1: % version of 11/14/06

2: 

3: % Choose pra, prb, prc, prd, pre, prl, prstab, or rmp for journal

4: %  Add 'draft' option to mark overfull boxes with black boxes

5: %  Add 'showpacs' option to make PACS codes appear

6: %  Add 'showkeys' option to make keywords appear

7: %\documentclass[aps,prl,preprint,showpacs,groupedaddress]{revtex4}

8: %\documentclass[aps,prb,preprint,showpacs,superscriptaddress]{revtex4}

9: \documentclass[aps,prl,twocolumn,showpacs,groupedaddress]{revtex4}

10: 

11: \bibliographystyle{apsrev}

12: 

13: \usepackage{graphicx}

14: \usepackage{dcolumn}

15: \usepackage{bm}

16: \usepackage{amssymb}

17: 

18: \begin{document}

19: 

20: %\preprint{}

21: 

22: %Title of paper

23: \title{Swinging and Tumbling of Fluid Vesicles in Shear Flow}

24: 

25: \author{Hiroshi Noguchi}

26: \email[]{e-mail: hi.noguchi@fz-juelich.de}

27: \author{Gerhard Gompper}

28: %\email[]{e-mail: g.gompper@fz-juelich.de}

29: %\homepage[]{Your web page}

30: %\altaffiliation{}

31: \affiliation{

32: Institut f\"ur Festk\"orperforschung, Forschungszentrum J\"ulich, 

33: 52425 J\"ulich, Germany}

34: 

35: \date{\today}

36: 

37: \begin{abstract}

38: The dynamics of fluid vesicles in simple shear flow is studied

39: using mesoscale simulations of dynamically-triangulated surfaces, as well

40: as a theoretical approach based on two variables, a shape parameter and 

41: the inclination angle, which has no adjustable parameters. We show 

42: that between the well-known tank-treading and tumbling states,  

43: a new ``swinging'' state can appear. We predict the dynamic phase diagram 

44: as a function of the shear rate, the viscosities of the 

45: membrane and the internal fluid, and the reduced vesicle volume.

46: Our results agree well with recent experiments.

47: \end{abstract}

48: %\pacs{87.16.Dg Membranes, bilayers, and vesicles --

49: %      83.50.-v Deformation and flow

50: %      87.17.Aa Theory and modeling; computer simulation

51: %     }

52: \pacs{87.16.Dg,83.50.-v,87.17.Aa}

53: %\keywords{}

54: 

55: \maketitle

56: 

57: The dynamics of deformable objects such as liquid droplets~\cite{rall84,bart85}, 

58: lipid vesicles~\cite{haas97,kant06,seif99,misb06,krau96,beau04,nogu04,nogu05}, 

59: red blood cells~\cite{fisc78,kell82,tran84,pozr03}, 

60: and elastic capsules~\cite{chan93,skot06,fink06} in flows has received 

61: increasing attention experimentally, theoretically, and numerically in recent 

62: years.  All of these objects show phenomenologically similar behaviors in 

63: shear flows --- either a tank-treading rotation with a stationary 

64: shape and a finite inclination angle $\theta>0$ with respect to the flow 

65: direction, or an unsteady tumbling motion.  However, the qualitative and 

66: quantitative behavior of the various objects can be very different, 

67: because the energies governing their shapes and thermal fluctuations are

68: very different.  In the case of fluid vesicles, which we want to investigate

69: in this letter, the membrane conformations are determined by the 

70: curvature elasticity together with the constraints of membrane 

71: incompressibility and constant internal volume.

72:  

73: In simple shear flows, with flow velocity ${\bf v}=\dot\gamma y {\bf e}_x$, 

74: a transition occurs from tank-treading to tumbling with an increasing viscosity 

75: of the internal fluid $\eta_{\rm {in}}$~\cite{kant06,beau04,kell82} 

76: or membrane viscosity $\eta_{\rm {mb}}$~\cite{nogu04,nogu05}.

77: This transition 

78: is described very well by the theory of Keller and Skalak 

79: (K-S)~\cite{kell82}, which assumes a fixed ellipsoidal vesicle shape.

80: However, shear forces can be large enough to induce shape transformations of

81: fluid vesicles, for example from discocyte to prolate at small membrane 

82: viscosity and small viscosity contrast between inside and outside, or 

83: from prolate to discocyte at larger membrane viscosities \cite{nogu04,nogu05}.

84: In this case, it is essential to take the membrane deformability into 

85: account.

86: 

87: Recently, Kantsler and Steinberg~\cite{kant06} reported the first observation

88: of a new type of vesicle dynamics in shear flow, which is characterized by  

89: small oscillations of the inclination angle $\theta$ and the deformation,

90: where $-\theta_0\lesssim \theta(t) \lesssim \theta_0$ with $\theta_0/\pi <1$ 

91: and time average $\langle\theta\rangle\simeq 0$. 

92: The vesicles were found to transit from tumbling to this oscillatory motion 

93: with increasing shear rate $\dot\gamma$. It is worth mentioning that such

94: an oscillatory motion was also observed in our previous simulations 

95: (see Fig. 6 in Ref.~\onlinecite{nogu05}), but not further analyzed.

96: Simultaneously with the experiment, Misbah~\cite{misb06} predicted 

97: a ``vacillating breathing" mode for quasi-spherical fluid vesicles. 

98: This mode exhibits very similar dynamical behavior as seen experimentally; 

99: however, it ``coexists" with the tumbling mode, and its orbit depends on 

100: the initial deformation i.e. it is not a limit cycle. Furthermore, the shear

101: rate only enters the theory as the basic time scale, and therefore cannot

102: induce any shape transitions.

103: Hence it does not explain the tumbling-to-oscillatory transition seen in 

104: the experiments \cite{kant06}.

105: 

106: In this letter, we study the oscillatory dynamics of fluid vesicles, which

107: we will refer to as the ``swinging mode'', by mesoscale hydrodynamics 

108: simulations and a simplified non-linear theoretical model. The main questions

109: we want to address are: How does the bending energy affect the 

110: dynamics? Can transitions between modes be induced by varying the shear rate?

111: What happens beyond the quasi-spherical limit, which is the typical 

112: experimental situation? What is the effect of the membrane viscosity?

113: Is the swinging mode stable when thermal membrane 

114: undulations are taken into account? We will show that  

115: the experiments of Ref.~\onlinecite{kant06} can be understood very well on 

116: the basis of our theory. Furthermore, we will present a complete phase 

117: diagram of vesicle motion as a function of shear rate and viscosity.

118: 

119: It is worth mentioning that elastic capsules~\cite{chan93} 

120: and red blood cells~\cite{vall06} can also exhibit a swinging mode; however,

121: in this case,  the angle $\theta(t)$ is always positive during the oscillation.

122: Very recently, this dynamics was explained by the K-S theory with an 

123: addition of an energy barrier for the tank-treading rotation caused by the 

124: membrane shear elasticity~\cite{skot06}. Therefore, this mechanism cannot

125: be employed to explain the swinging mode of fluid vesicles. 

126: 

127: The vesicle dynamics is described by several dimensionless quantities.

128: For a vesicle of volume $V$ and surface area $S$,  

129: the reduced volume $V^*$ and the excess area $\Delta_{\rm S}$ are 

130: defined by

131: $V^*= (R_{\rm V}/R_{\rm S})^3= (1+\Delta_{\rm S}/4\pi)^{-3/2}$ and

132: $\Delta_{\rm S}=S/R_{\rm V}^2 -4\pi$,

133: where $R_{\rm V}= (3V/4\pi)^{1/3}$ and $R_{\rm S}=(S/4\pi)^{1/2}$.

134: The relative viscosity of the inside fluid and membrane are 

135: $\eta_{\rm {in}}^*=\eta_{\rm {in}}/\eta_{\rm 0}$

136: and $\eta_{\rm {mb}}^*=\eta_{\rm {mb}}/\eta_{\rm 0}R_{\rm S}$, where 

137: $\eta_{\rm 0}$ is the viscosity of the outside fluid. 

138: The shape relaxation time of vesicles with bending rigidity $\kappa$ is given

139: by $\tau=\eta_{\rm 0}R_{\rm S}^3/\kappa$ (for $\eta_{\rm {in}}^*=1$). This

140: time is used to define a reduced shear rate $\dot\gamma^*=\dot\gamma \tau$.

141: 

142: The hydrodynamics of fluid vesicle can be studied very well by hybrid 

143: simulations of a dynamically-triangulated membrane model~\cite{gg:gomp97f} and 

144: a particle-based mesoscale solvent, multi-particle collision dynamics

145: \cite{male99}.  A detailed description of this method can be 

146: found in Ref.~\onlinecite{nogu05}.

147: 

148: We have simulated a prolate vesicle with a viscous membrane at $V^*=0.78$ 

149: and $\eta_{\rm {in}}^*=1$. 

150: Figure~\ref{fig:rq} shows the time development of the vesicle shape and 

151: $\theta$.  The shape parameter is $\alpha_{\rm D}=(L_1-L_2)/(L_1+L_2)$, 

152: where $L_1$ and $L_2$ are the maximum lengths in the direction

153: of the eigenvectors of the gyration tensor in the vorticity ($x,y$) plane.

154: The vesicle is found to exhibit a tumbling motion for $\dot\gamma^*=0.92$,

155: but a swinging motion for $\dot\gamma^*=3.68$.  As $\dot\gamma^*$ increases,

156: the tumbling frequency $f_{\rm {tmb}}$ decreases, see Fig.~\ref{fig:ftum}, 

157: while the frequency of the swinging motion increases.

158: The shear rate $\dot\gamma^*$ at the tumbling-to-swinging

159: transition is found to increase with increasing membrane viscosity 

160: $\eta_{\rm {mb}}^*$. These results are consistent with the experiments

161: of Ref.~\onlinecite{kant06}.

162: 

163: \begin{figure}

164: \includegraphics{fig1.eps}

165: \caption{ \label{fig:rq}

166: (Color online)

167: Temporal evolution of $\alpha_{\rm D}$ and inclination angle 

168: $\theta$, for $V^*=0.78$ and $\eta_{\rm {mb}}^*=2.9$.

169: The solid (red) and dashed (blue) lines represent simulation data

170: for $\dot\gamma^*=3.68$ and $0.92$ ($\kappa/k_{\rm B}T=10$ and $40$ with 

171: $\dot\gamma \eta_{\rm 0}R_{\rm S}^3/k_{\rm B}T=36.8$), respectively.

172: The solid lines in (b) are obtained from Eqs.~(\ref{eq:ald}),  

173: (\ref{eq:thetks}) without thermal noise for $\dot\gamma^*=1.8$, $3$ and 

174: $10$ (from top to bottom).

175: }

176: \end{figure}

177: 

178: \begin{figure}

179: \includegraphics{fig2.eps}

180: \caption{ \label{fig:ftum}

181: (Color online)

182: $\dot\gamma$ dependence of

183: the tumbling frequency $f_{\rm {tum}}$ normalized by 

184: the frequency $f_{\rm {tum}}^{\rm {KS}}=\dot\gamma\sqrt{1-B^2}/2\pi$ of 

185: K-S theory at $V^*=0.78$.

186: The bending rigidity $\kappa$ is varied to change $\dot\gamma^*$

187: with $\dot\gamma \eta_{\rm 0}R_{\rm S}^3/k_{\rm B}T=36.8$.

188: The symbols represent simulation data for $\eta_{\rm {mb}}^*=2.9$ ($\bullet$) 

189: and $4.1$ ($\blacksquare$). The solid and dashed lines are obtained from 

190: Eqs.~(\ref{eq:ald}), (\ref{eq:thetks}) with and without thermal noise, 

191: respectively.

192: }

193: \end{figure}

194: 

195: In order to obtain a better understanding of the vesicle dynamics, we

196: now derive an approximate theoretical model. First, we follow 

197: Refs.~\onlinecite{seif99,misb06} and employ the Stokes approximation 

198: and perturbation theory for quasi-spherical vesicles.

199: The vesicle shape is expanded in spherical harmonics $Y_{l,m}$ as 

200: $R= R_{\rm V}(1+\sum_{l,m}u_{l,m}Y_{l,m})$. 

201: The free energy of the membrane with bending rigidity $\kappa$ and surface 

202: tension $\sigma$ is $F =  \int dS \{\sigma + (\kappa/2)(C_1+C_2)^2 \}$,

203: where $C_1$ and $C_2$ are the 

204: principal curvatures at each point of the membrane.

205: With the harmonic approximation, it is given by

206: $F = (\kappa/2)\sum_{l,m} E_l |u_{l,m}|^2 + O({u_{l,m}}^3)$,

207: where $E_l=(l+2)(l-1)\{l(l+1)+ \sigma {R_{\rm V}}^2/\kappa\}$.

208: The flow fields inside and outside of the vesicle are described by the Lamb 

209: solution~\cite{seif99}, with no-slip boundary conditions on the membrane.

210: The flow stress on the membrane is balanced with the elastic forces due 

211: to bending and tension. This implies that the undulation amplitudes 

212: $u_{l,m}$ are determined by

213: \begin{equation}

214: \label{eq:lamb}

215: \frac{\partial u_{l,m}}{\partial t} = \frac{i \dot\gamma m}{2} u_{l,m}- 

216: \frac{\kappa\Gamma_l E_l}{2\eta_0 R_{\rm V}^3}{u_{l,m}}

217: \mp i h\dot\gamma \delta_{l,2}\delta_{m,\pm 2},

218: \end{equation}

219: where $h=60\sqrt{2\pi/15}/(32+23\eta_{\rm {in}}^*)$ and

220: $\Gamma_l = l(l+1)/\{(l+2)(2l^2-l+2)+(l-1)(2l^2+5l+5)\eta_{\rm {in}}^*\}$.

221: The detailed derivation of Eq.~(\ref{eq:lamb}) for

222: $\eta_{\rm {in}}^*=1$ is described in Ref.~\onlinecite{seif99}.

223: 

224: In this letter, we focus on vesicles, which are symmetric with 

225: respect to the vorticity plane through their center.

226: Hence, we only consider $u_{2,\pm 2},u_{2,0}$, and

227: decompose $u_{2,\pm 2}$ into amplitude and phase, 

228: $u_{2,\pm 2}=r \exp(\mp 2i \theta)$,

229: where $\theta$ corresponds to the inclination angle \cite{misb06}.

230: Since the curvature energy does not contribute in this 

231: case, compare Ref.~\onlinecite{misb06}, we have to go beyond the harmonic 

232: approximation.

233: Therefore, we replace the force $\kappa E_l r$ in Eq.~(\ref{eq:lamb}) 

234: by $\partial F/\partial r$. This implies 

235: \begin{eqnarray}

236: \label{eq:r2f}

237: \frac{\partial r}{\partial t} &=&  -\frac{\Gamma_2}{2\eta_0 R_{\rm V}^3} 

238:    (\frac{\partial F}{\partial r} + 2\lambda r) + \dot\gamma h\sin(2\theta),\\

239: \label{eq:thetlam}

240: \frac{\partial \theta}{\partial t} &=& \frac{\dot\gamma}{2}

241:                 \{ -1 + \frac{h}{r}\cos(2\theta)\},\\

242: \label{eq:u20}

243: \frac{\partial u_{2,0}}{\partial t} &=&  -\frac{\Gamma_2}{\eta_0 R_{\rm V}^3} 

244:          (\frac{\partial F}{\partial u_{2,0}} + \lambda u_{2,0}).

245: \end{eqnarray}

246: A Lagrange multiplier $\lambda\equiv \sigma R_V^2$ is employed to

247: satisfy the area constraint \cite{miln87}; it is determined by

248: $d{\Delta}_{\rm S}/dt=0$, which implies 

249: $\Delta_{\rm S}= 2 {u_{2,0}}^2 + 4 r^2$. Thus, Eq.~(\ref{eq:r2f}) 

250: becomes

251: \begin{equation}

252: \label{eq:r2ds}

253: \frac{d r}{dt} = \left\{1-\frac{4r^2}{\Delta_{\rm S}}\right\}

254:        \left\{ -\frac{\Gamma_2}{2\eta_0 R_{\rm V}^3} 

255:        \frac{\partial F}{\partial r}\Bigl|_{\Delta_{\rm S}} 

256:                       + \dot\gamma h\sin(2\theta)\right\},

257: \end{equation}

258: where $\partial F/\partial r|_{\Delta_{\rm S}} = 

259:         \partial F/\partial r-(2r/u_{2,0})\partial F/\partial u_{2,0}$.

260: In the harmonic approximation of $F$,

261: the first term in Eq.~(\ref{eq:r2ds}) disappears because of 

262: $\partial F/\partial r|_{\Delta_{\rm S}}=0$, so that 

263: we recover the description of Ref.~\onlinecite{misb06}.

264: The prolate shape appears as an energy minimum,

265: when higher-order terms in the free energy $F$ are taken into 

266: account~\cite{miln87}.

267: 

268: \begin{figure}

269: \includegraphics{fig3.eps}

270: \caption{ \label{fig:cv}

271: (Color online)

272: $\dot\gamma$ dependence of the amplitudes of the oscillations of 

273: $\alpha_{\rm D}$ and $\theta$ 

274: for $V^*=0.9$, $\eta_{\rm {in}}^*=6$, and $\eta_{\rm {mb}}^*=0$.

275: The dashed and solid lines represent the tumbling and swinging states, 

276: respectively, given by Eqs.~(\ref{eq:ald}), (\ref{eq:thetks})

277: without thermal noise.  Symbols represent  

278: estimates from the experimental data~\cite{kant06} for

279: $\alpha_{\rm D}$ ($\blacktriangle$) and $\theta$ ($\blacklozenge$).

280: The inset shows the curvature energy $F(\alpha_{\rm D})$ of an ellipsoid

281: for the reduced volumes $V^* = 0.78$, $0.85$, and $0.9$.

282: }

283: \end{figure}

284: 

285: Since an expansion of the vesicle shape in spherical harmonics

286: is difficult experimentally, more easily accessible measures of the 

287: deformation are desirable. Furthermore, a description is needed which

288: goes beyond the quasi-spherical limit. Thus, instead of $r$, we employ 

289: the shape parameter $\alpha_{\rm D}=(L_1-L_2)/(L_1+L_2)$,

290: which is easily measurable by microscopy.

291: Since $\alpha_{\rm D}= (\sqrt{15/2\pi}/2) r + O(r^2)$,

292: Eq.~(\ref{eq:r2ds}) implies

293: \begin{equation}

294: \label{eq:ald}

295: \frac{d \alpha_{\rm D}}{dt} = \left\{1-\left(\frac{\alpha_{\rm D}}

296:                    {\alpha_{\rm D}^{\rm {max}}}  \right)^2\right\}

297:     \left\{ -\frac{A_0}{\tau\kappa V^*} 

298:        \frac{\partial F}{\partial \alpha_{\rm D}} 

299:                 + \dot\gamma A_1\sin(2\theta)\right\},

300: \end{equation}

301: where $A_0= 15\Gamma_2/16\pi= 45/8\pi(32+23\eta_{\rm {in}}^*)$ and 

302: $A_1= h\sqrt{15/2\pi}/2= 30/(32+23\eta_{\rm {in}}^*)$.

303: Since an accurate evaluation of the free energy $F$ is very important, 

304: we calculate it numerically for ellipsoidal vesicles with 

305: $(x_1/a_1)^2 + (x_2/a_2)^2 + (x_3/a_3)^2=1$, see the inset in Fig.~\ref{fig:cv}.

306: The prolate ($a_1>a_2=a_3$) and oblate ($a_1=a_2>a_3$) shapes are energy 

307: minima and maxima, respectively,

308: and $\partial F/\partial \alpha_{\rm D}$ diverges in the limit of maximum

309: extension, $\alpha_{\rm D} \to \alpha_{\rm D}^{\rm {max}}(V^*)$.

310: Eq.~(\ref{eq:ald}) has the same form as the simplified model studied previously 

311: \cite{nogu04,nogu05}, but now has {\em no adjustable parameters}.

312: 

313: In a final step, in order to obtain a reliable description also for large

314: excess areas, we replace Eq.~(\ref{eq:thetlam}) by the equation of K-S 

315: theory~\cite{kell82,tran84}, which reads

316: \begin{eqnarray}

317: \label{eq:thetks}

318: \frac{d \theta}{dt} &=& \frac{\dot\gamma}{2}\{-1+B\cos(2\theta)\},\\

319: \label{eq:KS-B} \nonumber

320: B &=& f_0\left\{f_1+ \frac{f_1^{-1}}

321:       {1+f_2(\eta_{\rm {in}}^* -1) 

322:                   + f_2f_3 \eta_{\rm {mb}}^*}\right\},

323: \end{eqnarray}

324: where the factors $f_0$, $f_1$, $f_2$, and $f_3$ are

325: the functions of the ellipsoidal shape ($a_2/a_1$, $a_3/a_1$),

326: and are given in Appendix B of Ref.~{\onlinecite{nogu05}}.

327: The K-S theory in general shows very good agreement with simulation 

328: results~\cite{krau96,nogu05}.

329: When thermal fluctuations are taken into account,

330: Gaussian white noises $g_{\alpha}(t)$ and $g_{\theta}(t)$

331: are added to Eqs.~(\ref{eq:ald}) and (\ref{eq:thetks}), respectively,

332: which obey the fluctuation-dissipation theorem, so that

333: $\langle g_i(t)\rangle = 0$,

334: $\langle g_i(t) g_j(t')\rangle =

335:       2 k_{\text B}T/\zeta_i \delta_{ij}\delta(t-t')$, 

336: where $i,j \in \{\alpha_{\rm D},\theta\}$.

337: The friction coefficients are $\zeta_{\alpha}=

338:     \tau\kappa V^*/A_0\{1-(\alpha_{\rm D}/\alpha_{\rm D}^{\rm {max}})^2\}$ 

339: and the rotational friction coefficient of a sphere,  

340: $\zeta_{\theta}=8\pi \eta_{\rm {0}}R_{\rm S}^3$.

341: We numerically integrate Eqs.~(\ref{eq:ald}) and (\ref{eq:thetks})

342: with or without thermal noise 

343: using the second or fourth-order Runge-Kutta method, respectively.

344: When $\alpha_{\rm D}(t)$ becomes zero, we set 

345: ($\alpha_{\rm D}$,$\theta$)=($0,\pi/4$), since $\theta=\pi/4$ is 

346: the inclination angle in the spherical limit.

347: 

348: Eqs.~(\ref{eq:ald}) and (\ref{eq:thetks}) reproduce the simulation results 

349: semi-quantitatively (see Figs.~\ref{fig:rq} and \ref{fig:ftum}).

350: This is a very good agreement, given the fact that our model certainly

351: does not systematically take into account all higher order terms, and

352: furthermore has no adjustable parameters.

353: At small $\dot\gamma^*$, in the tumbling phase,

354: $\theta$ rotates with $\alpha_{\rm D}$-oscillation of small amplitude.

355: This $\alpha_{\rm D}$ amplitude becomes larger at larger $\dot\gamma^*$,

356: see Fig.~\ref{fig:cv}.  Then, when $\alpha_{\rm D}(t)$ reaches zero,

357: $\theta$ jumps to $\pi/4$, see Fig.~\ref{fig:rq}.

358: This type of $\theta$ oscillations with a jump is reminiscent  

359: of the behavior predicted for

360: very viscous liquid drops within perturbation theory~\cite{bart85}.

361: Finally, at even larger $\dot\gamma^*$,

362: $\alpha_{\rm D}$ and $\theta$ exhibit oscillations without jumps. 

363: 

364: The physical mechanism of swinging can be understood

365: on the basis of Eqs.~(\ref{eq:ald}), (\ref{eq:thetks}) as follows.

366: At finite $\kappa$, the shear force elongates the vesicle 

367: ($\alpha_{\rm D}$ increases) for $0<\theta<\pi/2$,

368: but compresses it ($\alpha_{\rm D}$ decreases) for $-\pi/2<\theta<0$,

369: since the $\sin(2\theta)$-term in Eq.~(\ref{eq:ald}) changes sign.

370: Thus, the swinging motion is caused 

371: by a shape deformation, where $B$ in Eq.~(\ref{eq:thetks})

372: crosses the tank-treading-to-tumbling threshold periodically.

373: First, a prolate vesicle starts tumbling because $B<1$, 

374: $\alpha_{\rm D}$ decreases when $\theta<0$, which implies that $B$ increases;

375: then $\theta$ starts to increase again because $B>1$ at small 

376: $\alpha_{\rm D}$; finally $\alpha_{\rm D}$ increases when $\theta>0$.

377: In the swinging phase, the amplitudes of $\alpha_{\rm D}$ and $\theta$ 

378: decrease and saturate to finite values with increasing $\dot\gamma^*$, 

379: see Fig.~\ref{fig:cv}.

380: 

381: A linear stability analysis of the fixed points of Eqs.~(\ref{eq:ald}), 

382: (\ref{eq:thetks}) shows that the tank-treading-to-tumbling transition 

383: always occurs as a saddle-node bifurcation at $\theta\simeq 0$.

384: The tank-treading-to-swinging transition is a saddle-node bifurcation for 

385: small $\dot\gamma^*$, but becomes a Hopf bifurcation at $\theta<0$ for larger 

386: $\dot\gamma^*$.  Near the boundary of these two bifurcations,

387: a second range of stable fixed points appears between the saddle and 

388: unstable fixed points at $\theta<0$.

389: However, these stable points at $\theta<0$ will likely disappear

390: when the shape degrees of freedom in the vorticity direction 

391: are taken into account. 

392: 

393: \begin{figure}

394: \includegraphics{fig4.eps}

395: \caption{ \label{fig:phase}

396: (Color online)

397: Dynamical phase diagrams as a function of (a) $\eta_{\rm {mb}}^*$ for 

398: $\eta_{\rm {in}}^*=1$, and (b) $\eta_{\rm {in}}^*$ for $\eta_{\rm {mb}}^*=0$,

399: for various reduced volumes $V^*$, obtained from Eqs.~(\ref{eq:ald}), 

400: (\ref{eq:thetks}) without thermal noise.

401: The tank-treading phase is located on the left-hand-side of the dashed 

402: lines. The solid lines represent the tumbling-to-swinging transitions.

403: }

404: \end{figure}

405: 

406: Fig.~\ref{fig:phase} shows the full phase diagrams, both for variations of

407: the internal and the membrane viscosities.

408: With increasing reduced volume $V^*$ the smallest shear rate 

409: $\dot\gamma^*_{\rm {os}}$, at which swinging can be observed,  

410: decreases, since the dependence of the energy $F$ on $\alpha_{\rm D}$

411: becomes more shallow, compare Fig.~\ref{fig:cv}. Also, the region of

412: stability of swinging shifts to higher viscosity ($\eta_{\rm in}$ or 

413: $\eta_{\rm mb}$) with increasing $V^*$. On the other hand, for fixed $V^*$,

414: $\dot\gamma^*_{\rm {os}}$ is a monotonically increasing function of

415: $\eta_{\rm {mb}}^*$ or $\eta_{\rm {in}}^*$. This increase is found to be

416: much less pronounced for $\eta_{\rm {mb}}^*$. The reason is that in our

417: derivation of Eq.~(\ref{eq:lamb}), we have neglect the $\eta_{\rm {mb}}^*$ 

418: dependence. In analogy with Eq.~(\ref{eq:thetks}), it is plausible to expect 

419: that in a more detailed calculation, a linear combination of $\eta_{\rm {mb}}^*$ 

420: and $\eta_{\rm {in}}^*$ will appear in the both $h$ and $\Gamma_l$.  

421: If this $\eta_{\rm {mb}}^*$ dependence were taken into account,

422: the $\dot\gamma^*_{\rm {os}}$ lines in Fig.~\ref{fig:phase}(a) will curve 

423: upwards more strongly; this is consistent with

424: the larger $\eta_{\rm {mb}}^*$ dependence observed in the simulations in 

425: Fig.~\ref{fig:ftum}.

426: 

427: In the experiments of Ref.~\onlinecite{kant06}, the swinging (tumbling) motion 

428: was observed for $\dot\gamma^*\simeq 17$ ($4.5$), $\eta_{\rm {in}}^*=6$ ($8.4$), 

429: and  $V^*\simeq 0.9$, and a very small membrane viscosity 

430: $\eta_{\rm {mb}}^* \sim 0.1$ (calculated from 

431: $\eta_{\rm {mb}} \sim 10^{-9}$ Ns/m of Ref.~\onlinecite{dimo99}).

432: It is mentioned that the swinging motion is seen in particular in 

433: the close vicinity to the tank-treading-to-tumbling transition.

434: Thus, the experimental data confirm our predicted phase diagram, 

435: Fig.~\ref{fig:phase}, very well.

436: Furthermore, the oscillation amplitudes also show good agreement, 

437: see Fig.~\ref{fig:cv}.

438: 

439: In summary, we have studied the oscillatory motion of fluid vesicles in 

440: simple shear flow.

441: We have developed a simplified model for the ellipsoidal fluid vesicle,

442: which explains the simulation and experimental results very well.

443: This model could be extended in the future to study the coupling of  

444: different types of 

445: oscillation mechanisms, like membrane shear elasticity and viscosity, 

446: in elastic capsules or red blood cells.

447: 

448: \begin{acknowledgments}

449: We thank R. Finken (Stuttgart) for his help to rederive Eq.~(\ref{eq:lamb}).

450: GG acknowledges support of this work through the DFG priority program 

451: ``Nano- and Microfluidics''.

452: 

453: \end{acknowledgments}

454: 

455: \bibliographystyle{apsrev}

456: %\bibliography{tri}

457: \begin{thebibliography}{23}

458: \expandafter\ifx\csname natexlab\endcsname\relax\def\natexlab#1{#1}\fi

459: \expandafter\ifx\csname bibnamefont\endcsname\relax

460:   \def\bibnamefont#1{#1}\fi

461: \expandafter\ifx\csname bibfnamefont\endcsname\relax

462:   \def\bibfnamefont#1{#1}\fi

463: \expandafter\ifx\csname citenamefont\endcsname\relax

464:   \def\citenamefont#1{#1}\fi

465: \expandafter\ifx\csname url\endcsname\relax

466:   \def\url#1{\texttt{#1}}\fi

467: \expandafter\ifx\csname urlprefix\endcsname\relax\def\urlprefix{URL }\fi

468: \providecommand{\bibinfo}[2]{#2}

469: \providecommand{\eprint}[2][]{\url{#2}}

470: 

471: \bibitem[{\citenamefont{Rallison}(1984)}]{rall84}

472: \bibinfo{author}{\bibfnamefont{J.~M.} \bibnamefont{Rallison}},

473:   \bibinfo{journal}{Ann.\ Rev.\ Fluid\ Mech.} \textbf{\bibinfo{volume}{16}},

474:   \bibinfo{pages}{45} (\bibinfo{year}{1984}).

475: 

476: \bibitem[{\citenamefont{Barth{\`e}s-Biesel and Sgaier}(1985)}]{bart85}

477: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Barth{\`e}s-Biesel}}

478:   \bibnamefont{and} \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Sgaier}},

479:   \bibinfo{journal}{J.\ Fluid\ Mech.} \textbf{\bibinfo{volume}{160}},

480:   \bibinfo{pages}{119} (\bibinfo{year}{1985}).

481: 

482: \bibitem[{\citenamefont{de~Haas et~al.}(1997)\citenamefont{de~Haas, Blom,

483:   van~den Ende, Duits, and Mellema}}]{haas97}

484: \bibinfo{author}{\bibfnamefont{K.~H.} \bibnamefont{de~Haas}},

485:   \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Blom}},

486:   \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{van~den Ende}},

487:   \bibinfo{author}{\bibfnamefont{M.~H.~G.} \bibnamefont{Duits}},

488:   \bibnamefont{and} \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Mellema}},

489:   \bibinfo{journal}{Phys.\ Rev. E} \textbf{\bibinfo{volume}{56}},

490:   \bibinfo{pages}{7132} (\bibinfo{year}{1997}).

491: 

492: \bibitem[{\citenamefont{Kantsler and Steinberg}(2006)}]{kant06}

493: \bibinfo{author}{\bibfnamefont{V.}~\bibnamefont{Kantsler}} \bibnamefont{and}

494:   \bibinfo{author}{\bibfnamefont{V.}~\bibnamefont{Steinberg}},

495:   \bibinfo{journal}{Phys.\ Rev.\ Lett.} \textbf{\bibinfo{volume}{96}},

496:   \bibinfo{pages}{036001} (\bibinfo{year}{2006}).

497: 

498: \bibitem[{\citenamefont{Seifert}(1999)}]{seif99}

499: \bibinfo{author}{\bibfnamefont{U.}~\bibnamefont{Seifert}},

500:   \bibinfo{journal}{Eur.\ Phys.\ J. B} \textbf{\bibinfo{volume}{8}},

501:   \bibinfo{pages}{405} (\bibinfo{year}{1999}).

502: 

503: \bibitem[{\citenamefont{Misbah}(2006)}]{misb06}

504: \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Misbah}},

505:   \bibinfo{journal}{Phys.\ Rev.\ Lett.} \textbf{\bibinfo{volume}{96}},

506:   \bibinfo{pages}{028104} (\bibinfo{year}{2006}).

507: 

508: \bibitem[{\citenamefont{Kraus et~al.}(1996)\citenamefont{Kraus, Wintz, Seifert,

509:   and Lipowsky}}]{krau96}

510: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Kraus}},

511:   \bibinfo{author}{\bibfnamefont{W.}~\bibnamefont{Wintz}},

512:   \bibinfo{author}{\bibfnamefont{U.}~\bibnamefont{Seifert}}, \bibnamefont{and}

513:   \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Lipowsky}},

514:   \bibinfo{journal}{Phys.\ Rev.\ Lett.} \textbf{\bibinfo{volume}{77}},

515:   \bibinfo{pages}{3685} (\bibinfo{year}{1996}).

516: 

517: \bibitem[{\citenamefont{Beaucourt et~al.}(2004)\citenamefont{Beaucourt, Rioual,

518:   S\'eon, Biben, and Misbah}}]{beau04}

519: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Beaucourt}},

520:   \bibinfo{author}{\bibfnamefont{F.}~\bibnamefont{Rioual}},

521:   \bibinfo{author}{\bibfnamefont{T.}~\bibnamefont{S\'eon}},

522:   \bibinfo{author}{\bibfnamefont{T.}~\bibnamefont{Biben}}, \bibnamefont{and}

523:   \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Misbah}},

524:   \bibinfo{journal}{Phys.\ Rev.\ E} \textbf{\bibinfo{volume}{69}},

525:   \bibinfo{pages}{011906} (\bibinfo{year}{2004}).

526: 

527: \bibitem[{\citenamefont{Noguchi and Gompper}(2004)}]{nogu04}

528: \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Noguchi}} \bibnamefont{and}

529:   \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Gompper}},

530:   \bibinfo{journal}{Phys.\ Rev.\ Lett.} \textbf{\bibinfo{volume}{93}},

531:   \bibinfo{pages}{258102} (\bibinfo{year}{2004}).

532: 

533: \bibitem[{\citenamefont{Noguchi and Gompper}(2005)}]{nogu05}

534: \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Noguchi}} \bibnamefont{and}

535:   \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Gompper}},

536:   \bibinfo{journal}{Phys.\ Rev.\ E} \textbf{\bibinfo{volume}{72}},

537:   \bibinfo{pages}{011901} (\bibinfo{year}{2005}).

538: 

539: \bibitem[{\citenamefont{Fischer et~al.}(1978)\citenamefont{Fischer,

540:   St\"ohr-Liesen, and Schmid-Sch\"onbein}}]{fisc78}

541: \bibinfo{author}{\bibfnamefont{T.~M.} \bibnamefont{Fischer}},

542:   \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{St\"ohr-Liesen}},

543:   \bibnamefont{and}

544:   \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Schmid-Sch\"onbein}},

545:   \bibinfo{journal}{Science} \textbf{\bibinfo{volume}{202}},

546:   \bibinfo{pages}{894} (\bibinfo{year}{1978}).

547: 

548: \bibitem[{\citenamefont{Keller and Skalak}(1982)}]{kell82}

549: \bibinfo{author}{\bibfnamefont{S.~R.} \bibnamefont{Keller}} \bibnamefont{and}

550:   \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Skalak}},

551:   \bibinfo{journal}{J.\ Fluid\ Mech.} \textbf{\bibinfo{volume}{120}},

552:   \bibinfo{pages}{27} (\bibinfo{year}{1982}).

553: 

554: \bibitem[{\citenamefont{Tran-Son-Tay et~al.}(1984)\citenamefont{Tran-Son-Tay,

555:   Sutera, and Rao}}]{tran84}

556: \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Tran-Son-Tay}},

557:   \bibinfo{author}{\bibfnamefont{S.~P.} \bibnamefont{Sutera}},

558:   \bibnamefont{and} \bibinfo{author}{\bibfnamefont{P.~R.} \bibnamefont{Rao}},

559:   \bibinfo{journal}{Biophys.\ J.} \textbf{\bibinfo{volume}{46}},

560:   \bibinfo{pages}{65} (\bibinfo{year}{1984}).

561: 

562: \bibitem[{\citenamefont{Pozrikidis}(2003)}]{pozr03}

563: \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Pozrikidis}},

564:   \bibinfo{journal}{Annals\ Biomed.\ Eng.} \textbf{\bibinfo{volume}{31}},

565:   \bibinfo{pages}{1194} (\bibinfo{year}{2003}).

566: 

567: \bibitem[{\citenamefont{Chang and Olbricht}(1993)}]{chan93}

568: \bibinfo{author}{\bibfnamefont{K.~S.} \bibnamefont{Chang}} \bibnamefont{and}

569:   \bibinfo{author}{\bibfnamefont{W.~L.} \bibnamefont{Olbricht}},

570:   \bibinfo{journal}{J. Fluid Mech.} \textbf{\bibinfo{volume}{250}},

571:   \bibinfo{pages}{609} (\bibinfo{year}{1993});

572: \bibinfo{author}{\bibfnamefont{Y.}~\bibnamefont{Navot}},

573:   \bibinfo{journal}{Phys.\ Fluids} \textbf{\bibinfo{volume}{10}},

574:   \bibinfo{pages}{1819} (\bibinfo{year}{1998}).

575: 

576: \bibitem[{\citenamefont{Skotheim and Secomb}()}]{skot06}

577: \bibinfo{author}{\bibfnamefont{J.~M.} \bibnamefont{Skotheim}} \bibnamefont{and}

578:   \bibinfo{author}{\bibfnamefont{T.~W.} \bibnamefont{Secomb}},

579:   \eprint{arXiv:cond-mat/0605664}.

580: 

581: \bibitem[{\citenamefont{Finken and Seifert}(2006)}]{fink06}

582: \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Finken}} \bibnamefont{and}

583:   \bibinfo{author}{\bibfnamefont{U.}~\bibnamefont{Seifert}},

584:   \bibinfo{journal}{J. Phys. Condens. Matter} \textbf{\bibinfo{volume}{18}},

585:   \bibinfo{pages}{L185} (\bibinfo{year}{2006}).

586: 

587: \bibitem[{\citenamefont{Viallat}()}]{vall06}

588: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Viallat}}, \eprint{(private

589:   communication)}.

590: 

591: \bibitem[{\citenamefont{Gompper and Kroll}(1997)}]{gg:gomp97f}

592: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Gompper}} \bibnamefont{and}

593:   \bibinfo{author}{\bibfnamefont{D.~M.} \bibnamefont{Kroll}},

594:   \bibinfo{journal}{J. Phys. Condens. Matter} \textbf{\bibinfo{volume}{9}},

595:   \bibinfo{pages}{8795} (\bibinfo{year}{1997}).

596: 

597: \bibitem[{\citenamefont{Malevanets and Kapral}(1999)}]{male99}

598: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Malevanets}} \bibnamefont{and}

599:   \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Kapral}},

600:   \bibinfo{journal}{J.\ Chem.\ Phys.} \textbf{\bibinfo{volume}{110}},

601:   \bibinfo{pages}{8605} (\bibinfo{year}{1999}).

602: 

603: \bibitem[{\citenamefont{Milner and Safran}(1987)}]{miln87}

604: \bibinfo{author}{\bibfnamefont{S.~T.} \bibnamefont{Milner}} \bibnamefont{and}

605:   \bibinfo{author}{\bibfnamefont{S.~A.} \bibnamefont{Safran}},

606:   \bibinfo{journal}{Phys.\ Rev.\ A} \textbf{\bibinfo{volume}{36}},

607:   \bibinfo{pages}{4371} (\bibinfo{year}{1987}).

608: 

609: \bibitem[{\citenamefont{Dimova et~al.}(1999)\citenamefont{Dimova, Dietrich,

610:   Hadjiisky, Danov, and Pouligny}}]{dimo99}

611: \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Dimova}},

612:   \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Dietrich}},

613:   \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Hadjiisky}},

614:   \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Danov}}, \bibnamefont{and}

615:   \bibinfo{author}{\bibfnamefont{B.}~\bibnamefont{Pouligny}},

616:   \bibinfo{journal}{Eur.\ Phys.\ J.\ B} \textbf{\bibinfo{volume}{12}},

617:   \bibinfo{pages}{589} (\bibinfo{year}{1999}).

618: 

619: \end{thebibliography}

620: 

621: \end{document}

622: 

623: