* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Nonlinear waves and shocks in relativistic two-fluid hydrodynamics
Euler equations (fluid dynamics) wikipedia , lookup
Partial differential equation wikipedia , lookup
Magnetic field wikipedia , lookup
Field (physics) wikipedia , lookup
Navier–Stokes equations wikipedia , lookup
Neutron magnetic moment wikipedia , lookup
State of matter wikipedia , lookup
Equations of motion wikipedia , lookup
Electromagnetism wikipedia , lookup
Woodward effect wikipedia , lookup
Maxwell's equations wikipedia , lookup
Superconductivity wikipedia , lookup
Magnetic monopole wikipedia , lookup
Aharonov–Bohm effect wikipedia , lookup
Lorentz force wikipedia , lookup
Electromagnet wikipedia , lookup
Theoretical and experimental justification for the Schrödinger equation wikipedia , lookup
Time in physics wikipedia , lookup
Nonlinear waves and shocks in relativistic two-fluid hydrodynamics Thesis submitted in partial fulfillment of the requirement for the degree of master of science in the Faculty of Natural Sciences Submitted by: Lev Haim Advisor: Prof. M. Gedalin Department of Physics Faculty of Natural Sciences Ben-Gurion University of the Negev August, 2009 Abstract Studies of nonlinear waves in relativistic plasmas are important in connection with collisionless shocks which are expected to form in a number of astrophysical objects dominated by relativistic plasmas. We study stationary one-dimensional nonlinear waves propagating in the direction perpendicular to the external magnetic field, within the two-fluid approach. The research covers the cold and hot pair plasma cases as well as electron-ion plasma. In the symmetric pair plasma the absence of the electric field along the propagation direction allows to derive a pseudopotential-type equation for the magnetic field or density in the cold as hot cases as well. We reproduce the well-known results regarding compressive solitons and show that the pair plasma allows also rarefactive solitary wave solutions. In the case of electron-ion plasma we invoke the assumption of the quasineutrality in the species rest frame and solve the equations in the approximation of the weak longitudinal electric field. This approach allows to derive a single pseudopotential equation for the density. The equation is solved numerically in a wide range of plasma parameters. It is found that the compressive solitons are similar to those found for the pair plasma and do not exist for Mach numbers exceeding some critical Mach number. There are no rarefactive solitons in the electron-ion plasma in this approximation. 2 Contents 1 Introduction 4 1.1 Scientific Background and Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 1.2 Thesis Outline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 2 Multi-Fluid Hydrodynamics Model 9 2.1 4-Representation and Continuity Equation . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 2.2 Energy-Momentum Tensor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 2.2.1 Energy, Momentum, State & Maxwell’s Equations . . . . . . . . . . . . . . . . . . 10 One-dimensional Stationary Oblique Equations . . . . . . . . . . . . . . . . . . . . . . . . 12 2.3 3 Two Fluid MHD: Perpendicular Case 14 3.1 Two Fluid Perpendicular Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 3.2 Normalization and Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 3.3 Conservation Laws . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 4 Electron-Positron Plasma 19 4.1 Cold Plasma Classical Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 4.2 Cold Plasma Non Classical Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 4.3 Hot Plasma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 4.3.1 Singularity & Maximum Amplitude . . . . . . . . . . . . . . . . . . . . . . . . . . 30 4.3.2 Numerical Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Magnetic Profile & Pair Plasma Summarize . . . . . . . . . . . . . . . . . . . . . . . . . . 34 4.4 5 Electron-Ion Plasma 39 5.1 Numerical Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2 Magnetic Profile & Electron-Ion Plasma Summarize . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Discussions & Conclusions 44 48 52 3 Chapter 1 Introduction 1.1 Scientific Background and Motivation Observations have shown that active galactic nuclei (AGNs), quasars (QSOs), micro quasars (µQSOs) and gamma-ray bursts (GRBs) eject relativistic jets [Pearson & Zensus 1987; Biretta et al. 1999; Junor et al. 1999; Mirabel & Rodriguez 1998; Kulkarni 1999]. In spite of the drastic difference of their characteristic scales and powers, it is believed that the activities of these objects, such as the relativistic jet ejection, are supported commonly by the drastic phenomena of accretion disks around black holes [Mirabel & Rodriguez 1998]. However, the distinct mechanism of the activity is not confirmed yet. Models with interaction between the plasma and magnetic field in the very strong gravity of the black hole are thought to be most promising. A black hole magnetosphere is the region where the plasma interact with the magnetic field of the black hole, which consist of an accretion disk and corona around a black hole and it is thought to be composed of various kinds of plasma. In the case of an AGNs, it is suggested that the disk is made of electron-ion plasma [Ford et al. 1994], the core of the relativistic jet is mainly of electron-positron (pair) plasma [Wardle et al. 1998], and the corona is both of them, while the actual components of such plasmas have not been conformed observationally yet. Propagating relativistic shock waves are the leading candidates for the origin of the emission from GRBs and their afterglows [Piran 1999, 2000; Mészáros 2000 ] as well as of the spectral flares and apparent superluminal motions exhibited by the blazar class 4 of AGNs [Ulrich, Maraschi, & Urry 1997]. The energy for the accelerated, synchrotron emitting particles is though to come from the kinetic energy and Poynting flux of relativistic jets emanating from energizing compact objects. Shocks waves in these relativistic flows are most plausible candidate for conversion of flow energy into random energy of the synchrotron emitting particles. In the case of blazars, the measurement of high optical linear polarization in some objects directly supports the synchrotron interpretation [ Angel & Stockman 1980]. In the case of GRBs there are two confirmed detections of optical polarization [Covino et al. 1999; Wijers et al. 1999; Rol et al. 2000], which, despite the relatively low measured values , appear to be consistent with a synchrotron origin. In both blazars and microquasars the case for synchrotron radiation is supported also by measurements of radio polarization. These shock waves are expected to be collisionless [Granot & Königl 2001]. Nonlinear waves were extensively studied in the framework of magnetohydrodynamics (MHD) and two-fluid hydrodynamics. A typical analysis sequence is as follows: onedimensionality assumption, model equation construction, and stationary solution analysis with particular attention to solitons. Although deviations from one-dimensionality are significant in some cases, as e.g. in the case of whistler self-focusing [Karpman et al. 1992] , it is widely believed that in most cases one-dimensional waves are a good approximation. In most papers on nonlinear waves in MHD and two-fluid hydrodynamics cold plasmas are considered [Kennel & Pellat 1976; Aslop & Arons 1988]. Both [Kennel & Pellat 1976] and [Aslop & Arons 1988] have concluded that fast waves propagating perpendicular to the magnetic field are restricted to low amplitudes in relativistic limit and have shown that compressive solitons can exist in electron-positron plasma as well as ion-electron plasma, respectively, in the regime where the flow of cold plasma is perpendicular to the magnetic field. Moreove, [Kennel & Pellat 1976] have assumed that for electron ion plasmas particle inertia is determined by wave amplitude, not by the rest masses in the relativistic limit. p Hence, have found a critical Mach number, that is MC = 2(1 + γo ), and largest possible ∗ magnetic field amplitude when MA = MC , that is Bmax = 1 + 2/γo . The importance of the parallel/perpendicular distinction is clear when we consider the motion of a particle that is initially headed away from the shock. In the de Hoffman-Teller frame the motion 5 has two parts: an unimpeded motion along the direction of the magnetic field, and a gyration around (i.e., transverse to) it. In the case of a parallel shock, the field lines pass through the shock, and a particle’s motion along the field will carry the particle through (and away from) the shock relatively easily. On the other hand, at a perpendicular shock, the field lines are parallel to the shock surface, and so particle motion along the magnetic field does not let the particle pass away from the shock. Indeed, the particle gyration at a perpendicular shock brings the particle back to the shock. There is a further simple conclusion to be drawn from the parallel/perpendicular distinction. Because particle motion in the normal direction is ”easier” at the parallel shock, compared with at the perpendicular shock, then we can expect that the magnetic field does not play an important role in the former case. However, in the context of a collisionless plasma, the only way for dissipation to occur is via field-particle processes, and it is certain that here the fields will play a crucial role [Kivelson & Russell 1996 ]. Moreover, assuming that magnetic field is isotropic in the wave frame, which moves relatively to laboratory frame with velocity ~v . Thus, according to Lorentz field transformations, in the laboratory frame magnetic field parallel to the flow is not change where as transverse magnetic field is increased by Lorentz factor. Hence, in this paper we restrict our self to the perpendicular case. Investigations of various aspects of relativistic hot electron-positron/ion plasma have bean carried out. But there is little progress in this trend due to the mathematical complexity and difficulty in the understanding of the phenomena. In addition, it should be noted that consideration of the problem is very complicated at relativistic temperatures and also when the particles have relativistic hydrodynamical velocities. Nevertheless, a number of papers dealt with a hot plasma. Using the water-bag model for the distribution function and neglected the electron inertia, electron internal energy and pressure in the face of relativistic proton energy scale [Chiueh & Lai 1991], those assumption generally are wrong and can not be used in hot relativistic case. Latter conformed that small amplitude soliton of sech2 type do exist. Moreover, the peak amplitude for cold ionp electron plasma of the charge density is npeak = [2 γo2 σ(1 + σ)/(γo2 − 1) − 1 − σ]−1 , while for the hot case npeak is of order γo , where γo is the Lorentz factor corresponding to the mean fluid velocity far upstream (sub index o) and σ = (uA )/c2 , with uA being the 6 proper Alfvenic speed. Considerable deviation from the above steps scenario was done [Medvigy & Loeb 2001]. Showing that no stationary, continuous shock solution exist for hot relativistic pair plasmas. Moreover, they have found that soliton solutions exist only for Bo2 /8πme no c2 γo2 1, where me is electron mass and Bo /γo plasma’s self magnetic field defined at the proper reference frame with charge density no . Above references confirmed the existence of the compressive (bright) solitons. Existence of a different nature of solitary wave, rarefactive (dark) solitons were noticed [Lakhina & Verheest 1997], as well as compressive (bright) solitons in relativistic hot pair plasma, where an expansion for all variables were used to deduce a KdV equation, which implied that in ultrarelativistic pair plasmas perpendicularly propagating solitons of sech2 type have a compressive/rarefactive nature. However, typical length scales or peak amplitudes of those solitons were not found. It should be noted that compressive and rarefactive terms are not really appropriate for the changes in magnetic field but more suitable for changes in a charge density. Dark solitons are known in optical systems [Farina & Bulanov and references in them], and they have been observed and discussed also in the Bose-Einstein condensate [Farina & Bulanov and references in them]. The objective of the present research is to study the properties of the nonlinear stationary one-dimensional waves (NS1DW) in relativistic magnetized plasmas. The main emphasis is on the case where the plasma flow is perpendicular to the magnetic field. 1.2 Thesis Outline In Chapter 2, we derive general equations and relations for NS1DW from the multifluid relativistic hydrodynamics with finite nonzero temperature and arbitrary number of species, emphasizing the differences between the oblique and perpendicular propagation of the wave with respect to the magnetic field. In Chapter 3, we reduce our set of oblique multi-fluid equations to the two-fluid perpendicular equations, meaning θ = π/2 and s = 1, 2, where 1 stands for positive electric charge e.g. positrons or ions depends on the type of plasma and 2 stands for negative electric charge e.g. electrons. The new normalized set of equations reveal seven constants of the motion. The first is continuity of 7 charge density. This relates the plasma density to the x component of the velocity. The xcomponent of the current is zero with the consequence that both species speeds along the x axis are equal by virtue of quasineutrality. The energy and x, y momentum equations, where z component was left out for the sake of convenience, are used to construct four conservations laws, which correspond to conservation of energy flux, transverse momentum flux, longitudinal momentum flux and electric potential flux. Faraday’s law requires that the component of the electric field, transverse to the direction of propagation, is constant. In perpendicular case Gauss’s law requires that the component of the magnetic field in the direction of propagation be zero. The conditions for existence of soliton structure are derived in chapter 4, which covers the electron positron plasma whereas chapter 5 covers the electron ion plasma. Hot and cold regimes for each plasma are investigated in chapters 4 and 5 using Sagdeev’s pseudopotential method. Compressive and rarefactive soliton solutions are found. Chapter 6 presents main conclusions of the analysis and suggestions for future work on the subject. 8 Chapter 2 Multi-Fluid Hydrodynamics Model In this section we derive fully nonlinear equations from the multi-fluid hydrodynamics. We derive a complete set of equations with finite nonzero temperature and arbitrary number of species. 2.1 4-Representation and Continuity Equation We choose Minkowski metrics with positive signature, gαβ = [1, −1, −1, −1]. From now on all four dimensional tensor indexes will be denoted by Greek letters α, β, ... and take the values t, x, y, z. Here, Usα is the four velocity for each species defined as Usα = (γs , γs~vs /c) where ~vs is the plasma velocity that measured in the lab frame and U α Uα = 1. The index s denotes different species. Continuity equation [Lichnerowicz 1967] reads ∂α (ns Usα ) = 0 where ∂α = 1 ~ ∂ ,∇ c t (2.1) and ns is the proper particle density while Ñs = γs ns is particle density at the laboratory frame. Using those notations we get ~ s γs~vs ) = 0 ∂t (ns γs ) + ∇(n (2.2) 9 2.2 Energy-Momentum Tensor The Energy-Momentum Tensor [Landau & Lifshitz 1986] for an ideal fluid is Tβα = ( + p)U α Uβ − pδβα (2.3) where , p are the energy density and the pressure, respectively, in the rest frame of the fluid. We adopt the ideal fluid description for each species and do not specify the species index unless it is necessary. In this section and its subsections the species index is suppressed, because each species fulfill all equations independently. In what follows we shell use also the notation w = ( + p)/n. The motion (Euler) equation for the charged fluid is ∂β T αβ = qUβ F αβ (2.4) where Fαβ is the electromagnetic field tensor [Landau & Lifshitz 1986], defined as 0 Ex Ey Ez 0 −Ex −Ey −Ez −Ex E 0 −B B 0 −Bz By x z y αβ Fαβ = , F = Ey Bz −Ey Bz 0 −Bx 0 −Bx Ez −By Bx 0 −Ez −By Bx 0 Together with continuity equation we have nU α ∂α (wUβ ) − δβα ∂α p + qnU α Fαβ = 0 (2.5) It should be noted that the usual Einstein summation convention is been used. 2.2.1 Energy, Momentum, State & Maxwell’s Equations Multiplying the electromagnetic field tensor, Fαβ , by four velocity, U α , we obtain: γ ~ U α Fαt = − ~v · E c ~ + ~v × B/c] ~ U α Fαi = γ[E (2.6) (2.7) where i is a spatial index, that is i = x, y, z. Energy and Momentum equations are obtained using equations (2.5)-(2.7): 10 Setting β = t, we obtain Energy Equation, that is h i ~ (wγ) − ∂t p − qnγ~v · E ~ =0 nγ ∂t (wγ) + ~v · ∇ (2.8) Setting β = i, we obtain Momentum Equations, that is i nγ h ~ ~ + qnγ[E ~ + ~v × B/c] ~ − 2 ∂t (wγ~v ) + ~v · ∇ (wγ~v ) − ∇p =0 c (2.9) State Equation Multiplying equation (2.5) by four velocity, U α , and using U β ∂α Uβ = 0 with continuity equation, Eq.(2.1) we have 1 α nU ∂α w − ∂α p = 0 n Defining the convective derivative D = U α ∂α one has Dp = nDw (2.10) If we assume that pressure has a polytropic behavior p = CnΓ then = mc2 n + 1 p Γ−1 and w= Γ p +p = mc2 + n Γ−1n (2.11) Above equations are valid provided that the flow does not come to a halt, that is v > 0. Maxwell’s Equations The derived equations should be completed with Maxwell’s Equations: X ~ 1 ∂E ~ ×B ~ = 4π qs Ñs~vs + ∇ c s c ∂t (2.12) ~ ~ ×E ~ = − 1 ∂B ∇ c ∂t ~ ~ ∇·B = 0 X ~ ·E ~ = 4π ∇ qs Ñs (2.13) (2.14) (2.15) s 11 In above equations magnetic field, electric field, plasma velocity and charge density are measured in an arbitrary reference frame which will henceforth designate as the laboratory frame. 2.3 One-dimensional Stationary Oblique Equations From now on and to the end of the paper we are restricting our equations to stationary one-dimensional waves, meaning we let everything depend on x, that is (∂/∂t) = (∂/∂y) = ~o = Bo (cos θ, 0, sin θ) be the magnetic field in some reference point. (∂/∂z) = 0. Let us B We will normalize all variables with the fluid parameters in this reference point, where we define the rest-frame density n0 and 4-velocity u = (γo , uo , 0, 0) for each species. We are especially interested in soliton solutions, where the plasma state is asymptotically uniform P at x → ±∞. In this case the reference point is at x → −∞, where s qs nos γos = 0 and P qs nos uαos = 0. In what follows subscript ⊥ denotes components perpendicular to the ~ ⊥ = (0, By , Bz ), and subscript o denotes constant components far x direction, e.g. B upstream. Using above notations the complete set of the equations takes the following form ∂x (ns usx ) = 0 → ns usx = const ≡ Js dps = ndws h i ~⊥ Js ∂x (ws γs ) = qs ns usx Ex + ~u⊥s · E Js 1 ∂x (ws usx ) = qs ns γs Ex + (~u⊥s × B⊥ ) · x̂ − ∂x ps c2 c Js usx Bo cosθ ∂x (ws~u⊥s ) = qs ns γs E⊥ + (x̂ × B⊥ ) + (~u⊥s × x̂) c2 c c X ~ ⊥ = 4π (x̂∂x ) × B qs ns~u⊥s c s ~ ⊥ = const → E X = 4π qs ns γs (2.16) (2.17) (2.18) (2.19) (2.20) (2.21) ~⊥ = 0 (x̂∂x ) × E (2.22) ∂x Ex (2.23) s X qs ns usx = 0 (2.24) s 12 Those are multi-fluid one dimensional stationary oblique equations, which should be accompanied with state equation, we shell assume that the pressure have a polytropic behavior p = CnΓ . Now it is easy to emphasize the difference in our equations between the oblique case and the perpendicular case. The main difference is an additional term at the transverse momentum equation, Eq.(2.20). θ = 0 and θ = π/2 corresponds to parallel and perpendicular cases, respectively, where it is easy to see that in the latter case parallel magnetic field is equal to zero. Combining these equations one has (we omit indices for simplicity) d X c ~⊥ × B ~ ⊥ )] = 0 [ nwγux + ~xˆ · (E dx 4π (2.25) or X nwγux + c ˆ ~ ~ ⊥ ) ≡ Q = const ~x · (E⊥ × B 4π (2.26) Further one has X X X d X ~⊥ + ( ~⊥ ( nwux~u⊥ ) = ( qnγ)E qn~u⊥ ) × ~xˆBx + ( qnux )~xˆ × B dx ~⊥ d d X E Bx d ~ ( nwux~u⊥ ) = Ex + B⊥ dx 4π dx 4π dx X ~ ⊥ Bx B ~⊥ Ex E ~ = const nwux~u⊥ − − ≡S 4π 4π (2.27) (2.28) (2.29) Similarly one also has X (nwu2x + p) − 2 Ex2 − B⊥ ≡ P = const 8π (2.30) Now multiply the equation d ~ ⊥) (nwu2x + p) = qnγEx + qn~xˆ · (~u⊥ × B dx (2.31) by wux /q and sum up to get X wux d X X ~ ⊥ × ~xˆ) nwγux )Ex + ( nwγ~u⊥ ) · (B (nwu2x + p) = ( q dx (2.32) Using (2.26) and (2.29) one gets X wux d ~ ~ ~ ×B ~ ⊥ ) − Ex~xˆ · E⊥ × B⊥ = QEx (nwu2x + p) − ~xˆ · (S q dx 4π 13 (2.33) Chapter 3 Two Fluid MHD: Perpendicular Case In this section we reduce our set of oblique multi-fluid equations to the two-fluid perpendicular equations, meaning θ = π/2 and s = 1, 2, where 1 stands for positive electric charge e.g. positrons or ions depends on the type of plasma and 2 stands for negative electric charge e.g. electrons, q1 = −q2 = q. The new set equations will be normalized and used to construct four conservations laws, which correspond to conservation of energy, transverse momentum, longitudinal momentum and electric potential. 3.1 Two Fluid Perpendicular Equations We are restricted to a 1 + 2 dimension analysis (one spatial coordinate x and two velocity components ux̂ and uys ŷ). Although a finite constant uzs , z component of the velocity, in principle exists, it can be trivially incorporated and we shell leave it out in this analysis for the sake of convenience. Latter and transverse momentum equation (2.20) implies that the z component of the electric field is equal to zero as well as the y component of the magnetic field, that is Ez = By = 0. Finally, we consider only situations in which the magnetic field ~ = (0, 0, Bz ), and no variations in the z and y direction is is always in the z direction, B allowed. But the electric field begins to develop the x component with in the soliton where as according Faraday’s law (2.22) the y component of the electric field remain constant. 14 It follows from longitudinal momentum equation (2.19) that Ey = uo B cγo o = βo Bo , given the boundary conditions at infinity. Charge conservation requires X qs ns usx = 0 → n1 ux1 = n2 ux2 s We add the assumption of quasineutrality (verified aposteriori), n1 = n2 = n. This is obviously valid exactly for a symmetric pair plasma but is an approximation for electronion plasmas. As a result, ux1 = ux2 = u. Equations (2.16)-(2.23) take the following form: nu = no uo (3.1) Γ s ps Γs − 1 n qs n [uEx + uys Ey ] h uys i qs n γs Ex + Bz − ∂x ps c i h u qs n γs Ey − Bz c 4πqn − (uy1 − uy2 ) c 4πqn (γ1 − γ2 ) ws = ms c2 + no uo ∂x (ws γs ) = no uo ∂x (ws u) = c2 no u o ∂x (ws uys ) = c2 ∂x Bz = ∂x Ex = (3.2) (3.3) (3.4) (3.5) (3.6) (3.7) Those are two-fluid one dimensional stationary perpendicular equations. It should be noticed that our analysis is carried out in the rest frame of the soliton. Lets us summarize the boundary condition far upstream, that is x → −∞: x n p s , ws ~u1 = (u, uy1 , 0) ~u2 = (u, uy2 , 0) ~ = (0, 0, Bz ) B ~ E = (Ex , βo Bo , 0) x → −∞ no pos , wos → ~uo1 = (uo , 0, 0) ~uo2 = (uo , 0, 0) ~ o = (0, 0, Bo ) B ~ o = (0, βo Bo , 0) E 15 3.2 Normalization and Definitions The phase (and group) speed of the wave in the rest frame of the plasma is cβA , defined [Aslop & Arons 1988; Gedalin 1993 ] by βA2 = Bo2 /4πno wo γo2 σ = 2 2 1 + Bo /4πno wo γo 1+σ where σ= 2 /4π) Poynting flux (cβo )(Brz Bo2 = = plasma energy flux (cβo )(no wo ) 4πno wo γo2 Here, σ is Lorenz invariant quantity, since Brz = Bo /γo is the strength of the upstream magnetic field (fields transformations from [Landau & Lifshitz 1986]) which is measured in an Proper/Rest reference frame and wo = wo1 + wo2 is total enthalpy. Constructing the momentum of the particle uA , when the particle’s speed equal the wave speed, where √ cβA = c uA = γA vA = γA cβA = p σ 1 − βA2 Here, uo is the upstream momentum of the particle, that is uo = γo vo = γo cβo Shocks and solitons are expected to form when uo ≥ uA . One might expect, therefore, the structure of solitary waves to expressed in terms of the magnetosonic Mach number MA , defined by 2 uo u2 γ 2β 2 2 = 2o = o o MA = uA cσ σ (3.8) We proceed by expressing the equations in terms of normalized variables,that is Ws = ws , wo W = W1 + W2 , N= B= Ps = ps = N Γs , po1 Bz , Bo E= Ex , Bo P = P 1 + P2 , n , no U= wo = wo1 + wo2 u , uo x = XL, βs = 16 8πpos , Bo2 uys Vs = c r wo L= 4πno q 2 β = β1 + β2 where L is the typical length scale. Equations (3.1)-(3.7) take the following form: NU = 1 (3.9) Ws = Wos + ∂X (Ws γs ) = βo2 γo2 ∂X (Ws U ) = ∂X (Ws Vs ) = ∂X B = ∂X E = Γs Γs − 1 βs σγo2 2 N Γs −1 − 1 (3.10) √ qs σ [γo E + N Vs ] q √ qs βs σγo2 γo σ N [γs E + Vs B] − ∂ X Ps q 2 √ qs σ N [γs − γo U B] q N − √ (V1 − V2 ) γo σ N √ (γ1 − γ2 ) γo σ (3.11) (3.12) (3.13) (3.14) (3.15) and the normalized boundary condition far upstream, that is X → −∞, are: X → −∞ X N 1 U , Vs 1, 0 → B, E 1, 0 Ps , P 1, 2 2 2 ms c W , W + Γs βs σγo , 1 s 3.3 wo Γs −1 2 Conservation Laws The new normalized set equations (3.9)-(3.15) can be used to construct four conservation laws. W1 γ1 + W2 γ2 = γo [1 + σ (1 − B)] B 2 − E 2 = 1 + β1 (1 − P1 ) + β2 (1 − P2 ) + W1 V1 + W2 V2 = σγo E √ 2γo2 σ(1 + σ)Φ = − W12 − W22 1 + βo2 γo2 U 2 , (3.16) 2βo2 σ (1 − W U ) (3.17) (3.18) E = −dΦ/dX (3.19) Those are conservation laws of energy, transverse momentum, longitudinal momentum and electric potential, respectively. First three equations are easily obtained by summing 17 up the equations, (3.11), (3.12) and (3.13) respectively, for all species and making use of normalized Ampere’s law Eq(3.14) together with normalized Gauss’s law Eq(3.15) the resultant equation must be integrated once. Last equation (3.19) is more complex, so we specify how it was obtained in more details. First step is to replace the pressure term in longitudinal momentum equation (3.12) by βs σγo2 ∂X Ps 2 = N ∂X Ws . Second step is to multiply the resultant equation by qs U Ws and summing up it for all species. In the last step we use conservation of the energy flux (3.16) and the transverse momentum flux (3.18) so that the resultant equation can be integrated once. It should be noticed that we set the constant electric potential far upstream Φo so it eliminates the constant value after integration. 18 Chapter 4 Electron-Positron Plasma In this chapter we examine relativistic pair plasmas in the perpendicular regime using the pseudopotential method. We first study a cold plasma and generalize further onto the hot plasma case. We reproduce classical results for cold plasmas using our model as well as some new results (dark solitons). Dark solitons have been covered by the literature but not in an magnetized plasma [Pillay & Bharuthram 1992; Lee 2008]. Before proceeding to our analysis we briefly describe the pseudopotential method. Figure 4.1: The typical shape of the pseudopotential well that permits solitary wave solutions. φinitial is the value far upstream, x → −∞. φeq is the root of Ψ = 0 and φcn is the root of ∂Ψ/∂φ = 0. In the pseudopotential theory, it is well known that for the solitary wave to exist the 19 pseudopotential should have the form shown in Fig. 4.1 [Lee 2008], the shape of which can be characterized by the following conditions: I. Ψ(φinitial ) = 0 II. ∂Ψ(φinitial )/∂φ = 0 III. ∂ 2 Ψ(φinitial )/∂φ2 < 0 IV. there exist φ = φeq such that Ψ(φeq ) = 0 and no singularities Ψ (φinitial < φ < φeq ) It should be noticed that compressive (bright) solitons are corresponds to φinitial < φeq while rarefactive (dark) solitons are corresponds to φinitial > φeq . 4.1 Cold Plasma Classical Solutions In a symmetric pair plasma ms = m, Ws = W , γs = γ, uy1 = −uy2 = V , Wx = 0. The derived equations take the form 1 N N √ 2V ∂X B = γo σ γ = γo [1 + σ (1 − B)] 2β 2 B 2 = 1 + o (1 − U ) σ U = (4.1) (4.2) (4.3) (4.4) where γ= p 1 + βo2 γo2 U 2 + V 2 (4.5) and W = 1/2 for the cold case. It is easy to derive the equation for the magnetic field in the form (∂X B)2 = −Ψ̃(B) (4.6) where the pseudopotential U (B) is a function of the magnetic field only: h i2 σ 2 2 2 2 2 γo [1 + σ(1 − B)] − 1 + βo γo 1 + 2β 2 (1 − B ) o Ψ̃(B) ≡ − h i 2 γo2 σ 1 + 2βσ2 (1 − B 2 ) 4 o 20 (4.7) Then equation (4.6) becomes formally analogous to the equation of motion of a particle in the a force field of potential Ψ̃(B), where the role of time and positions are played by X and B, respectively. For solitary wave to exist conditions I-IV should be satisfied. Condition I, Ψ̃(1) = 0, is automatically satisfied by the initial conditions as well as condition II, ∂B Ψ̃(1) = 0. Condition III, ∂BB Ψ̃(1) < 0, which is often called the soliton condition by itself, yields a well known result, Mach number should be higher than one, that is MA2 > 1. Condition IV, Ψ(B) = 0, yields three possible roots, which are marked with subscript one two and max: B1 = B2 = 1, Bmax = 2 γo q MA2 − 1 + γo2 − 1 Those are well behaved as long as the singularity in the denominator of equation (4.6) is avoided. The denominator is simply a squared plasma velocity, U 2 , so the singularity corresponds to U → 0 and N → ∞. Choosing B = Bmax we can find such Mach number which sets the plasma velocity to be a zero, U = 0, using equation (4.4): MC = p 2(1 + γo ) (4.8) this is the maximum Mach number permissible for solitary solutions. Using this critical Mach number MC we can find the largest possible magnetic field amplitude. ∗ Bmax = 1 + 2/γo (4.9) ∗ = 3 when γo = 1. In the non relativistic limit (4.8) and (4.9) reduces to MC = 2 and Bmax While the relativistic limits permits an apparently much larger range of permissible Mach numbers 1 < MA < p 2(1 + γo ) than the non relativistic limit, this hole range corresponds to solitary waves of very small maximum amplitudes. 1 < B ∗ < 1 + 2/γo Thus, fast waves propagating perpendicular to magnetic field are restricted to low amplitudes in the relativistic limit. 21 Figure 4.2: Pseudopotential as function of magnetic field, where γo = 2 for all curves. Dashed upper and dashed lower curves are corresponding to cases when there are no solitons, MA < 1 and MA > MC , respectively. The other two corresponds to 1 < MA < MC but different mach numbers. In all cases . Figure (4.2) emphasize the classical soliton (compressive) solution for γo = 2 so √ √ MC = 6 then in the range of 1 < MA < 6 solitons solutions exists. Moreover, in this range it is easy to see that the amplitude rises from B = 1 at x → +∞ to maximum magnetic filed Bmax and then returns to B = 1 at x → −∞, not reaching the singularity when the denominator of the pseudopotential equal to zero or the plasma velocity. The compressive solitons from the right side of figure (4.2) is well known and fully described in the literature [e.g. Kennel & Pellat 1976]. But what happens when the magnetic field reaches zero, the figure (4.2) tell us that the magnetic field continue to decrease, however we know this is wrong, because the magnetic field cannot be negative, so there should be a cutoff when magnetic field reaches zero. 22 4.2 Cold Plasma Non Classical Solutions The problem when the magnetic field reaches zero can be avoid if we express every parameters as function of charge density, because charge density have positive values when B = 0. Equation (4.6) will take the following form: (∂X N )2 = −Ψ(N ) (4.10) where Ψ(N ) ≡ − γo2 [1 + γo2 −1 2 (1 MA − B)] − 1 + γo2 γo2 −1 1 2 N2 4 MA 2 γo2 −1 N2 (∂N B)2 (4.11) The potential is pure function of the density (the prefixes pseudo and charge were neglected for simplicity), because the magnetic field is pure function of the density as well as the derivative of the magnetic field with respect to density. 2MA2 1 2 B = 1+ 2 1− γo N 2 MA 1 ∂N B = γo2 BN 2 (4.12) (4.13) For solitary wave to exist conditions I-IV should be satisfied. Condition I, Ψ(1) = 0, is automatically satisfied by the initial conditions as well as condition II, ∂N Ψ(1) = 0. Condition III, ∂N N Ψ(1) < 0, yields non surprising result, that is MA2 > 1. Condition IV, Ψ(N ) = 0, yields four possible roots, which are marked with subscript 1, 2, max and min: N1 = 0 N2 = 1 Nmax = Nmin = MA2 p = N (Bmax ) 2 − MA2 − 2γo2 + 2 γo2 (−1 + MA2 + γo2 ) 1 1+ γo2 2 2MA N2 is the usual one which we could expect where as N1 is non physical. Nmax correspond to Bmax which is not surprising. Nmin obtained by setting the magnetic field to be zero, because Ψ ∝ (∂N B)−1 ∝ B. Those are well behaved as long as the singularity in the denominator of equation (4.10) is avoided. The denominator is proportional to plasma 23 velocity, U , so the singularity corresponds to U → 0 and N → ∞ as in previous section, p so the critical Mach number is the same, MC = 2(1 + γo ), this is the maximum Mach number permissible for solitary solutions. Using this critical Mach number MC we can find the largest possible magnetic field amplitude, density and the smallest possible density. ∗ ∗ ∗ = = ∞, Nmin = 1 + 2/γo , Nmax Bmax 4(1 + γo ) (2 + γo )2 (4.14) ∗ ∗ = 8/9 = 3 and Nmin In the non relativistic limit (4.8) and (4.14) reduces to MC = 2, Bmax when γo = 1. While the relativistic limits permits an apparently much larger range of √ permissible Mach numbers and density, that is 1 < MA < 2γo and N > 4γo−1 , than the non relativistic limit, this whole range corresponds to solitary waves of very small √ maximum amplitudes , 1 < B ∗ < 1 + 2/γo . Setting γo = 2 and 1 < MA < 6 we obtain two solitons, compressive for N > 1 and rarefactive for 0 < N < 1 figure (4.3). (a) (b) Figure 4.3: Two curves for different MA and γo = 2. Each curve represents two solitons compressive from the right and rarefactive from the left. (a) Zoom in on 0 < N < 1. Rarefactive solitons and (b) Zoom in on 1 < N < 6. Compressive solitons. In figure (4.3) the intersection points with x axis are corresponding to Nmin from the left of N = 1 and to Nmax from the right of N = 1, which are increased when Mach number is growing up. It should be noted that the only condition for existence of compressive solution is that the Mach number should be higher than one and less than the critical Mach number, 1 < MA < MC , as demonstrated in figure(4.4a) by upper solid 24 curve. While for rarefactive solutions the Mach number should be higher than one only, MA > 1, and can exceed the critical mach number as it shown in figure(4.4b) by lower solid curve. However, for Mach numbers that less then one there are no solitons of any kind, (4.4) dashed curve. (a) (b) Figure 4.4: Same Plot for different N where γo = 3 and MA = 0.9, 2, 4. Each curve represent different regime. Dashed curve for MA < 1 no solitons of any kind. Solid upper curve for 1 < MA < MC both rarefactive and compressive solitons are shown. Solid lower curve for MA > MC only rarefactive solitons is shown. (a) Zoom in on 0 < N < 1 and (b) Zoom in on 1 < N < 6. To understand why the rarefactive solution is less restricted one has to remember that the restriction of MA < MC is coming from the singularity of the pseudopotential which corresponds to N → ∞ where as the rarefactive solution corresponds to Nmin < N < 1. Finally, we describe in details the magnetic profile, figure(4.5). Compressive solitary waves exist when 1 < MA < MC , in this regime magnetic field amplitude rises from B = 1 at x → +∞ to maximum magnetic filed Bmax and then returns to B = 1 at x → −∞ where the first derivative of the magnetic field with respect to normalized distance at B = Bmax is zero, that is ∂X B(Bmax ) = 0, meaning magnetic profile looks like a parabola in the vicinity of Bmax . Increasing MA leads to the increase in the amplitude of the maximum magnetic field until MA → MC . Moreover, the profile has cusp behavior ∗ ∗ at B = Bmax , because [∂X B(Bmax )]2 > 0. 25 Rarefactive solitary waves exist when MA > 1, in this regime magnetic field amplitude decreases from B = 1 at x → +∞ to zero magnetic filed and then returns to B = 1 at x → −∞. The profile has cusp behavior at B = 0, because [∂X B(0)]2 > 0 according to Eq(4.6). Figure 4.5: Magnetic profile sketch. The upper curve is compressive soliton while the lower is rarefactive soliton. 4.3 Hot Plasma In the hot pair plasma both species have identical ratio of pressure far upstream to the magnetic pressure far upstream, that is β1 = β2 = β/2, because we have assumed that both species have the same pressure far upstream, po1 = po2 = po , and same polytropic indexes, Γ1 = Γ2 = Γ . Hence, both species have same enthalpy, W1 = W2 = W/2 and far upstream Wo1 = Wo2 = Wo /2. The equations read N √ 2V γo σ W γ = γo [1 + σ (1 − B)] ∂X B = (4.15) (4.16) 2βo2 W 2 Γ B = 1 + β(1 − N ) + 1− σ N 2 Γ βσγo W = 1+ N Γ−1 − 1 p Γ−1 2 γ = 1 + βo2 γo2 /N 2 + V 2 26 (4.17) (4.18) (4.19) Using ∂X B = (∂X N )(∂N B) it is easy to derive: (∂X N )2 = −Ψ(N ) (4.20) where Ψ(N ) is the pseudopotential Ψ(N ) ≡ γo2 [1 + γo2 −1 2 (1 MA − B)]2 − W 2 1 + 2 2 −1 W2 o − γ4o γM 2 N2 A 2 (∂N B) γo2 −1 N2 ≡ f1 (N ) f2 (N ) (4.21) where Γ βγo2 (γo2 − 1) Γ−1 N − 1 Γ−1 2MA2 Γβγo2 (γo2 − 1) Γ−2 ∂N W = N 2MA2 2MA2 W 2 Γ B = 1 + β(1 − N ) + 2 1− γo N 2 2MA N ∂N W − W 1 Γ−1 −ΓβN − 2 ∂N B = 2B γo N2 W = 1+ (4.22) (4.23) (4.24) (4.25) For solitary wave to exist conditions I-IV should be satisfied. Condition I, Ψ(1) = 0, is automatically satisfied by the initial conditions. For condition II we need the first derivative of the pseudopotential, that is f10 (N )f2 (N ) − f1 (N )f20 (N ) f22 (N ) γo2 − 1 γo2 − 1 0 2 f1 (N ) = 2γo 1 + (1 − B) (−) ∂N B MA2 MA2 2 γo2 − 1 2 (γo − 1) ∂ W + 2W −2W 1 + N N2 N3 2 2 γ γ − 1 W2 f20 (N ) = −∂N o o 2 (∂N B)2 4 MA N 2 Ψ0 (N ) = (4.26) (4.27) (4.28) Prime is derivative respect to the density. It easy to see that f20 (1) = 0, because Ψ(1) = 0. Imposing initial conditions together with Eq(4.23) and Eq(4.25) we obtain that f10 (1) = 0. Hence, condition II is fulfil, ∂N Ψ(1) = 0. For condition III we need the second derivative 27 of the pseudopotential, that is 0 f1 (N )f20 (N ) f100 (N ) f10 (N )f20 (N ) + (4.29) ∂N N Ψ(N ) = − + f2 (N ) f22 (N ) f22 (N ) 2 2γo2 (γo2 − 1) γo2 − 1 γo − 1 2 00 f1 (N ) = − 1+ (1 − B) ∂N N B − (∂N B) MA2 MA2 MA2 γo2 − 1 γo2 − 1 2 −2 1 + (∂ W ) + W ∂ W + 4W ∂N W N N N N2 N3 2W N 3 ∂N W − 3N 2 W 2 2 (4.30) +2(γo − 1) N6 where the second derivative of the enthalpy and the magnetic field are produced from Eq(4.23) and Eq(4.25), respectively, that is ∂N N W = Γ(Γ − 2)βγo2 (γo2 − 1) Γ−3 N 2MA2 (4.31) 2B∂N N B = −2 (∂N B)2 − Γ(Γ − 1)βN Γ−2 N 2 ∂N W − 2N W 2MA2 N ∂N N W − ∂N W − − 2 γo N2 N4 (4.32) f 00 (1) Remembering that f1 (N ) = f10 (N ) = 0 at N = 1, so ∂N N Ψ(1) = − f12 (1) . Imposing boundary condition we obtain γo2 − 1 Γβγo4 Γβγo4 00 2 2 f1 (1) = 2 2 2 MA − MA − 1 − γo MA 2 2 2 Γβγo4 γo2 − 1 2 M − f2 (1) = A 2γo2 MA2 2 (4.33) (4.34) Condition III, ∂N N Ψ(1) < 0, is fulfilled at two regimes, that is a. MA2 − Γβγo4 2 > 0 and MA2 − 1 − Γβγo4 2 >0 b. MA2 − Γβγo4 2 < 0 and MA2 − 1 − Γβγo4 2 <0 The β parameter is related to the magnetic field. In order to characterize the plasma temperature we introduce T = β= po1 +po2 , 2mc2 no so that 2MA2 T Γ 2 2 γo (γo − 1) 1 + Γ−1 T 28 The two regimes ( a and b ) take the following form: a. 1 − T /T ∗ > 0 and MA2 (1 − T /T ∗ ) − 1 − ΓT Γ−1 >0 b. 1 − T /T ∗ < 0 and MA2 (1 − T /T ∗ ) − 1 − ΓT Γ−1 <0 where 1 T∗ =Γ h γo2 γo2 −1 − 1 Γ−1 i . We do not deal specially with the second regime, (b), because if 1 − T /T ∗ < 0 then condition III is fulfilled regardless the Mach number. According to the first regime, (a), the minimum Mach number can be found for solitary waves to exists, that is s MA > MA,min ≡ Γ 1 + Γ−1 T (1 − T /T ∗ ) (4.35) if T ∗ < 0 or T ∗ > T > 0, where former is true for Γ < 2 and γo2 > true when Γ ≥ 2 or Γ < 2 and γo2 < 1 . 2−Γ 1 2−Γ while latter is Equation (4.35) can be reduced to the cold case by setting T = 0, that is MA > 1. Let us summarize when condition III, ∂N N Ψ(1) < 0, is fulfilled : III.a) T /T ∗ > 1 no restriction on MA III.b) Γ < 2, γo2 > 1 , 2−Γ MA > MA,min III.c) Γ < 2, γo2 < 1 , 2−Γ MA > MA,min , T ∗ > T III.d) Γ ≥ 2, MA > MA,min , T ∗ > T However, condition IV, Ψ(N ) = 0, can not be solved analytically. Nevertheless, we can plot the pseudopotential using Mathematica for different set of parameters, that is Γ, T, γo and MA , which be able to predict if there are solitary waves. For this purpose we present four examples, where Γ, T, γo remain constants while MA vary. Each example is exhibit one of the four cases of condition III. 29 4.3.1 Singularity & Maximum Amplitude We can determine the maximum density, that is Nmax , at the center of a soliton, X = 0, by setting the pseudopotential to zero, that is f1 (Nmax ) = 0 from Eq(4.21). However, the exact expression of Nmax can not be found as function of T, Γ, MA and γo , since f1 (Nmax ) = 0 is transcendental equation, in contrast to the cold case where Nmax = h i−1 p MA2 2 − MA2 − 2γo2 + 2 γo2 (−1 + MA2 + γo2 ) . Numerically solving it, one should remember that according to condition IV (defined in chapter 4) Nmax ≤ Nsp , here Nsp is a singular point where pseudopotential diverge. Moreover, as MA increases, Nmax increases figures (4.6c, 4.7c, 4.8c) and (4.9a) until Nmax = Nsp the corresponding MA is MC . We can determine the singular point, that is Nsp , by letting denominator of pseudopotential be equal to zero, f2 (N ) ∝ [(∂N B)W/N ]2 = 0 from Eq(4.21). Magnetic field is restricted by the pressure term in Eq(4.24), thus maximum density must be finite, in contrast to the cold case where it diverge. Hence, W/N 6= 0 and the only term which can be zero is the derivative of magnetic filed with respect to the density, that is ∂N B(Nsp ) = 0, so Nsp is a solution to N Γ (γo2 − 1)(Γ − 2) + N Γ+2 (Γ − 1) = (γo2 − 1)(Γ − 1) N ΓT (4.36) This transcendental equation cannot be solved by algebraic methods. Nevertheless, for Γ ≥ 2 the left and right functions of Eq(4.36) are monotonically increasing functions, thus the intersection point corresponds to Nsp . It is easy to see that MA has no influence on Nsp , while increasing γo or decreasing T enlarges the value of Nsp . Moreover, for 1 < Γ < 2, as N increases the left side of Eq(4.36) decreases monotonically, reaches a q 2 o −1)(2−Γ) minimum at N = (γ(Γ−1)(Γ+2) , and then increases monotonically eventually intersecting with the right side of Eq(4.36). Hence, for ultrarelativistic plasma, γo 1, Nsp is of order γo as well as Nmax when MA = MC . 30 4.3.2 Numerical Analysis Example 1 (IIIb) (a) (b) (c) Figure 4.6: Pseudopotential as function of density which fulfill IIIb, where Γ = 1.8, T = 0.7, MA,min ≈ 1.49, MC ≈ 2.143, γo = 3 and MA = 1.4, 1.85, 2, 2.5 corresponding to curves 1,2,3,4 respectively. (a) Zoom in on 0 < N < 1. Rarefactive solitons exist when MA > MA,min curves 2,3,4, (b) Zoom in on 1 < N < 2. Compressive soliton exists when MA,min < MA < MC , moreover, as MA increases Nmax increases curves 2,3. (c) Zoom in on 1 < N < 12. Nsp ≈ 2.29 for all curves . 31 Example 2 (IIIc) (b) (a) (c) Figure 4.7: Pseudopotential as function of density which fulfill IIIc, where Γ = 1.9, T = 0.2, MA,min ≈ 1.19, MC ≈ 1.875, T ∗ ≈ 37.89, γo = 3 and MA = 1.1, 1.7, 1.8, 2.5 corresponding to curves 1,2,3,4 respectively. (a) Zoom in on 0 < N < 1.1. Rarefactive solitons exist when MA > MA,min curves 2,3,4, (b) Zoom in on 1 < N < 2.5. Compressive soliton exists when MA,min < MA < MC , moreover, as MA increases Nmax increases curves 2,3. (c) Zoom in on 1 < N < 20. Nsp ≈ 3 for all curves . 32 Example 3 (IIId) (b) (a) (c) Figure 4.8: Pseudopotential as function of density which fulfill IIId, where Γ = 2, T = 0.9, MA,min ≈ 1.19, MC ≈ 2.491, T ∗ ≈ 7.5, γo = 4 and MA = 1.6, 2, 2.1, 2.5 corresponding to curves 1,2,3,4 respectively. (a) Zoom in on 0 < N < 1.1. Rarefactive solitons exist when MA > MA,min curves 2,3,4, (b) Zoom in on 1 < N < 1.5. Compressive soliton exists when MA,min < MA < MC , moreover, as MA increases Nmax increases curves 2,3. (c) Zoom in on 1 < N < 15. Nsp ≈ 2.03 for all curves . 33 Example 4 (IIIa) (a) (b) Figure 4.9: Pseudopotential as function of density which fulfill IIIa, where Γ = 2.1, T = 0.9, T ∗ ≈ 0.88, γo = 1.7 and MA = 0.8, 1.5, 2 corresponding to curves 1,2,3 respectively. (a) Zoom in on 0 < N < 1.1. Rarefactive solitons do not exists, since condition IV is not fulfilled for all curves and Nsp ≈ 0.944. (b) Zoom in on 1 < N < 4. Compressive soliton exists regardless the value of MA , as MA increases Nmax increases for all curves. 4.4 Magnetic Profile & Pair Plasma Summarize Cold Pair Plasma • Compressive solitons emerge when 1 < MA < MC = p 2(1 + γo ), figure (4.4a), rarefactive solitons emerge when MA > 1, figure (4.4b). • As MA increases, the magnetic field amplitude of the compressive soliton increases, figure (4.10a), until R = MA /Mc = 1, where cusp appears (see figure (4.11a)), since ∗ [∂x B(Bmax )]2 > 0. The minimum magnetic field for rarefactive solitons is zero and the magnetic field profile has a cusp for any value of MA , (figure (4.10b)), since [∂x B(0)]2 > 0. • The maximum magnetic field and density for compressive solitons when R 6= 1 are q MA2 2 2 2 p Bmax = MA − 1 + γo − 1 , Nmax = γo 2 − MA2 − 2γo2 + 2 γo2 (−1 + MA2 + γo2 ) 34 • The largest possible magnetic field amplitude is reached when R = 1 ∗ Bmax = 1 + 2/γo ∗ which corresponds to U → 0 and Nmax → ∞. The non relativistic limit permits a much larger range of magnetic field amplitudes, B ≤ 3. Fast waves propagating perpendicular to magnetic field are restricted to low amplitudes in the relativistic limit. (a) (b) Figure 4.10: Magnetic profiles for γo = 10 with different Mach numbers, R = MA /MC = 1, 0.8, 0.6, 0.4 . (a) Compressive solitons and (b) Rarefactive solitons. (a) (b) Figure 4.11: Zooming on Magnetic profiles for γo = 10 at B = Bmax . (a) For R = 1 there is a cusp behavior (b) For R 6= 1 (e.g. R = 0.8) there is no cusp, since ∂X B(Bmax ) = 0. 35 Hot Pair Plasma • For solitary waves to exists conditions I-IV should be satisfied. Condition III is fulfilled in two different regimes (see examples 1-4 from Chapter 4), that is a) T /T ∗ > 1 - no rarefactive solitons (figure (4.9a)), compressive solitons emerge regardless the value of MA , figure (4.9b). We do not deal in detail with this regime, since it valid for Γ ≥ 2 or when Γ < 2 and γo2 < 1 . 2−Γ b) T /T ∗ < 1 and MA > MA,min - rarefactive solitons emerge when MA > MA,min while compressive solitons exist when MA,min < MA < MC , (see examples 1-3). r h i where 1 T∗ =Γ γo2 γo2 −1 − 1 Γ−1 and MA,min ≡ Γ T 1+ Γ−1 . (1−T /T ∗ ) Latter can be reduced to the cold case by setting T = 0, that is MA,min = 1. There is a critical Mach number MC , figures (4.6)-(4.8), however it can not be found analytically because of complexity of our equations. ∗ has a finite value at R = 1, in contrast to the cold case where • Maximum density Nmax it diverges. This maximum density increases with the increase of γo or decrease of ∗ T . Moreover, for ultrarelativistic plasma γo 1, Nmax is of order γo . • As MA increases while Γ, γo and T remain constant, the magnetic field amplitude increases (figure (4.12a)), until R = 1 for compressive solitons. At R = 1 the magnetic profile has a cusp which is related to the singularity of the pseudopotential there. An increase of MA leads to the decrease of the width of the soliton at half maximum of the magnetic field for compressive solitons and rarefactive solitons as well (figure (4.12b)). • As γo increases while Γ, R and T remain constant, the magnetic field amplitude decreases, figure (4.13a,b). Moreover, same increase of γo leads to increase of the width of rarefactive solitons figure (4.13c), while the width of the compressive solitons decreases. • Decreasing T and keeping Γ, R and γo as constants leads to the increase in the magnetic field amplitude (figure (4.14a)). Moreover, same decrease leads to the 36 increase of the width for compressive solitons while the width of the rarefactive solitons does not change, figure (4.14b). • For the reasons above we can confirm that fast waves propagating perpendicular to magnetic field are restricted to low amplitudes in the hot relativistic limit as well as cold. (a) (b) Figure 4.12: Magnetic profiles for γo = 10, Γ = 4/3, T = 1 and different Mach numbers where MA,min = 1.169 and MC = 3.579. (a) Compressive solitons for MA = 1.8, 2, 3 (b) Rarefactive solitons for MA > 1.16. It should be noted that for comparison of the the typical length scales (in physical units) of the hot solitons with the cold ones, one must multiply the width by the factor q Γ 1 + Γ−1 T , because of the normalization of X. 37 (a) (b) (c) Figure 4.13: Magnetic profiles for R = 0.5, Γ = 4/3, T = 1 and different γo . (a) Compressive solitons, (b) Zoom in on the center of a compressive solitons, no cusp behavior. and (c) Rarefactive solitons. (a) (b) Figure 4.14: Magnetic profiles for γo = 10, Γ = 4/3, R = 0.5 and different T . (a) Compressive solitons (b) Rarefactive solitons . 38 Chapter 5 Electron-Ion Plasma In this chapter we examine relativistic electron-ion plasmas in the perpendicular regime using the pseudopotential method. All equations, (3.9)-(3.19), and boundary conditions that were deduced in chapter 3 are applicable here, as well as conditions for existence of solitons, I-IV, from chapter 4. W1 γ1 + W2 γ2 = γo [1 + σ (1 − B)] (5.1) B 2 − E 2 = 1 + β1 (1 − N Γ1 ) + β2 (1 − N Γ2 ) + W1 V1 + W2 V2 = σγo E √ 2γo2 σ(1 + σ)Φ = − W12 − W22 1 + βo2 γo2 /N 2 Γs βs σγo2 N Γs −1 − 1 Ws = Wos + Γs − 1 2 N ∂X B = − √ (V1 − V2 ) γo σ N √ (γ1 − γ2 ) ∂X E = γo σ and X N, U , Vs B, E Ps , P Ws , W X → −∞ 1 ,1 , 0 → 1, 0 1, 2 c2 ms wo + Γs βs σγo2 Γs −1 2 39 , 1 2βo2 σ (1 − W/N ) (5.2) (5.3) (5.4) (5.5) (5.6) (5.7) The lower indexes 1 and 2 stand for ion and electron, respectively. First, let us express the normalized electric field, E, as function of density and it’s derivatives with respect to normalized distance, using the expression for the electric potential, Eq(5.4), and E = −∂X Φ. E = (∂X N )∂N (W12 − W22 ) (1 + γo2 βo2 /N 2 ) √ ≡ (∂X N )(∂N Φ) 2γo2 σ (1 + σ) (5.8) Using Ampere’s law, Eq(5.6), and conservation of longitudinal momentum, Eq(5.3), we can express V1 and V2 as function of N, E, and ∂X B. γo σ W2 V1 = E−√ ∂X B W σN γo σ W1 V2 = ∂X B E+√ W σN (5.9) (5.10) Besides the quasineutrality assumption we shell assume that magnetic pressure is much bigger than electric pressure, that is B 2 E 2 , so the conservation of transverse momentum reads, Eq(5.2): B 2 = 1 + β1 (1 − N Γ1 ) + β2 (1 − N Γ2 ) + 2βo2 (1 − W/N ) σ (5.11) and 2βo2 N ∂N W − W 1 Γ1 −1 Γ2 −1 ∂N B = −Γ1 β1 N − Γ2 β2 N − 2B σ N2 Γs βs σγo2 Γs −2 N ∂N Ws = 2 (5.12) (5.13) Moreover, we can express V1 and V2 as function of N and ∂X N , that is Vs = gs ∂X N (5.14) where g1 g2 γo σ ≡ ∂N Φ − W γo σ ≡ ∂N Φ + W W2 √ ∂N B σN W1 √ ∂N B σN (5.15) (5.16) Using the definition of Lorentz factor, γs2 = 1 + βo2 γo2 /N 2 + Vs2 , in conservation of energy, Eq(5.1) together with equation (5.14) we obtain q q W1 1 + βo2 γo2 /N 2 + g12 (∂X N )2 + W2 1 + βo2 γo2 /N 2 + g22 (∂X N )2 = γo [1 + σ (1 − B)] 40 Squaring above equation twice to eliminant the roots we have Λ(∂X N )4 − 2Π(∂X N )2 + Ξ = 0 (5.17) where Λ(N ) = (W12 g12 − W22 g22 )2 βo2 γo2 2 2 2 2 (W1 + W2 ) (W12 g12 + W22 g22 ) Π(N ) = γo [1 + σ(1 − B)] − 1 + 2 N 2 2 β γ +2W12 W22 1 + o 2o (g12 + g22 ) N 2 βo2 γo2 2 2 2 2 Ξ(N ) = γo [1 + σ(1 − B)] − 1 + (W1 + W2 ) N2 2 βo2 γo2 2 2 −4W1 W2 1 + N2 (5.18) (5.19) (5.20) Equation (5.17) is usual quadratic equation with respect to (∂X N )2 having the following solutions 2 (∂X N ) = Π± √ Π2 − ΛΞ Λ We can neglect the plus solution of equation (5.17) because it does not satisfy the boundary conditions, that is Ξ = 0 and Π > 0 when N = 1. Finely (∂X N )2 = −Ψ(N ) where Ψ(N ) is the pseudopotential which is a pure function of density, defined as √ Π − Π2 − ΛΞ Ψ(N ) ≡ − (5.21) Λ For solitary wave to exist conditions I-IV should be satisfied. Condition I, Ψ(1) = 0, is automatically satisfied by the initial conditions as well as condition II, Ψ0 (1) = 0, because Ξ = 0, Π > 0 and Ξ0 = 0 when N = 1. The latter is somewhat tedious but straightforward (see Appendix A). Prime symbolize derivative with respect to the density. For condition III we need to differentiate twice the pseudopotential, Eq(5.21), and impose initial conditions, thus Ψ00 (1) = − Ξ00 (1) 2Π(1) (5.22) 41 According to Eq(5.19) Π(1) is positive, that is 2 2 >0 + Wo2 go2 Π(1) = 2γo2 Wo1 Wo2 Wo1 go1 and 2 2 Ξ00 (1) = 4γo4 (Wo1 + Wo2 )δ(1) where (for details see Appendix A) βo2 γo4 γo4 2 2 δ(1) = MA − (β1 Γ1 + β2 Γ2 ) MA − 1 − (β1 Γ1 + β2 Γ2 ) MA2 γo2 2 2 The only requirement that condition III, Ψ00 (1) < 0, will be fulfilled is δ(1) > 0. We express βs in terms of Ts = β1 = pos , ms c2 no 2T1 /γo2 σ 1 1 + Γ1Γ−1 T1 + µ1 1 + so that Γ2 T Γ2 −1 2 , Above relations are easily obtained using δ(1) = β2 = 4π Bo2 = 2T2 /γo2 σ 1 µ 1 + Γ1Γ−1 T1 + 1 + 1 no wo γo2 σ Γ2 T Γ2 −1 2 where µ = m1 /m2 . Hence, βo2 2 2 G M − 1/G A γo2 where G≡1− γo2 γo2 − µT1 Γ1 + T2 Γ2 1 µ 1 + Γ1 T + 1 + Γ1 −1 1 Γ2 T Γ2 −1 2 (5.23) Condition III, is fulfilled in two cases: a. G < 0 b. G > 0 and MA2 > 1/G In the cold limit, Ts = 0, where G = 1 only the second case of condition III is valid, hence MA > 1. Condition IV, Ψ(N ) = 0, can not be solved analytically in the hot limit. Nevertheless, we can plot the pseudopotential using Mathematica for different set of constant values, that is Γ1 , Γ2 , T1 , T2 , γo and MA , which be able to predict if there are solitary waves. However, in the cold limit when Ts = 0 condition IV, Ψ(N ) = 0, can be 42 solved analytically. According to Eq(5.21) there is such N which set the pseudopotential to zero, when Ξ = 0, where βo2 γo2 2 2 · Ξ(N ) = γo [1 + σ(1 − B)] − 1 + N2 βo2 γo2 2 2 2 γo [1 + σ(1 − B)] − 1 + (Wo1 − Wo2 ) N2 2β 2 B 2 = 1 + o (1 − 1/N ) σ It is sufficient to find roots of the first brackets of Eq(5.24), since Wos = (5.24) (5.25) ms m1 +m2 and m1 m2 , by squaring them once together with Eq(5.25), obtaining # 2 2 " 1 σ 1 1− + 1 + 2σ − 4(1 + σ) 2 = 0 N N βo using the definition of σ = βo2 γo2 2 , MA solutions to above are N = 1 Nmax = MA2 − (5.26) MA2 2s2 + (5.27) 2γo s those are the roots of Ξ = 0, where we have the neglected the negative root, since N > 0, where s2 = MA2 + γo2 − 1 and the magnetic field at N = Nmax , according to Eq(5.25) is Bmax = 2s −1 γo Critical Mach number can be found by setting the denominator of the pseudopotential to zero, Π = 0. In the cold limit Π ∝ N −3 f (γo , Wo1 , Wo2 , MA ). Hence, the singularity corresponds to N → ∞ and U → 0 or f = 0. The latter is true for negative MA , so it may be neglected. While the former leads to the same MC which were obtained in pair plasma, that is MC = p 2(1 + γo ) which in turn leads to same results that where obtain in pear plasma at least for compressive solitons. 43 5.1 Numerical Analysis Example 1 (Cold Plasma) The setup is as follows: Ts = 0, γo = 2, m1 /m2 = 2000 and MA = 0.9, 2, 2.5 so MC ≈ 2.45. According to condition III there should be no solitons for MA = 0.9 < 1 see figure (5.1a), continuous curve. Compressive solitons do exist for MA = 2 < MC , since conditions I-IV are fulfilled, as demonstrated in figure (5.1) by dashed curve, where the intersection of the x axis is corresponds to Nmax = 6.86. For MA = 2.5 > MC there are no solitary waves see figure (5.1b), dotted-dashed curve, since Λ = 0. In our model for cold electron ion plasma we can not confirm or disconfirm existence of rarefactive solitons, since the assumption of B 2 E 2 is no longer valid see figure (5.2a), where E 2 /B 2 is plotted as function of N . (a) (b) Figure 5.1: Same Plot for different N where γo = 2, Ts = 0. Dashed, dotted-dashed and continuous curves corresponds to MA = 2, 2.5, 0.9, respectively. (a) Zoom in on 0 < N < 1.5 and (b) Zoom in on 1 < N < 7. As a result to our assumption there is a minimum density, Nmin = 1 1+ 2 γo 2M 2 A , which cor- responds to zero magnetic field. Otherwise, according to Eq(5.25) magnetic field becomes imaginary. Hence, the curves’ cutoff in figure (5.1a) which occurs when N = Nmin . 44 The maximum deviation of E 2 /B 2 , which will be defined as η, in the regions where there are compressive solitons, 1 < MA < MC , is a function of two parameters MA and γo , that is η(MA , γo ). When γo = 2 the deviation η ∼ 1% for MA = 1 and η ∼ 10% for MA = MC see figure (5.2c), dashed curves. Thus, enlargement of MA causes enlargement of η, breaking down our assumption. (a) (b) (c) Figure 5.2: Same Plot for different N where T = 0, γo = 2. Lower dashed, continuous and upper dashed curves corresponds to MA = 1, 2, MC , respectively. (a) Zoom in on 0 < N < 1, (b) Zoom in on 1 < N < 7 and (c) Zoom in on 1 < N < 70. However, small enlargement of γo causes a significant decreasing of η. When γo = 5 the deviation η ∼ 0.006% for MA = 1 and η ∼ 0.73% for MA = MC see figure (5.3), dashed curves. (a) (b) (c) Figure 5.3: Same Plot for different N where T = 0, γo = 5. Lower dashed, continuous and upper dashed curves corresponds to MA = 1, 2, MC , respectively. (a) Zoom in on 0 < N < 1, (b) Zoom in on 1 < N < 7 and (c) Zoom in on 1 < N < 70. 45 Example 2 (Hot Plasma, G < 0) As we mentioned before the roots of Ψ(N ) = 0 in hot plasma can not be found because of complexity of our equations. However, it can be done graphically. In accordance with condition III there are two possible regimes, G < 0 and G > 0, MA2 > 1/G. The former is treated in this section while the letter in the next one. In electron ion plasma µ → 2000 so G, Eq(5.23), highly depend on ions’ temperature and polytropic index. In our model for cold electron ion plasma we can not confirm or disconfirm existence of compressive solitons, since the assumption of B 2 E 2 is no longer valid see Fig.(5.4), when G < 0 which corresponds to high Γ1 > 2, Fig.(5.4a), or low γo → 1, Fig.(5.4d), there is no dependents on Mach number. However, there is no rarefactive solitons. (a) (b) (c) (d) Figure 5.4: Plot for G < 0 where , Γ2 = 1.8, T2 = 0.8, µ = 2000. Dashed and continuous curves corresponds to E 2 /B 2 and Ψ(N ), respectively. (a) Const γo = MA = 2, T1 = 0.9 and Γ1 = 2.7, 2.9 (b) Const MA = 2, T1 = 0.9, Γ1 = 3 and γo = 1.5, 3 (c) Const γo = 2, T1 = 0.9, Γ1 = 3 and MA = 2, 3 (d) Const MA = 2, T1 = 0.8, Γ1 = 1.8 and γo = 1.2, 1.4 . 46 Example 3 (Hot Plasma, G > 0) When G > 0 there is a minimum Mach number for condition III to be satisfied, that is MA2 > 1/G, demonstrated in figure (5.5) where T1 = T2 = 1, Γ1 = Γ2 = 1.6 and γo = 4 so G = 0.53. According to figure (5.5c) when MA = 2.5 and η ∼ 0.1%, there is no compressive soliton because the singularity of the pseudopotential, Λ = 0, is encountered before condition IV is fulfilled. Hence, implies on existence of critical Mach number. Compressive solitons do exist for MA = 2 < MC and η ∼ 0.04%, since conditions I-IV are fulfilled, as demonstrated in figure (5.5b). In our model for cold electron ion plasma we can not confirm or disconfirm existence of rarefactive solitons, since the assumption of B 2 E 2 is no longer valid see figure (5.5d). (a) (b) (c) (d) Figure 5.5: Plot for G > 0 where T1 = T2 = 1, Γ1 = Γ2 = 1.6, γo = 4 so G = 0.53 and MA = 1.3, 2, 2.5. (a) No rarefactive solitons in 0 < N < 1, (b) Zoom in on 1 < N < 2.5 where there is compressive soliton for MA = 2 (c) Zoom in on 1 < N < 7 singularity of Ψ, (d) deviation from E 2 /B 2 1 . 47 5.2 Magnetic Profile & Electron-Ion Plasma Summarize Cold Electron-Ion Plasma • In our model for cold electron ion plasma we can not confirm or disconfirm existence of rarefactive solitons, since the assumption of B 2 E 2 is no longer valid, figure (5.1a). • Compressive solitons emerge when 1 < MA < MC = p 2(1 + γo ), figure (5.1b). • Enlargement of MA causes to enlargement of η, meaning for braking down of our assumption, B 2 E 2 . However, small enlargement of γo causes a significant decrease of η, figure (5.2) and figure (5.3). • Increasing MA leads to increase in the magnetic field amplitude and decrease of the soliton width, figure (5.6), until R = 1 thus causing to cusp behavior, since ∗ )]2 > 0. [∂x B(Bmax • Increasing γo and keeping R as constant leads to decrease in the magnetic field amplitude and the soliton width, figure (5.7). • The largest possible magnetic field amplitude ∗ Bmax = 1 + 2/γo is attain when R = 1, which corresponds to U → 0 and N → ∞ is identical to the Pair plasma case. It is reasonable that the particle inertia be determined by the wave amplitude, not by the rest masses in the relativistic limit. Thus confirming that fast waves propagating perpendicular to magnetic field are restricted to low amplitudes in the relativistic limit. 48 Figure 5.6: Magnetic profiles for γo = 10 with different Mach numbers, R = MA /MC = 0.99, 0.8, 0.6, 0.4 , where for R = 0.99 there is cusp. While l = 0.7, 0.8, 1.1, 1.7 from the highest to the lowest R, respectively. (a) (b) (c) Figure 5.7: Magnetic profiles for R = 0.5 with different γo , γo = 101 , 102 , 103 , 104 . And l = 0.69, 0.7, 0.75, 1.3 from the highest to the lowest γo , respectively. (a) Compressive solitons for γo = 10, 102 , (b) Compressive solitons for γo = 103 , 104 and (c) There is no cusp for any γo . 49 Hot Electron-Ion Plasma • For solitary waves to exists conditions I-IV should be satisfied. Condition III is fulfilled in two different regimes (see examples 2,3 from Chapter 5), that is a) G < 0 - No rarefactive solitons figure (5.4), because the singularity of the pseudopotential is encountered before condition IV is fulfilled. No compressive solitons since the assumption is not valid, figure (5.4). b) G > 0 and MA2 > 1/G - No rarefactive solitons since the assumption, B 2 E 2 is not valid, figure (5.5d). Compressive solitons emerge when MA,min < MA < MC , figure (5.5b,c). γo2 where G ≡ 1 − γ 2 −1 o µ µT1 Γ1 +T 2 Γ2 Γ1 T1 1+ Γ −1 1 Γ2 T2 +1+ Γ −1 2 and MA,min ≡ p 1/G. Latter can be reduced to the cold case by setting T = 0, that is MA,min = 1. It is easy to see that G is controlled by Ions’ parameters till T2 ∝ µT1 . There is a critical Mach number MC , figure (5.5), however it can not be found analytically because of complexity of our equations. • Magnetic profile for compressive soliton has a cusp when MA = MC , this fact is not accompanied by any of our figures, but can easily be explained. The pseudopotential has a singularity when B = Bmax and MA = MC , thus (∂X B)2 6= 0. • According to Table 5.1, Increasing Te , Ti or γo lowers the magnetic field amplitude, thus we confirm that fast waves propagating perpendicular to magnetic field are restricted to low amplitudes in the hot relativistic limit as well as cold. • According to Table 5.1, the critical Mach number is highly influenced by γo . • According to Table 5.2, the width of the soliton at half maximum of the magnetic field amplitude, l, strongly depends on Ti . But, when Te is bigger than µTi , the electron temperature becoming more significant. 50 γo = 102 , η < 10−4 % γo = 101 , η < 0.02% Ti 10−1 Te 10−1 102 10−1 104 1.07563 1.07091 1.05624 3.571 102 3.541 3.632 1.05556 1.05556 1.05555 3.650 104 3.650 3.650 1.05552 1.05552 1.05552 3.651 3.651 3.651 102 104 γo = 104 , η < 10−9 % 10−1 102 1.01255 1.01295 1.01668 1.0003 11.572 11.804 13.741 174.007 190.786 276.701 1.017 1.017 1.017 1.00079 1.00079 1.00079 13.898 13.898 13.898 282.234 282.235 282.249 1.01701 1.01701 1.01701 1.00079 1.00079 1.00079 13.906 282.54 13.906 13.906 1.00036 1.00076 282.54 Table 5.1: The First Raw divided to three which correspond to different γo and having maximum deviation η. The second raw and the first column is Te and Ti , respectively. Each sell have two values where the upper is the maximum permitted magnetic field ampletude and the lower value is the critical Mach number. γo = 101 , η < 10−2 % γo = 102 , η < 10−5 % γo = 104 , η < 10−10 % 10−1 102 104 10−1 102 104 10−1 102 104 10−1 2.5 2.3 11.7 0.99 0.86 3.96 0.8 0.7 3 102 57.4 57.4 57.8 18.8 18.7 17.9 104 574.4 574.4 574.1 188.2 188 187.8 Ti Te 104 14.09 14.1 141 140 13.4 140 Table 5.2: The First Raw divided to three which correspond to different γo and having maximum deviation η. The second raw and the first column is Te and Ti , respectively. Each cell represents a ratio between the width of the hot soliton and the width of the cold soliton. 51 282.54 Chapter 6 Discussions & Conclusions First, let us remind to the reader that we have analyzed stationary one-dimensional nonlinear waves propagating in the direction perpendicular to the external magnetic field, within the two-fluid approach and have covered the cold and hot pair plasma cases as well as electron-ion plasma. The quasineutrality assumption in the species rest frame allowed us to derive a pseudopotential-type equation for the magnetic field or density in the cold as hot cases as well. We reproduce the well-known results regarding compressive solitons, that are summarized in detail at section 4.4, and show that the pair plasma allows also rarefactive solitary wave solutions, which are absent in the literature for the relativistic magnetized pair plasma. In the case of electron-ion plasma we derive a single pseudopotential equation for the density by solving the equations in the approximation of the weak longitudinal electric field. The equation is solved numerically in a wide range of plasma parameters, and summarized in section 5.2. It is found that the compressive solitons are similar to those found for the pair plasma and do not exist for Mach numbers exceeding some critical Mach number or inferior than some minimum Mach number. There are no rarefactive solitons in the electron-ion plasma in this approximation and shock like solution as well. Stability of the solitons is an important issue that deserve attention, but has not been addressed in this study . Moreover, examining some features e.g correction of perpendicular propagation equations to the oblique case, generalization into the anisotropic plasma theory onto generally relativistic cases, consideration to non stationarity together 52 with non one-dimensionality and deviation from quasineutrality are adequate and have a fundamental importance for understanding better the structure of relativistic nonlinear waves and shocks in magnetized plasma. 53 Appendix A A.1 This section is devoted to calculation of Ξ(1), Ξ0 (1) and Ξ00 (1). 2 2 βo2 γo2 βo2 γo2 2 2 2 2 2 2 (W1 + W2 ) − 4W1 W2 1 + Ξ(N ) = γo [1 + σ(1 − B)] − 1 + N2 N2 = (c1 − c2 c3 )(c1 + c2 c4 ) = c5 c6 where c1 = γo2 [1 + σ(1 − B)]2 , βo2 γo2 N2 c1 (1) = γo2 , c2 (1) = γo2 c3 = W 2 , c3 (1) = 1 c4 = (W1 − W2 )2 , c4 (1) = (Wo1 − Wo2 )2 c5 = (c1 − c2 c3 ) , c5 (1) = 0 c6 = (c1 + c2 c4 ) , 2 2 ) + Wo2 c6 (1) = 2γo2 (1 − 2Wo1 Wo2 ) = 2γo2 (Wo1 c2 = 1 + First derivative: Ξ0 (N ) = c05 c6 + c5 c06 c01 = −2γo2 [1 + σ(1 − B)]σB 0 , 2 2 where c01 (1) = −2γo2 σB 0 c02 = −2 βNo γ3o , c02 (1) = −2βo2 γo2 c03 = 2W W 0 , c03 (1) = 2W 0 (1) c05 = c01 − c02 c3 − c2 c03 , c05 (1) = 0 Second derivative Ξ00 (N ) = c005 c6 + c05 c06 + c05 c06 + c5 c006 c001 = 2γo2 [σ 2 B 02 − [1 + σ(1 − B)]σB 00 ] , 2 2 where c001 (1) = 2σγo2 σB 0 2 − B 00 c002 = 6 βNo γ4o , c002 (1) = 6βo2 γo2 c003 = 2W 02 + 2W W 00 , c003 (1) = 2W 0 2 (1) + 2W 00 (1) c005 = c001 − c002 c3 − 2c02 c03 − c2 c003 , c005 (1) = 2γo2 δ(1) where δ(1) = σ 2 B 0 2 − σB 00 − 3βo2 + 4βo2 W 0 − W 0 2 − W 00 . So Ξ(1) = c5 (1)c6 (1) = 0 Ξ0 (1) = c05 (1)c6 (1) + c5 (1)c06 (1) = 0 2 2 Ξ00 (1) = c005 (1)c6 (1) + c05 (1)c06 (1) + c05 (1)c06 (1) + c5 c006 (1) = 4γo4 (Wo1 + Wo2 )δ(1) 54 A.2 This section is devoted to calculation of Ws0 (1), Ws00 (1), W 0 (1) and W 00 (1). 2 s βs σγo Ws = Wos + ΓsΓ−1 N Γs −1 − 1 , Ws (1) = Wos 2 2 Ws0 = Γs βs σγ2o N Γs −2 2 Ws00 = Γs βs (Γs − 2) σγ2o N Γs −3 2 , Ws0 (1) = Γs βs σγ2o , Ws00 (1) = Γs (Γs − 2)βs σγ2o 2 So W (1) = 1 σγo2 2 σγo2 (κ2 − 2κ1 ) = 2 W 0 (1) = κ1 W (1)00 where κ1 = β1 Γ1 + β2 Γ2 and κ2 = β1 Γ21 + β2 Γ22 A.3 This section is devoted to calculation of B 0 (1), B 00 (1) and δ(1). B 2 = 1 + β1 (1 − N Γ1 ) + β2 (1 − N Γ2 ) + 2 B B 0 00 2βo2 (1 − W/N ) σ 0 N −W −β1 Γ1 N Γ1 −1 − β2 Γ2 N Γ2 −1 − 2βσo W N 2 = 2B 02 −2B − β1 Γ1 (Γ1 − 1)N Γ1 −2 − β2 Γ2 (Γ2 − 1)N Γ2 −2 − = 2B So B(1) = 1 β 2 γ 2 κ1 B 0 (1) = o − o σ 2 2 2 2 κ2 γo2 2βo2 κ1 γo βo γo2 κ1 00 2 B (1) = (1 + 3βo ) − − − − 2 2 σ σ 2 Using all above we obtain: 2 2 δ(1) = σ 2 B 0 − σB 00 − 3βo2 + 4βo2 W 0 − W 0 − W 00 = σ σ 4 2 βo4 4 2 2 = βo + κ1 + γo κ1 + − βo γo κ1 − βo2 2 4 σ βo2 κ1 γo4 κ1 γo4 2 2 = MA − MA − 1 − MA2 γo2 2 2 55 2βo2 σ W 00 N − 2W 0 N2 − 2W N3 Bibliography [1] Pearson, T.J., & Zensus, J.A., Superluminal radio sources - Introduction, Cambridge and New York, Cambridge University Press, pp. 1-11 (1987) [2] Biretta, J.A., Sparks W.B., Macchetto F., Hubble Space Telescope Observations of Superluminal Motion in the M87 Jet, ApJ 520(2) pp. 621-626 (1999) [3] Junor W., Biretta J.A., Livio M., Formation of the radio jet in M87 at 100 Schwarzschild radii from the central black hole Nature 401 pp. 891-892 (1999) [4] Mirabel, I.F., Rodrguez, L.F., Microquasars in our Galaxy, Nature 392 pp. 673-676 (1998) [5] Kulkarni S.R., Djorgovski S.G., Odewahn S.C., Bloom J.S., Gal R.R., Koresko C.D., Harrison F.A., Lubin L.M., Armus L., Sari R., Illingworth G.D., Kelson D.D., Magee D.K., Dokkum P.G. Van, Frail D.A., Mulchaey J.S., Malkan M.A., McClean I.S., Teplitz H.I., Koerner D., Kirkpatrick D., Kobayashi N., Yadigaroglu I.A., Halpern J., Piran T., Goodrich R. W., Chaffee F. H., Feroci, M., Costa E., The afterglow, redshift and extreme energetics of the γ-ray burst of 23 January 1999, Nature 398 pp. 389-394 (1999) [6] Ford H.C., Harms R.J., Tsvetanov Z.I., Hartig G.F., Dressel L.L., Kriss G.A., Bohlin R.C., Davidsen A.F., Margon B.; Kochhar A.K., Narrowband HST images of M87: Evidence for a disk of ionized gas around a massive black hole, ApJ 435 pp. L27-L30 (1994) 56 [7] Wardle, J.F.C., Homan D.C., Ojha R., Roberts D.H., Electron-positron jets associated with the quasar 3C279, Nature 395 pp. 457-461 (1998) [8] Piran T., Gamma-ray bursts and the fireball model, Phys.Rep 314 pp. 575-667 (1999) [9] Piran T., Gamma-ray bursts and the fireball model, Phys.Rep 333 pp. 529-553 (2000) [10] Mészáros P., Gamma-Ray Bursts and Bursters, Nucl.Phys.Proc.Suppl. 80 pp. 63-77 (2000) [11] Ulrich M.H., Maraschi L.,, Urry C.M., Variability of Active Galactic Nuclei, ARA&A 35 pp. 445-502 (1997) [12] Angel J.R.P., Stockman H.S., Optical and infrared polarization of active extragalactic objects, ARA&A 18 pp. 321-361 (1980) [13] Covino S., Lazzati D., Ghisellini G., Saracco P., Campana S., Chincarini G., di Serego S., Cimatti A., Vanzi L., Pasquini L., Haardt F., Israel G.L., Stella L., Vietri M., GRB 990510: linearly polarized radiation from a fireball, A&A 348 pp.L1-L4 (1999) [14] Wijers R.A.M.J., Vreeswijk P.M., Galama T.J., Rol E., van Paradijs J., Kouveliotou C., Giblin T., Masetti N., Palazzi E., Pian E., Frontera F., Nicastro L., Falomo R., Soffitta P., Piro, L., Detection of Polarization in the Afterglow of GRB 990510 with the ESO Very Large Telescope, ApJ 523 pp. L33-L36 (1999) [15] Rol E., Wijers R.A.M.J., Vreeswijk P.M., Kaper L., Galama T.J., van Paradijs J., Kouveliotou C., Masetti N., Pian E., Palazzi E., Frontera F., van den Heuvel E.P.J., GRB 990712: First Indication of Polarization Variability in a Gamma-Ray Burst Afterglow, ApJ 544 pp. 707-711 (2000) [16] Granot J., Königl A., Radiative Hydromagnetic Shocks in Relativistic Outflow Sources, ApJ 560 pp. 145-159 (2001) [17] Karpman V.I., Kaufman R.N., Shagalov A.G., Self-focusing of whistler waves, Phys.Fluids B 4 pp. 3087-3100 (1992) 57 [18] Kennel C.F., Pellat R., Relativistic nonlinear plasma waves in a magnetic field, J.Plasma.Phys 15 pp. 335-355 (1976) [19] Aslop D., Arons J., Relativistic magnetosonic solitons with reflected particles in electron-positron plasmas, Phys Fluids 31 pp. 839-847 (1988) [20] Kivelson M.G., Russell C.T , Introduction to Space Physics, Cambridge University Press (1996) [21] Medvigy D., Loeb A., Steady-state structure of relativistic collisionless shocks, Phys.Rev.E [22] Chiueh T., Lai T.C., Magnetosonic soliton in a relativistically hot plasma, Phys.Rev.A 44 pp. 6944-6947, (1991) [23] Lakhina G.S., Verheest F., Alfvenic Solitons in Ultrarelativistic Electron-Positron Plasmas, Ap&&SS 253 pp. 97-106, (1997) [24] Farina D., Bulanov S.V., Dark solitons in electron-positron plasmas, Phys.Rev.E 64 066401, (2001) [25] Lichnerowicz A., Relativistic hydrodynamics and magnetohydrodynamics : lectures on the existence of solution, New York, Benjamin Inc , (1967) [26] Landau L.D., Lifshitz E.M., Fluid Mechanics, Pergamon, New York , (1986) [27] Gedalin M., Linear waves in relativistic anisotropic MHD, Phys. Rev E47 pp. 43544357, (1993) [28] Pillay R., Bharuthram R., Large amplitude solitons in a multi-species electronpositron plasma, Ap&&SS 198 pp. 85-93, (1992) [29] Lee N.C., Ion-acoustic solitary waves in fully relativistic electron-ion plasmas, Phys.Plas. 15 022307, (2008) 58