1 Non-equilibrium evolution of Long-range Interacting Systems and Polytrope Kyoto University Masa-aki SAKAGAMI Collaboration with T.Kaneyama (Kyoto U.) A.Taruya (RESCEU, Tokyo U.) 2 Outlines of this talk Two systems with long-range interaction (1) Self-gravitating N-body system A short review of our previous work. Extremal of generalized entropy by Tsallis polytrope f (x, v) A(0 ) q /(1q) Taruya and Sakagami, PRL90(2003)181101, Physica A322(2003)285 generalization of B-G dist. 12 v2 ( x) It well describes non-equilibrium evolution of the system. (2) 1-dim. Hamiltonian Mean-Field model (HMF model) We show another example where non-equilibrium evolution of the system moves along a sequence of polytrope. Thermodynamical instability implies a superposition of polytrope. (negative specific heat) Self-gravitating N-body System A System with N Particles (stars) (N>>1) Particles interact with Newtonian gravity each other 2 H i Gmi m j pi 2mi i j i | x i x j | Typical example of a system under long-range force Key word Negative Specific Heat Long-term (thermodynamic) instability ( t t relax ) Gravothermal instability Antonov 1962 Lynden-Bell & Wood 1968 3 Maxwell-Boltzmann distribution as 6 Equilibrium BoltzmannGibbs entropy S BG d 3x d 3v f ( x, v ) ln f ( x, v ) Extremization 0 S BG M E Maxwell-Boltzmann distribution l – reE/GM 2 f ( x, v) exp( ) ; Large lcrit = 0.335 v2 / 2 (x) D rc / r re e rc r e Dcrit = 709 Adiabatic wall D= rc/ re (perfectly reflecting boundary) A naïve generalization of BG statistics 7 Thermostatistical treatment by generalized entropy q-entropy q Sq 1 d 3x d 3v p(x, v) p(x, v) q1 BG limit q→1 One-particle distribution function identified with escort distribution f ( x , v) M Power-law distribution Tsallis, J.Stat.Phys.52 (1988) 479 { p( x, v)}q 3 3 q d x d v { p ( x , v )} p p p f (x, v) A(0 ) q /(1q ) 0 Sq M E v 2 / 2 (x) 9 Energy-density contrast relation for stellar polytrope f (x, v) A(0 ) q /(1q) n=6 Polytropic equation of state P( r ) Kn r 11 / n ( r ) (e.g., Binney & Tremaine 1987) Polytrope index 1 1 n 1 q 2 n→∞ BG limit Stellar polytrope as quasi-equilibrium state Energy-density contrast relation for stellar polytrope n=6 stable 10 unstable unstable state appears at n>5 (gravothermal instability) Survey results of group (A) 11 The evolutionary track keeps the direction increasing the polytrope index “n”. Once exceeding the critical value “Dcrit“, central density rapidly increases toward the core collapse. Run n3A : N-body simulation Density profile Initial cond: 12 Stellar polytrope 4 (n=3,D=10 ) One-particle distribution function Fitting to stellar polytropes is quite good until t ~ 30 trh,i. 1 2 v ( x) 2 Overview of the results in sel-gravitating system 13 Stellar polytropes are not stable in timescale of two-body relaxation. However, focusing on their transients, we found : Quasi-equilibrium property Transient states approximately follow a sequence of stellar polytropes with gradually changing polytrope index “n”. Quasi-attractive behavior Even starting from non-polytropic states, system soon settles into a sequence of stellar polytropes. 14 Application of generalized entropy and polytrope to another long-range interacting system, 1-dimensional Hamiltonian mean-field model 1D-HMF model Antoni and Ruffo, PRE 52(1995)2361 15 Antoni and Ruffo, PRE 52(1995)2361 A naïve generalization of BG statistics 16 Thermostatistical treatment by generalized entropy q-entropy q 1 Sq d dp h( , p) h( , p) q1 BG limit q→1 One-particle distribution function identified with escort distribution Tsallis, J.Stat.Phys.52 (1988) 479 {h( , p )}q f ( , p ) N q d d p h ( , p )} Power-law distribution (polytrope) h h h f ( , p) A(0 ) q /(1q ) p 2 / 2 ( ) Physical quantities by 1-particle distribution number N d dp f ( , p) energy E d dp 12 p 2 ( ) f ( , p) 1 magnetization M ( M x , M y ) d dp cos , sin f ( , p ) N 1 ( ) d ' dp ' 1 cos( ' )) f ( ' , p' ) potential N ( ) 1 M cos( ) 17 Polytropic distribution function f ( , p) A0 n1/ 2 18 Chavanis and Campa, Eur.Phys.J. B76(2010)581 Taruya and Sakagami, unpublished polytrope index BG limit q→1 n →∞ 1 2 q 1 2 p ( ) ( ) 1 M cos n 1 q 2 0 (n 1)Tphys 1 M 2 M n d cos ( 1 M cos ) 0 n d ( 1 M cos ) 0 Tphys 2 K / N 1 N For given U and n, this eq. self-consistently determines magnitization M. 2 d dp p f ( , p) U E / N 12 Tphys 1 M 2 As for defenition of Tphys, Abe, Phys.Lett. A281(2001)126 Thermal equilibrium,BG Limit n→∞ Antoni and Ruffo, PRE 52(1995)2361 19 inhomogeneous state M 0 homogeneous state generalized to seqences of polytrope, describing non-eqilibrium (?) n=1000 n=10 n=4 n=2 n=0.5 n=1000 n=10 n=4 n=2 n=0.5 Time evolution of M and Tphys (1) U E / N 0.69 N=10000, 10 samples of simulations Initial cond. : Water Bag Spatially homogeneous M 0 20 Time evolution of M and Tphys (2) U E / N 0.69 N=10000, 10 samples of simulations Initial cond. : Water Bag M 0 Spatially inhomogeneous 21 Three stages of evolution of M and Tphys U E / N 0.69 N=10000, 10 samples of simulations nearly equilibrium transient state quasi-stationary state (qss) M 0 The evolution over three stages are totally well described by sequences of polytropes. 22 Tphys – U relation at equilibrium U E / N 0.62 Long term behavior N=1000, single sample 23 early stage behavior Theoretical prediction for thermal equilibrium. n Tphys – U relation at qss U E / N 0.62 Long term behavior N=1000, single sample early stage behavior Tphys at early stage of qss 24 Tphys – U relation at qss U E / N 0.62 Long term behavior N=1000, single sample 25 polytrope n=0.5 early stage behavior Tphys at early stage of qss are explained by polytrope with n=0.5. Evolutionary track on the polytrope sequence U E / N 0.26 69 N=10000, 10 samples of simulations U E / N 0.69 n 0.90 U 0.69 t 60000 n 0.90 f by simulation prediction by polytrope Evolutionary track on the polytrope sequence U E / N 0.69 N=10000, 10 samples of simulations n 1.58 t 300300 n 1.58 Even in qss, polytrope index n and distribution f change. 27 Evolutionary track on the polytrope sequence U E / N 0.69 N=10000, 10 samples of simulations t 600800 n 2.69 n 2.69 28 Evolutionary track on the polytrope sequence U E / N 0.69 N=10000, 10 samples of simulations t 700300 n 3.95 n 3.95 29 Evolutionary track on the polytrope sequence U E / N 0.69 N=10000, 10 samples of simulations t 800800 n 6.46 n 6.46 30 Evolutionary track on the polytrope sequence U E / N 0.69 N=10000, 10 samples of simulations t 1000800 n 13.9 n 13.9 31 Evolutionary track on the polytrope sequence U E / N 0.69 N=10000, 10 samples of simulations t 1800800 n 58 .6 n 58 .6 32 Evolutionary track on the polytrope sequence U E / N 0.69 N=10000, 10 samples of simulations t 2500100 n 66 .4 33 Evolutionary track on the polytrope sequence U E / N 0.69 N=10000, 10 samples of simulations t 3000700 n 597 n 597 34 35 Failure of single polytrope description due to Thermodynamical instability (negative specific heat). Stellar polytrope as quasi-equilibrium state Energy-density contrast relation for stellar polytrope n=6 stable 36 unstable unstable state appears at n>5 (gravothermal instability) 37 self-gravitating system Sel-similar core-collapse in Fokker-Planck eq. Halo could not catch up with core collapse. CV 0 Heat flow core halo CV 0 When self-similar core collapse takes place, polytrope could not fit distribution function. H.Cohn Ap.J 242 p.765 (1980) fitting of self-similar sol. with double polytrope f ( ) A 0 n1 c 0 38 n2 Black dots: Numerical Self-Similar sol. by Heggie and Stevenson f ( ) Magenta lines: fitting by double polytrope n1 9.538 n2 16.73 c 3.59 r r 39 Thermodynamical Instability due to negative specific heat After a short time, single polytrope fails to describe the simulated distribution. t 5000 n 1 We prepare the initial state as Polytrope with U=0.6, n=1. 40 A description by double polytropes might work. double polytrope t 5000 single polytrope t 5000 41 A description by double polytropes might work. double polytrope t 5000 parameters n1 0.95, n2 1.2, l1 0.62 coexistence conditions (preliminary) T l1 T1 l2 T2 1 l1 l2 U1 U2 U1 12 T1 12 l1 (1 M12 ) U2 12 T2 12 l2 Summary and Discussion 42 (1) Polytrope (Extremal of Generalized Entropy) describes evolution along quasi-stationary and transient states to thermal equilibrium Self-gravitating system, 1D-HMF (Long-range interaction) (2) Break down of single polytrope description (due to negative specific heat) implies superposition of polytropes. 43 How to determine polytropic index n. 2 2 2n 1 2 2 2 ( n 2 ) T ( 2 n 3 ) l M T l ( 1 M ) phys phys 2 4( n 2 ) 44 45 Polytrope による準定常状態の記述の限界 Fokker-Planck eq. によるCore-Collapse の解析 46 self-similar evolution CV 0 Heat flow core halo CV 0 self-similar core collapse が 始まると polytrope で fit できない H.Cohn Ap.J 242 p.765 (1980) Self-similar sol. of F-P eq. r Heggie and Stevenson, MN 230 p.223 (1988) 47 power law envelope ln r r r 2.21 n 9 .7 ln r isothermal core fitting of self-similar sol. with double polytrope f ( ) A 0 n1 c 0 48 n2 Black dots: Numerical Self-Similar sol. by Heggie and Stevenson f ( ) Magenta lines: fitting by double polytrope n1 9.538 n2 16.73 c 3.59 r r 49 2D HMFモデル Antoni&Torcini PRE 57(1998) R6233 Antoni, Ruffo&Torcini, PRE 66(2003) 025103R 1 N 2 H ( px ,i p 2y ,i ) 2 i 1 1 2N 3 cos(x x ) cos( y y ) cos(x x ) cos( y y ) N i j i j i j i j i, j Interaction by Mean-field: Long-range interacting system 2D HMF have the effect of energy transfer due to 2-body scattering process. Negative specific heat in some range of energy 50 Magnetization T-U curve Boltzmann case U 1.95 T U (energy) N 10 , N 9 10 4 4 Thermal equilibrium Transient st polytrope ? t/N Vlasov phase ? dist. func. dist. func. vx vx 51 Initial : polytrope U=1.9 52 Negative specific heat Magnetization dist. func. p t (logarithmic) Initial: polytrope U=1.7 53 positive specific heat Magnetization dist. func. p t (logarithmic) Initial WB U=1.9 54 Negative specific heat Magnetization dist. func. p t (logarithmic) Initial WB U=1.7 55 positive specific heat Magnetization dist. func. p t (logarithmic) 56 How to derive the evolution equation for polytropic index, q or n. self-gravitating systems Kinetic-theory approach 57 For a better understanding of the quasi-equilibrium states, Fokker-Planck (F-P) model for stellar dynamics orbit-averaged F-P eq. f ( ) ( ) t t ; 16 2G2m2 ln ln f ( ) ln f ( ) ( ) d f ( ) f ( ) min[ ( ), ( )] 16 2 3/ 2 2 ( ) dr r 2 [ ( r )] phase space volume 3 Complicated, but helpful for semi-analytic understanding Generalized Variational Principle for F-P eq. 58 Glansdorff & Prigogine (1971) Local potential Inagaki & Lynden-Bell (1990) 2 f 0 ln f ln f ( f , f 0 ) d ln f d d f f min( , ) 0 0 0 0 4 t Variation w.r.t. f ( f , f0 ) 0 f ff F-P equation for f 0 0 f0 fixed Absolute minimum at a solution f0 ( f , f 0 ) ( f 0 , f 0 ) 0 Application: Takahashi & Inagaki (1992); Takahashi (1993ab) The evolution eq. for “n” from generalized variational principle 59 Assuming stellar polytropes with time-varying polytrope index as transient state, trial function f ( ) A(t )[ 0 (t ) ]n (t )3 / 2 ( f , f 0 ) 0 n ff ( A and 0 are the functions of n) function of n, E , M 0 l n(t ) e n l l ln f d ( ) e n n e n n e e 2 l ln f l ln f d f ( ) e n n e n n e e Semi-analytic prediction: evolution of “n” Time evolution of polytrope index “n” fitted to N-body simulations Time-scale of quasi-equilibrium states is successfully reproduced from semi-analytic approach based on variational method. 60 Summary and Discussion (1) Polytrope (Extremum of Generalized Entropy) Transient states to thermal equilibrium Self-gravitating system, 2D-HMF Negative specific heat Long-range interaction (2) Generalized variational princple for F-P eq. Evolution eq. for polytropic index Works in Progress: Short-range attracting interaction negative specific heat Polytrope ? Superposition of Boltzmann dist. ? 61 62 63 Summary (1) 重力多体系 (2) 準定常状態 長距離力(引力) 比熱が負 small system 非平衡進化 準定常状態が存在 ポリトロープ状態の系列で記述できる P( r ) K n r 11 / n (r) n 1 1 1 q 2 (3) ポリトロープ指数 n の時間発展 一般化された変分原理 Fokker-Planck eq. ポリトロープ状態: Trial func 指数 n の時間発展方程式 が導出できる Work in Progress (1) ポリトロープ 準定常状態: 他の例はあるか 2次元HMFモデルの解析 (2) ポリトロープ 準定常状態: 長距離相互作用が本質? Yukawa 型相互作用での解析 (3) ポリトロープによる準定常状態の記述の限界 ポリトロープは core collapse 前しか適用できない? 64
© Copyright 2024 ExpyDoc