \int_{t}^{t+\tau} dt
156: \int_{0}^{1} J(x,t) dx .
157: \label{Avg_curr}
158: \end{equation}
159: %
160: It should be noted that for symmetric potential and $\lambda = 0$, $j
161: = 0$. Rectification of current is possible if the potential is either
162: asymmetric or $\lambda \neq 0$ with $\phi \neq 0, \pi$~\cite{Dan3}. $j$ is
163: independent of the initial phase $\theta$ of the driving force.
164:
165: We explore various parameter regimes of the problem
166: extensively by solving the FPE, Eq.~(\ref{FPE}) numerically with
167: finite difference method.
168: Throughout we have set current $j$ and all other physical quantities
169: such as $T$, $A$, $w$ in dimensionless form .
170:
171: In the Fig.~\ref{j-T}, the average current $j$ is plotted
172: \textit{vs} $T$ for various values of $w$. Here $\lambda = 0.1 , \phi =
173: 0.2 \pi$, $A = 0.5$ and $\mu = 1$. Unlike Bartussek et al. ~\cite{Main},
174: where they show that
175: current reversal is not possible under adiabatic conditions,
176: however current reversal even in adiabatic condition can be obtained
177: in the present case for
178: some values of $\phi$~\cite{Dan3}. For moderately high frequency $w = 4.0$
179: the current reverses its sign \textit{twice} at $T=0.1$ and
180: $T=0.22$ and asymptotically goes to zero for higher values of $T$ as shown
181: in the figure. This phenomena of twice current reversal with
182: temperature $T$ is the foremost feature of our system, previously
183: unseen in any overdamped system.
184: The inset shows the zero contour plot
185: of the current $j$ versus $T$ and $\phi$ for three values of
186: $w$. Crossing this zero contour line implies current reversal.
187: It can be seen that twice current reversal occurs only for a very narrow
188: range of $\phi$. For frequencies higher than certain
189: critical frequency $w_{c}(\phi)$ current flows in only one direction for
190: all temperatures, i.e no current reversal occurs. The plot
191: $w_{c}$ \textit{vs} $\phi$ is shown in Fig.~\ref{wc_phi}.
192: The adiabatic curve $w < 1$ is not shown in the figure as it
193: goes much beyond the scale of the graph. However it has a similar qualitative
194: shape as for $w = 3.0$ curve. As mentioned before, currents are
195: due to the combined effect of phase shift $\phi$ coming from space dependent
196: friction and asymmetry parameter $\mu$. In the regime $\phi = 0.2 \pi$
197: and $A = 0.5 $ and in the absence of asymmetry current flows in the negative
198: direction for all values of $T$. The asymmetric case ( $\mu = 1.0$) in the
199: absence of space dependent friction give current in the positive direction
200: only as a function of temperature. Separately in both these cases
201: absolute value of current
202: exhibits a maxima as a function of $T$, reminiscent of stochastic
203: resonance phenomena. In a purely asymmetric case ($\lambda = 0$) current
204: vanishes rapidly when $T$ exceeds the temperature associated with the
205: barrier height. Whereas, in the symmetric case due to space dependent friction
206: absolute values of currents are significantly higher and decay slowly to
207: zero in the large temperature regime. Naturally in the presence of
208: both asymmetry and space dependent friction for the case under study
209: the low temperature regime is dominated by the effect of asymmetry while the
210: high temperature regime is dominated by space dependent friction. From this
211: , one can qualitatively explain the current reversals from positive
212: to negative side as a function of temperature even in the adiabatic limit.
213: In the regime $\phi > \pi$ current due to space dependent friction and
214: potential asymmetry flows in same direction and hence no reversal is possible
215: as a function of $T$. For
216: frequencies higher than the interwell frequency $w_{0}$,
217: the low temperature scenario
218: changes. The direction of current in this regime is more of a
219: interplay between potential asymmetry and $w$ than that of
220: $\lambda$. Due to higher frequency the Brownian particles do not get
221: enough time to cross the right barrier which is at a larger distance
222: from the minima. Number of particles moving about the
223: potential minima increases. This fact is amply reflected in Fig.~\ref{P_av},
224: where the time averaged probability curve
225: $ P_{av}(x) = \frac{1}{\tau} \int_{0}^{\tau} P(x,t) dt$
226: is plotted as function of $x$ for various values of
227: $w$ with $T = 0.05$. It is to be noted that the distribution is
228: independent of $\eta (x)$. Figure~\ref{P_av} shows that the
229: probability of finding the
230: Brownian particles near the minima of the potential well increases with
231: increasing frequency, consequently the probable number of particles
232: near the potential barrier decreases. Since the distance from a
233: potential minima to the basin of attraction of next minima is less
234: from the steeper side than from the slanted side, hence in one period
235: the particles get enough time to climb the potential barrier from the
236: steeper side than from the slanted side, resulting in a negative
237: current. On increasing the temperature , the particles get kicks of
238: larger intensity and hence they easily cross the slanted barrier,
239: resulting in a current reversal and positive current. On further
240: increasing the temperature, the effect of $\lambda$ dominates as
241: mentioned previously and as a result the current
242: again flows in the negative direction, implying a second current
243: reversal as shown in dotted curve of Fig~\ref{j-T}. It is obvious from
244: the above argument that for higher frequency the first reversal
245: temperature will be higher as shown in the Fig.~\ref{j-T-w}. The
246: dotted line in the base of Fig.~\ref{j-T-w} shows the contour of zero
247: current. But as mentioned previously, since the effect of $\lambda$ dominates
248: for higher temperature, therefore the second reversal temperature
249: decreases with increasing frequency. Beyond the critical frequency $w_{c}$
250: there is no reversal of current as shown in
251: Fig.(~\ref{j-T-w},~\ref{wc_phi}). In the high frequency regime
252: the effect of space dependent friction dominates the nature of current.
253: We would like to emphasize here that in absence of asymmetry current reversals does
254: not take place.
255:
256: Multiple current reversals can also be seen when the amplitude ( $A$) of the
257: forcing term is varied in a suitable parameter regime of our system.
258: In Fig.~\ref{j-A}, the plot of $j$ versus $A$ is shown for
259: different values of $w$, keeping $\lambda$, $\phi$ and $T$ fixed at
260: 0.1, $0.88\pi$ and $0.05$ respectively. For $w = 4.0$ curve, we can see as
261: many as four current reversals. For very large value of $A$, the current
262: asymptotically goes to a constant value depending on the value of
263: $\phi$, as was previously shown for the adiabatic
264: case~\cite{Dan3}. This value was shown to be $-\frac{\lambda}{2} \sin
265: (\phi)$. This is special to the space dependent friction
266: ratchet where the currents saturate to a finite value in the large $A$
267: limit. In the absence of space dependent friction it is to be
268: noted that currents decay to zero in the same asymptotic
269: regime. As the asymptotic value depends on the phase $\phi$, so we can
270: choose it appropriately to make the asymptotic current positive or
271: negative. In the present case $\phi$ is chosen such that the
272: asymptotic current is negative which guarantees at least one current
273: reversal irrespective of frequency. The oscillatory behavior in the $j-A$
274: characteristics is the reminiscent of the deterministic dynamics
275: \cite{Main,Lich} which will be discussed later. The inset in Fig.~\ref{j-A} shows the
276: zero contour of $j$ versus $\phi$ and $A$ for $w = 4.0$. For $\phi > \pi$
277: only two current reversals can be seen. For $\pi > \phi > \pi - \epsilon$
278: (where $\epsilon$ is a small number ) or $0 < \phi < \epsilon$, four
279: or more current reversals can occur. The value of $\epsilon$ depends
280: critically on $T$, large for small $T$ and vice-versa. It is also to be noted
281: that we have even number of reversals for finite frequency driving.
282: These oscillatory features along with their associated current reversals
283: disappear in the high temperature regime as expected.
284: Figure~\ref{Deter} is for $w = 0.25$ and shows that
285: current reversal is also possible
286: in the deterministic regime of the overdamped system. This deterministic
287: reversal of current cannot be attributed to the chaotic motion of the
288: system like in an forced underdamped oscillator~\cite{Under}. It solely arises due the
289: presence of space dependent friction. As shown in Ref.~\cite{Main} here too
290: the deterministic current shows quantization and phase locking
291: behaviour. However all these complex features are not robust in the
292: presence of noise as discussed in earlier literature.
293:
294: In conclusion, we have studied the transport properties of
295: overdamped Brownian particles moving in an asymmetric potential
296: with space dependent
297: friction coefficient and rocked by periodic force. We observe several novel
298: and complex features arising due to the interplay between asymmetry and
299: inhomogeneous friction.
300: Currents in the low temperature regime is mostly influenced by the asymmetry
301: of the potential. At higher temperatures it is controlled by the modulation
302: parameter $\lambda$ of the friction coefficient. We find current reversal
303: with temperature even when the forcing is adiabatic. In the presence of
304: finite frequency, twice current reversals are seen. As function of
305: amplitude of the forcing term we observe multiple current reversals.
306: Current even
307: reverses its sign in the adiabatic deterministic regime.
308: All the above results can be understood in a qualitative manner.
309: We expect that our analysis should be applicable for the motion of particle
310: in porous media and for molecular motors where space dependent friction can
311: arise due to the confinement of particles.
312:
313: MCM acknowledges partial financial support and hospitality
314: from the Institute of Physics, Bhubaneswar. MCM and AMJ acknowledgepartial financial support from the Board of Research in Nuclear
315: Sciences, DAE, India.
316:
317: %
318: % FIG 1
319: %
320: %\begin{figure}
321: %\centerline{
322: %\psfig{file=fig1.ps,width=1.8in,%
323: %bbllx=109pt,bblly=107pt,bburx=529pt,bbury=934pt}}
324: %\vspace{0.1in}
325: %\setlength{\columnwidth}{3.2in}
326: %\centerline{\caption{
327: %$I$-$V$ characteristic given by the PT with $\gamma \approx 2$
328: %(solid lines). The dashed line is a conjecture.
329: %\label{Fig_I_V}
330: %}}
331: %\end{figure}
332: %
333:
334:
335:
336: \begin{thebibliography}{0}
337:
338: \bibitem{Motor}
339: S. Leibler, Nature {\bf370},(1994),412;
340: J. Maddox, ibid {\bf365}(1993),203;
341: ibid,{\bf368},(1994),287;
342: ibid,{\bf368},(1994),287.
343:
344: \bibitem{Rock}
345: M.O. Magnasco, Phys. Rev. Lett. {\bf 71}, 1477, (1993);
346: I. Derenyi and T. Vicsek, Phys. Rev. Lett. {\bf75},(1995),374.
347:
348: \bibitem{Rock2}
349: P. Jung, J. G. Kissner and P. H\"{a}nggi, Phys. Rev. Lett,
350: {\bf76}(1996),3436.
351:
352: \bibitem{Main} R. Bartussek, P. Hanggi, and J.G. Kissner, Europhys. Lett.
353: {\bf 28}, 459, (1994).
354:
355:
356: \bibitem{Flash}
357: J. Prost et al, Phys. Rev. Lett. {\bf72},(1994),2652;
358: J. F. Chauwin, A. Ajdari and J. Prost,
359: Europhys. Lett. {\bf27},(1994), 421;
360: J. Rousselet et al, Nature, {\bf370}, (1994), 446;
361: C. Van den Broeck et al, Lecture Notes in Physics:
362: ``\bf{Statistical Mechanics of Biocomplexity}'', Vol{527},
363: (Springer-Verlag Berlin, Heidelberg, New York), (1999), Page 93;
364: P. Hanggi and R. Bartussek, \emph{Nonlinear Physics of Complex
365: Systems - Current Status and Future Trends}, Lecture Notes in Physics,
366: Vol. 476, ed. by J. Parisi, S.C. Mueller, and W. Zimmermann (Springer,
367: Berlin, 1996), pp. 294-308.
368:
369: \bibitem{Flash2}
370: A. Ajdari and J. Prost, Europhys. Lett. {\bf32}, (1995), 373.
371:
372:
373: \bibitem{Diffu}
374: P. Reimann, R. Bartussek, R. H\"{a}uss ler and P. H\"{a}nggi,
375: Phys. Lett. A, {\bf215}, (1996), 26.
376:
377: \bibitem{Corr} C.R. Doering, W. Horsthemke, and J. Riordan, Phys. Rev.
378: Lett. {\bf 72}, 2984, (1994).
379:
380: \bibitem{Ma}A. Ajdari, D. Mukamel, L. Peliti, and J. Prost,
381: J. Phys. (Paris) {\bf 14}, 1551, (1994);
382: M.C. Mahato, and A.M. Jayannavar, Phys. Lett.
383: A{\bf 209}, 21, (1995);
384: D.R. Chialvo, and M.M. Millonas, Phys. Lett.
385: A{\bf 209}, 26, (1995).
386:
387:
388: \bibitem{Mill} M. M. Millonas, Phys. Rev. Lett {\bf 74}, 10 (1995).
389:
390:
391: \bibitem{Jay} A. M. Jayannavar, Phys. Rev. E {\bf 53}, 2957 (1996).
392:
393: \bibitem{IntMod} M. C. Mahato, T. P. Pareek and A. M. Jayannavar, Int. J. Mod. Phys
394: B {\bf 10}, 3857 (1996). Cond-Mat/9603103.
395:
396: \bibitem{But} M. B\"{u}ttiker, Z. Phys. B {\bf 68}, 161 (1987).
397:
398: \bibitem{Kam} N. G. van Kampen, IBM. J. Res. Develop {\bf 32}, 107 (1988).
399:
400: \bibitem{Spc_mob} Rolf H. Luchsinger, Phys. Rev. E, {\bf62}, (2000) 272.
401:
402: \bibitem{Dan3} D. Dan, M. C. Mahato and A. M. Jayannavar,
403: Int. J. Mod. Phys. B, (in press);
404: Cond-Mat, (0006244)
405:
406: \bibitem{Kep} C. Kettner et al,
407: Phys. Rev. E, {\bf61}, (2000), 312.
408:
409: \bibitem{PLA} D. Dan, M. C. Mahato and A. M. Jayannavar,
410: Phys. Lett. A {\bf 258}, 217 (1999).
411:
412: \bibitem{PRE} D. Dan, M. C. Mahato and A. M. Jayannavar, Phys. Rev. E
413: {\bf 60}, 6421, (1999).
414:
415: \bibitem{Under} Jose L. Mateos,
416: Phys. Rev. Lett, {\bf84}, (2000), 258.
417:
418: \bibitem{Reim}
419: P. Reimann
420: Phys. Rep., {\bf290}, (1997), 149.
421:
422: \bibitem{Bier}
423: M. Bier and R. D. Astumian
424: Phys. Rev. Lett., {\bf32}, (1995), 373.
425:
426: \bibitem{Conf}
427: Luc P. Faucheux and A. J. Libchaberm,
428: Phy. Rev. E, {\bf49}, (1994), 5158.
429:
430: \bibitem{Surf} H. Brenner,
431: Chem. Eng. Sc., {\bf16}, (1962), 242.
432:
433: \bibitem{Falco} C. M. Falco,
434: Am. J. Phys., {\bf44}, (1976), 733.
435:
436: \bibitem{Pram} A. M. Jayannavar and M. C. Mahato, Pramana- J. Phys {\bf 45},
437: 369 (1995).
438:
439:
440: \bibitem{Sancho} J.M. Sancho, M. San Miguel, and D. Duerr, J. Stat. Phys.
441: {\bf 28}, 291, (1982).
442:
443: \bibitem{Risk} H. Risken, {\em The Fokker Planck Equation}
444: (Springer Verlag, Berlin
445: , 1984).
446:
447:
448: \bibitem{Lich} T. E. Dialynas, K. Lindenberg and G. P. Tsironis,
449: Phys. Rev. E, {\bf56}, (1997), 3976.
450:
451: % \bibitem{PLA}
452: % \Name{D. Dan, M. C. Mahato \and A. M. Jayannavar}
453: % \REVIEW{Phys. Lett. A}{258}{1999}{217}.
454:
455: % \bibitem{PRE}
456: % \Name{D. Dan, M. C. Mahato \and A. M. Jayannavar}
457: % \REVIEW{Phys. Rev. E}{60}{1999}{6421}.
458:
459: \end{thebibliography}
460:
461: \newpage
462:
463: \begin{figure}
464: \centerline{\epsfbox{fig1.eps}}
465: \vspace{.0in}
466: \caption{The mean current $j$ \textit{vs} temperature $T$ for $\phi =
467: 0.2 \pi$, $A=0.5$ and $\lambda = 0.1$. The driving frequencies are $w=
468: 3.0, 4.0$ and $5.0$. The inset shows the contour of zero current for
469: same values of $T, \lambda$. Regions enclosed on the right hand side
470: of a given contour is the negative current region and vice
471: versa. Note that we have only one reversal for all values of $\phi >
472: \pi$.}
473: \label{j-T}
474: \end{figure}
475:
476: \begin{figure}
477: \centerline{\epsfbox{w_c.eps}}
478: \caption{The critical $w$ above which no current reversal with
479: $T$ occurs \textit{vs} $\phi$ for $A=0.5$ and $\lambda=0.1$.}
480: \label{wc_phi}
481: \end{figure}
482:
483: \begin{figure}
484: \centerline{\epsfbox{prob.eps}}
485: \vspace{.0in}
486: \caption{The time averaged probability $P_{av}$ \textit{vs} x for
487: various values of $w$ and $T = 0.05$ and $A=0.5$. The
488: distribution is independent of $\phi$ and $\lambda$.}
489: \label{P_av}
490: \end{figure}
491:
492: \begin{figure}
493: \centerline{\epsfbox{3D.ps}}
494: \vspace{-1.1cm}
495: \caption{The mean current $j$ \textit{vs} $w$ and $T$ for $A = 0.5,
496: \phi=0.2 \pi$ and $\lambda = 0.1$. The contour on the base shows the
497: zero current.}
498: \label{j-T-w}
499: \end{figure}
500:
501:
502: \begin{figure}
503: \centerline{\epsfbox{fig22.eps}}
504: \vspace{.0in}
505: \caption{The mean current $j$ with amplitude $A$ of the forcing term
506: for $\phi = 0.88 \pi,T=0.05$ and $\lambda=0.1$ with $w=3.0,
507: 4.0, 5.0$. The inset shows the contour of zero current for $T=0.05$
508: for $w = 4.0$. The dots in the inset shows the four values of
509: $A$ for $\phi = 0.9 \pi$ where the current reversal occurs. }
510: \label{j-A}
511: \end{figure}
512:
513: \begin{figure}
514: \centerline{\epsfbox{deter.ps}}
515: \vspace{.0in}
516: \caption{The deterministic current $j$ \textit{vs} amplitude of
517: forcing $A$ for $w=0.25$ and $\phi=0.2 \pi$ and $\lambda = 0.1$.}
518: \label{Deter}
519: \end{figure}
520:
521: %\begin{figure}
522: % \onefigure{3D.eps}
523: % \vspace{0.0in}
524: % \caption{}
525: % \label{Deter}
526: %\end{figure}
527:
528:
529: \end{document}
530:
531: