EUROPHYSICS LETTERS

15 December 2006

Europhys. Lett., 76 (6), pp. 11091115 (2006)

DOI: 10.1209/epl/i2006-10391-2

The eect of mechanical deformation on spiral turbulence

H. Zhang 1 , B.-W. Li 1 ( ), Z.-M. Sheng 1 , Z. Cao 2 and G. Hu 3,2 ( ) Zhejiang Institute of Modern Physics and Department of Physics Zhejiang University - Hangzhou 310027, PRC 2 Department of Physics, and The Beijing-Hong Kong-Singapore Joint Centre for Nonlinear and Complex Systems (Beijing), Beijing Normal University Beijing 100875, PRC 3 Chinese Center for Advanced Science and Technology (World Laboratory) Beijing 8730, PRC received 28 June 2006; accepted in nal form 23 October 2006 published online 23 November 2006
PACS. 47.54.-r Pattern selection; pattern formation.
1

Abstract. The behavior of spiral turbulence in mechanically deformed excitable media is investigated. Numerical simulations show that when the forcing frequency is chosen around the characteristic frequency of the system, complicated spiral turbulence may be quenched within a shorter evolution time, compared to the case free of mechanical deformation. It is shown that the observed phenomenon occurs due to enhancing the drift of spiral tips induced by mechanical deformation.

A wide range of self-organized phenomena exists in spatially distributed systems. One of the most paradigmatic examples of spatiotemporal self-organized structures is a class of spiral waves, which have been observed in a variety of physical [1], chemical [25] and biological systems [69]. The attractiveness of investigating the dynamics of spiral waves is not only because they own a special structure, i.e., the core regarded as a phase-singular in mathematical language, but more important they contribute to the underlying class of cardiac disease also, such as tachycardia and brillation [69]. It is thus that, from a practical point of view, controlling the behaviors of spiral waves, particularly eliminating spiral waves and spiral turbulence, is extremely important and meaningful. Up to now, various methods to control two-, and even three-dimensional spiral waves and turbulence have been put forward [1019]. On the other hand, Munuzuri et al. [20] designed an appropriate elastic excitable medium by incorporating the BZ reaction into a polyacrylamide-silica gel and experimentally studied the inuence of periodic mechanical contraction of excitable media on the behavior of rotating spiral waves. Recently, we derived, directly from origin reaction-diusion equations, a formula of drifting velocity using weak-deformation approximation and explained the drift of spiral waves under resonant frequency [21]. Besides, spiral breakup due to mechanical deformation was studied in our recent work [22]. To our knowledge, however, the inuences of mechanical
( ) ( ) E-mail: bwelllee@zju.edu.cn E-mail: ganghu@bnu.edu.cn.

c EDP Sciences

1110

EUROPHYSICS LETTERS

deformation on complex spiral turbulence have not been studied. Naturally, a question arises: what is the destination of spiral turbulence when the medium is deformed periodically? In this paper, we will study the eects of mechanical deformation on spiral turbulence. It will be shown that when the frequency of mechanical deformation is chosen in the vicinity of the characteristic frequency of the system, spiral turbulence can be eliminated during a short evolution time. Let us begin with a modied FitzHugh-Nagumo model with two elds, e, g, and piecewise linear Pushchino kinetics [23]: e = 2 e f (e) g, t g = (e, g)(ke g). t (1)

Here f (e) = C1 e when e < e1 ; f (e) = C2 e + a when e1 e e2 ; f (e) = C3 (e 1) when e > e2 with e1 = 0.0026, e2 = 0.837, C1 = 20, C2 = 3, C3 = 15, a = 0.06 and k = 3, respectively. (e, g) = 1 , if e < e2 ; (e, g) = 2 , if e > e2 ; (e, g) = 3 , if e < e1 and g < g1 , with 1 = 1/75, 2 = 1.0, 3 = 0.3, and g1 = 1.8. To mechanical deformation, it is modeled typically in the following manner: any xed point r is changed to r (t) due to its deformation. We consider here only the one direction (xdirection) and the case of a simple oscillation x = x[1+A cos(t+)], so eqs. (1) become [22]: e 1 = exx + eyy f (e) g, t [1 + A cos(t + )]2 g = (e, g)(ke g). t

(2)

One can see that eqs. (2) are identical with the model of mechanical deformation in ref. [20]: the contraction of the medium is modeled by varying the size of the grid in the x-direction in the numerical simulation of eqs. (1). Compared to eqs. (1), one can see that the mechanical deformation in the x-direction leads to a time-dependent simple anisotropy in the diusion coecient Dx from a mathematical point of view. To integrate eqs. (2), the explicit Euler scheme is employed with time step t = 0.02 and space step x = 0.5. Firstly, we study the inuences of deformation on a certain spiral turbulence initialized from a breakup process of a single spiral without any deformation [23, 24]. As shown in [25], the spiral turbulence for this model is a long-lived transient whose lifespan increases rapidly with L, the linear size of the system. We have tested one initial turbulence state with L = 128 (g. 1(a)) of the system (1) and found that it will not evolve into the rest state when the evolution time is less than 5000. One of the qualitative eects of mechanical deformation on spiral turbulence is illustrated in g. 1. For values of the forcing frequency, , far smaller ( = 0.100, g. 1(b)) or far greater ( = 0.500, g. 1(c)) than the characteristic frequency ( 0.260, see g. 3(c)) of the system (1), the initial spiral turbulence (g. 1(a)) changes little after a period of time. To the contrary, spiral turbulence may be suppressed when the forcing frequency ( = 0.250, g. 1(d)) is chosen close to the characteristic frequency of the system (1). However, the behavior of spiral turbulence subjected to periodic mechanical deformation with a frequency around the characteristic frequency is very complicated, and this can be seen clearly from g. 2(a) showing the dependence of the inverse of lifetime, 1/ , on the forcing frequency. The lifetime is dened as the time of elimination of spiral turbulence after applying the forcing. 1/ = 0 means that the spiral turbulence is not eliminated. In g. 2(a), for the same initial condition (g. 1(a)), we keep the forcing amplitude at A = 0.25 and vary the values of forcing frequencies in the range [0.210, 0.295] with steps of 0.005. One can see

H. Zhang et al.: The effect of mechanical deformation etc.

1111

Fig. 1 (a) Initial condition. (b-d) States evolved from the initial condition in the presence of mechanical deformation with varied forcing frequencies at A = 0.25 and = /2. (b) = 0.100, t = 2000. (c) = 0.500, t = 2000. (d) = 0.250, t = 1600. The characteristic frequency of spiral turbulence 0.260, see g. 3(c). The system size is 128 128 with grid 256 256 points, and t = 0.02. No-ux boundary conditions are imposed.

that 1/ is sensitive to the forcing frequency . For example, the spatiotemporal chaos is eliminated within t = 1500 when the forcing frequency is chosen at = 0.250, while it cannot be suppressed for the chosen frequency = 0.255 even after t = 3000. Such irregularity or sensitivity is also observed clearly in g. 2(b), where the forcing frequency is xed at = 0.250 and the amplitude A is changed in a range [0.13, 0.27] with steps of 0.01, and in g. 2(c), where the forcing frequency and amplitude are kept constant and the initial conditions are
0.0008 0.0006 0.0004 0.0002 0.0000 0.21 0.003 0.22 0.23 0.24 0.25

1/

0.26

0.27

0.28

0.29

0.30

1/

0.002 0.001 0.000 0.12 0.003 0.002 0.14 0.16 0.18 0.20 0.22 0.24 0.26

0.28

1/

0.001 0.000 0 50 100 150 200 250 300 350 400 450 500

Ts
Fig. 2 The dependence of 1/ on A, , and initial conditions. The initial conditions of (a) and (b) are the same. (a) We keep A = 0.25 and vary the frequency in the range [0.210, 0.295] with a 0.005 step. (b) We keep the frequency = 0.250 and vary the forcing amplitude A from 0.13 to 0.27 with a 0.01 step. (c) We keep the forcing amplitude and frequency at A = 0.25, = 0.250, and vary initial conditions. Ts represents the time when we switch on periodic forcing in the course of the evolution of system (1). 1/ is set as zero if spatiotemporal turbulence is not eliminated within t = 3000. The other parameters are the same as g. 1.

1112
0.6 0.5 0.4 0.3 0.2 0.1 0.0 0.16 0.18 0.20 0.22 0.24 0.26 0.28 0.30 0.32 16 14 12 10 8 6 0.16 0.18 0.20 0.22 0.24 0.26 0.28 0.30 0.32 200

EUROPHYSICS LETTERS

()

<n>

power spectra

150 100 50 0 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8

2.0

Fig. 3 (a) The distribution of the probable successful control (). We dene that the spiral turbulence is controlled successfully if it can be suppressed within t = 2000. (b) The average number of spiral tips as a function of frequency. One hundred initial conditions with system size 128 128 are used in (a) and (b). (c) The power spectra (for a variable e) calculated from 1000 points of the system (1) free of mechanical deformation. One can see that the characteristic frequency of spiral turbulence 0.260. The values of the two vertical lines shown in (a), (b) and (c) are equal, representing 0.21 (left) and 0.28 (right), respectively.

changed by varying the time when periodic forcing is switched on. Ts represents the time when we switch on periodic forcing in the course of evolution of system (1), and it is thus that dierent Ts means dierent initial conditions in our simulations. As discussed above, the dynamics of spiral turbulence in the periodically mechanical deformed media is so complicated that we do not tend to study the eects of mechanical deformation on one certain initial condition. Instead, we study the distribution of probable successful control () on large numbers of initial conditions. The larger the values of , the more probably the rest state appears. For instance: for N dierent initial turbulence states, if n of them can be successfully suppressed at , () will be equal to n/N . One hundred initial conditions are used to calculate (). By changing the time when the external periodic force is applied to the system (1), we can obtain large numbers of initial conditions. The distribution of the probable successful control () is shown in g. 3(a). One can see that the principal contribution of () is mainly constrained to the region between the two vertical lines, representing = 0.21 (left) and = 0.28 (right) in g. 3(a). Furthermore, we also calculate the average number of spiral tips as a function of frequency from one hundred initial conditions (see g. 3(b)) and nd it consistent with g. 3(a). Figure 3(c) gives the power spectrum (for a variable e) calculated from 1000 points of the system (1) and the principal value also approximatively falls into the range from 0.21 to 0.28, indicated by left and right vertical lines. The agreement of the qualitative features of the distribution of the probable successful control, average tip number as well as the power spectrum permit us to draw a conclusion that the result observed here is a kind of single-frequency resonance behavior, which has been observed in the resonant drift of rotating spirals for equal frequencies of mechanical deformation and spiral rotation [20, 21].

H. Zhang et al.: The effect of mechanical deformation etc.

1113

Fig. 4 The trajectories of spiral tips motion for two cases from the same initial condition during the same evolution time t = 70. (a) Free of mechanical deformation; (b) deformed with A = 0.50, = 0.230. The system size is 200 200 with grid 400 400 points, and t = 0.02.

The mechanism, underlying the result of spiral turbulence being quenched due to external mechanical deformation around the characteristic frequency of the system, can be interpreted intuitively in the following way: spiral turbulence dominating the whole excitable medium consists of a variety of small spirals; when the excitable medium is subjected to an external periodic mechanical deformation, the drift of spiral tips will be enhanced, which is shown in g. 4. Generally, spiral tips drift randomly and are constrained to a limited region if the system is free of mechanical deformation (see g. 4(a)). However, when the system is deformed periodically with a forcing frequency close to the characteristic frequency of the system, the tips are no longer conned to a local region and instead the drift of tips is enhanced (see g. 4(b)). As a result, some of them will drift towards the boundary of the system and disappear through it, and some may collide with each other because of the enhancement of the drift. When two tips with dierent topological charges collide, they may be annihilated and the reconnection of waves will happen, which can be seen from g. 5. Figure 5(a) is an initial spiral-turbulence state consisting of a large number of small spirals. In general, this state evolves in time with

Fig. 5 Typical snapshots with unequal time intervals during the evolution of the system (2) within t = 2000. The parameters are the same as g. 4.

1114

EUROPHYSICS LETTERS

Fig. 6 Typical evolution snapshots under the deformed Br model with A = 0.27, = 0.120 and a = /2. (a) t = 0, (b) t = 2420; (c) t = 2422; (d) t = 2430. Other parameters are: a = 0.84, b = 0.07 and = 0.085. With these parameters, the characteristic frequency of the turbulence state is about 0 = 1.10. The initial turbulence state is obtained from the break-up of a single spiral. Grid 128 128 points, x = 0.35 and t = 0.01.

similar congurations and will not vary dramatically during the evolution of the system (1). Contrarily, the case will be changed if the medium is deformed around the characteristic frequency. As time elapses, fragments will reconnect due to the interaction of tips and form larger ones (gs. 5(b)-(j)). Ultimately, the complexity of the spatiotemporal dynamics is reduced and even eliminated (gs. 5(k)-(l)). The drift and interaction of spiral tips play a key role for the overall dynamics of spiral turbulence. Though the evolution process of the whole dynamics is very complicated, the spiral turbulence may disappear nally if the condition is suitable. To check whether our strategy presented here is sensitively model dependent, we also study the inuence of mechanical deformation on spiral turbulence in the Br model [26]: a 1 v+b u = u(1 u) u t a + 2 u, v = f (u) v, t (3)

where f (u) = 0, if 0 u < 1/3; f (u) = 1 6.75u(u 1)2 , if 1/3 u 1; and f (u) = 1, if u > 1. For > 0.07 with xed a = 0.84 and b = 0.07, a single spiral will break up due to Doppler eect and the system will quickly fall into a turbulence state. Numerical simulations of the Br model show that termination of spiral turbulence does occur when the chosen a frequency is close to the characteristic frequency (0 = 1.10) of system (3) (see g. 6). We also nd that spiral turbulence can be suppressed (not shown) when we periodically modulate other parameters, for example a of the modied FitzHugh-Nagumo model (1). The observed phenomenon is also a resonance-type behavior in frequency. The drift of spiral tips is enhanced by modulating a when the forcing frequency is chosen in the neighborhood of the characteristic frequency of the system. In conclusion, the eect of mechanical deformation on spiral turbulence has been studied. We have shown that the complicated spatiotemporal chaos may be suppressed when the excitable medium is subjected to mechanical deformation. Although the results are dependent on and sensitive to the parameters of mechanical deformation and even initial conditions, a large amount of statistic data indicate that the frequencies of mechanical deformation for successful control are mostly around the characteristic frequency of the system. The agreement of distribution of probable successful control (), average number of spiral tips as well as power spectra of the system (1) permits us to conclude that the observed phenomenon is a resonance-type behavior in frequency. At last, we hope that our results will be observed in experiments. By changing the concentrations of the BZ reaction, the spiral waves may break up and spiral turbulence will emerge in the elastic excitable medium [20, 27]. Then one can study the eect of mechanical deformation on spiral turbulence experimentally.

H. Zhang et al.: The effect of mechanical deformation etc.

1115

This work was supported by the National Nature Science Foundation of China. We would like to thank Ningjie Wu for useful discussions.
REFERENCES [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] Cross M. C. and Hohenberg P. C., Rev. Mod. Phys., 65 (1993) 851. Winfree A. T., Science, 175 (1972) 634. Ouyang Q. and Flesselles J.-M., Nature (London), 379 (1996) 143. Vanag V. K. and Epstein I. R., Science, 294 (2001) 835. Jakubith S., Rotermund H. H., Engel W., von Oertzen A. and Ertl G., Phys. Rev. Lett., 65 (1990) 3013. Davidenko J. M., Pertsov A. V., Salomonsz R., Baxter W. and Jalife J., Nature (London), 355 (1992) 349. Witkowski F. X., Leon L. J., Penkoske P. A., Giles W. R., Spano M. L., Ditto W. L. and Winfree A. T., Nature (London), 392 (1998) 78. Winfree A. T., When Time Breaks Down (Princeton University Press, Princeton, NJ) 1987. Focus Issue: Fibrillation in normal ventricular myocardium, Chaos, 8 (1998) 1. Aranson I., Levine H. and Tsimring L., Phys. Rev. Lett., 72 (1994) 2561. Baier G., Sahle S. and Chen J., J. Chem. Phys., 110 (1999) 3251. Osipov G. V. and Collins J. J., Phys. Rev. E, 60 (1999) 54. Panfilov A. V., Muller S. C., Zykov V. S. and Keener J. P., Phys. Rev. E, 61 (2000) 4644. Kim M., Bertram M., Pollmann M., von Oertzen A., Mikhailov A. S., Rotermund H. H. and Ertl G., Science, 292 (2001) 1357. Stamp A. T., Osipov G. V. and Collins J. J., Chaos, 12 (2002) 931. Zhang H., Hu B. and Hu G., Phys. Rev. E, 68 (2003) 026134; Fu Y. Q., Zhang H., Cao Z., Zheng B. and Hu G., Phys. Rev. E, 72 (2005) 046206. Alonso S., Sagues F. and Mikhailov A. S., Science, 299 (2003) 1722; Chaos, 16 (2006) 023124. Zhang H., Cao Z., Wu N. J., Ying H. P. and Hu G., Phys. Rev. Lett., 94 (2005) 188301. For a recent review, see Mikhailov A. S. and Showalter K., Phys. Rep., 425 (2006) 79. Munuzuri A. P., Innocenti C., Flesselles J.-M., Gilli J.-M., Agladze K. I. and Krinsky V. I., Phys. Rev. E, 50 (1994) R667. Zhang H., Hu B., Hu G. and Xiao J., J. Chem. Phys., 119 (2003) 4468. Zhang H., Ruan X. S., Hu B. and Ouyang Q., Phys. Rev. E, 70 (2004) 016212. Panfilov A. V. and Hogeweg P., Phys. Lett. A, 176 (1993) 295. Panfilov A. V., Chaos, 8 (1998) 57. Sinha S., Pande A. and Pandit R., Phys. Rev. Lett., 86 (2001) 3678. Bar M. and Eiswirth M., Phys. Rev. E, 48 (1993) R1635; Hildebrand M., Bar M. and Eiswirth M., Phys. Rev. Lett., 75 (1995) 1503. Ouyang Q., Swinney H. L. and Li G., Phys. Rev. Lett., 84 (2000) 1047.