![Arxiv:1703.07714V1 [Astro-Ph.HE] 22 Mar 2017 O Xml,Tejasaayi Ntesraeo Ro- Crite- of Toomre Surface So-Called Account](http://data.docslib.org/img/fa66b1f03cc73f8ea1713cf23925a44a-1.webp)
arXiv:1703.07714v1 [astro-ph.HE] 22 Mar 2017 o xml,teJasaayi ntesraeo ro- crite- of Toomre surface so-called account. the the into to mo- on leads rion, taken angular analysis galaxies be Jeans disk the tating should the of cri- system effects example, a complicated. For the stabilizing finding more of the much mediums, mentum case is rotating this instability the In Jeans var- In for the account investi- into terion i.e., widely taking effects. by length, been situations ious has specific different instability a in This gated than collapse. when larger length. will amplified Jeans are mass are scales perturbations Jeans perturbations prevent small their the words, case, not than other this can In larger In pressure masses gas with collapse. internal gravitational In the the hap- environment instability. when gravita- non-rotating Jeans in pens the so-called instability is Jeans the clouds fact, or gas instability interstellar tional the in mentation ain n iuain osqetyitrs nthe in interest Consequently obser- in reveal simulation. structures and Theses vations structures. mentary instabil- magneto-Jeans the as ity. known is instability This by studied been has dissipation gas with disk galactic field tidal in external studied an Gravita- been of stability. presence has the gravitational in local instability the tional to viscosity of , stud- in been has ied disk galactic multicomponent for criterion Elmegreen ( ( ( Toomre oicueteeeto h hcns fteds,see disk, the of thickness the of effect the include to esoscutr h tblzn ffc fteCoriolis over- the ( the of of regions collapse self-gravitating dense effect the allows stabilizing and force, the counters tensions 1996 2011 1992 o Solomon & Jog n ftekyfcosfrsa omto n frag- and formation star for factors key the of One h en nlsshsas enivsiae o fila- for investigated been also has analysis Jeans The two-component in stability local the Furthermore rpittpstuigL using typeset Preprint . ) , ) , ) Toomre Kato ytm n h osblt flclfamnain nsm relativist hig some to in applied fragmentations are local results of Keywords: the possibility Finally the lim and it. this systems preventing in Interestingly, of to instead increases. th mass instability system between Jeans the new difference c the of fractional and make ature/pressure the derived effects Furthermore, is relativistic mass. mass the Jeans Jeans limit, new this A found In is techniques. mass. perturbations different two small wh using of what systems tion than propagation for higher for othe are relation considered In field dispersion are gravitational limit. instability corresponding post-Newtonian and gravitational first system the local in the studied on is analysis Jeans The hder Khajenabi & Shadmehri ( 1964 Gammie ag&Silk & Wang ( 2011 ( eateto hsc,Fros nvriyo aha,P.O Mashhad, of University Ferdowsi Physics, of Department 1972 and ) ( 1964 lere 1987 Elmegreen nmgeie ein,temagnetic the regions, magnetized In . ) 1. , ) ( ( lh,bt n gamma and beta alpha, 1984 .Ti rtro a enmodified been has criterion This ). 1996 Jog INTRODUCTION Vandervoort A T etn&Romeo & Bertin la aai l aei amo Roshan Mahmood Kazemi, Ali Nazari, Elham E tl mltajv 12/16/11 v. emulateapj style X , ) ( effect the considered also has ) ( 2013 1994 Elmegreen . ) and ) ; i srkr2001 Ostriker & Kim ( 2012 ( 1970 OTNWOINJASANALYSIS JEANS POST-NEWTONIAN oe Wiegert & Romeo , ) .As Toomre Also ). ( ( 1988 1995 Rafikov , ) and ) Romeo ( 2001 ABSTRACT Jog ). ) odsrmnt ewe rvt hois o example for theories, gravity differences between see observable discriminate to lead to some may in system differences dif- astrophysical These are and gravity. perturbations rate Newtonian small from the growth ferent of the these relation dispersion consequently the In the from and different case. is case principle, standard Newtonian in the mass, gravita- Jeans from the the different theories which is in law theories tional gravitational some of text h yaiso h akrudsse.Freapein example For influence significantly system. background can the they of since dynamics interest the of are ties analysis. Jeans standard deviations significant the domi- some from Neutrino find and re- and disks stars the accretion neutron apply nated massive we hyper disks example For to accretion sults or (BHs). case, nucleus holes the black galactic is than gravi- around the This higher near the regime. example, are Newtonian and for the system in fluid the allowed the what and throughout where of mass forces systems velocity tational Jeans astrophysical characteristic the for post- the criterion find the stability we using theory, the specifically (PN) More Newtonian instability. Jeans the oe,hsbe netgtdin black investigated Kerr been of orbiting has stability tori holes, local relativistic The magnetized evolution stars. strongly the neutron on hypermassive impact the profound of a has instability tional hysoe htsc oiaesbett nunstable an topology. to field subject magnetic are mode. tori toroidal magnetorotational such non-axisymmetric purely that a showed They of case the al. et Siegel ( to al. config- reader et symmetric the Hosseinirad cylindrically refer works of for we recent stability, analysis urations and sys- Jeans global literature relevant these the and the on of local of review analysis their brief Jeans study a to years. helps recent tems in raised systems filamentary has within perturbations small of growth 2012 h en ntblt a enas tde ntecon- the in studied also been has instability Jeans The eaiitceet ntedffrn id finstabili- of kinds different the on effects Relativistic on effects relativistic the study we paper, this In o oewrso h eaiitcKelvin-Helmholtz relativistic the on works some For ohn&Abbassi & Roshan o 46 aha,Iran; Mashhad, 1436, Box . . ) s w assicesswe temper- when increases masses two ese * n hha Abbassi Shahram and , tpesr a eptegravitational the help can pressure it ( emte nNwoinlmt The limit. Newtonian in permitted 2013 nteps-etna approxima- post-Newtonian the in r hrceitcvlct fthe of velocity characteristic ere cssesi investigated. is systems ic mae otesadr Jeans standard the to ompared esalrta h Newtonian the than smaller be thsbe hw htmagnetorota- that shown been has it ) od,terltvsi effects relativistic the words, r ( 2017 eprtr astrophysical temperature h ( 2014 ). , [email protected] 2015a ilu tal. et Wielgus rudihe al. et Freundlich , b ), aozel tal. et Capozziello ( 2015 ( 2014 for ) ), 2 instability we refer the reader to Blandford & Pringle of small perturbations in PN approximation using two (1976); Ferrari et al. (1978); Hardee & Norman (1988); different approaches. In Sec. 3 we find a new Jeans Zhang et al. (2009) . Furthermore the dynamical in- mass, i.e., the PN Jeans mass and in Sec. 4 we apply it stability of inspiraling neutron-star binaries near coales- to some high temperature systems. We also compare it cence in the PN limit has been studied in Lai & Wiseman with the standard Jeans mass. Finally in Sec. 5, results (1996); Faber & Rasio (2000). are discussed. It is important mentioning that in general rela- tivity field equations can be solved analytically only 2. EQUATIONS OF POST-NEWTONIAN for some restricted cases where the matter distri- HYDRODYNAMICS bution possesses special symmetries. On the other Before moving on to introduce the PN hydrodynamics, hand self-gravitating systems in reality do not pos- it is important to mention that we assume a perfect fluid sess such perfect symmetries. As a result, beside energy-momentum tensor for our fluid systems through- numerical simulations, approximative methods are out this paper. In other words dissipative effects have used to solve Einstein’s equations. One of the pow- not been considered in our analysis. erful approaches is the PN theory (Chandrasekhar In this section we briefly review the PN hydrodynam- 1965, 1967, 1969; Chandrasekhar & Nutku 1969; ics. As mentioned in the introduction section, the realm Chandrasekhar & Esposito 1970). This theory works in of validity of the PN theory is restricted to systems where systems where the gravitational field is suitably weak and the gravitational fields are weak and the motions (all the motions inside the matter source also are appropriately motions within the matter distribution) are slow com- slow compared with the speed of light. But both velocity pared with the speed of light. Keeping in mind these and gravitational field are high enough to lie outside the assumptions, the space-time metric of a PN system can realm of validity of Newtonian description. For some be obtained. For a comprehensive review of the subject applications of this theory we refer to: the equation of we refer the reader to Poisson & Will (2014). The result- motion of binary pulsars (Blandford & Teukolsky 1976; ing components of the PN metric are given by Epstein 1977; Hulse & Taylor 1975; Damour & Taylor 1991), tests of general relativity in solar system (Will 2 2 g = 1+ U + (Ψ U 2)+ O(c−6) 1987, 1994), and gravitational radiation reaction 00 − c2 c4 − (Chandrasekhar & Esposito 1970; Burke 1971; Blanchet 3 −5 g0j = 4/c Uj + O(c ) (1) 2006). Also it has been used to study the relativistic ef- − fects in accretion discs around BHs (Demianski & Ivanov 2 g = δ 1+ U + O(c−4) 1997). jk jk c2 There are two different approaches to PN theory. One of these derivations, which is referred as the classic ap- where Ψ is defined as proach to PN theory, is based on the standard formu- 1 lation of the Einstein field equations. An alternative Ψ= ψ + ∂ttX (2) derivation, the modern approach, of PN theory is based 2 on the Landau-Lifshitz formulation of Einstein field The potentials that appear in the metric are defined by equations. In this method, the PN metric is restricted the following differential equations to a near zone1 within one characteristic wavelength of 2U = 4πGρ∗ (3) the radiation, while there is not any clear restriction in ∇ − the classic approach. For an excellent introduction to 2X =2U (4) PN theory we refer the reader to Poisson & Will (2014) ∇ where modern and classical formulations of the theory 3 3p are introduced; and to Will (2014) for a comprehensive 2ψ = 4πGρ∗ v2 U +Π+ (5) ∇ − 2 − ρ∗ review of the applications of this approach. It is impor- tant to mention that modern approach possesses some 2 ∗ important advantages and removes several ambiguities U = 4πGρ v (6) ∇ − in the classic approach. Here, we use the modern ap- Where U is the Newtonian potential, U is the vector proach as well as the classic approach to investigate the potential, and X is the superpotential which its source local gravitational stability. term 2U extends over all space. In fact these equations Naturally approximated methods for solving Einstein’s are replaced with the Poisson equation in the Newto- equations have some restrictions . More specifically, as nian gravity. These differential equations can be simply we mentioned before, the PN approximation is limited solved, namely to weak fields as well as slow motions. However re- ′ cent words claim that PN approximation is unreasonably ∗ ρ 3 ′ successful to describe systems which are not in the PN U(t, x)= G − ′ d x approximation(Will 2011). x x Z | ′ ′ | ′ The layout of the paper is the following. In Sec.2, we ρ∗ ( 3 v 2 U ′ + Π′ +3p′/ρ∗ ) ψ(t, x)= G 2 − briefly review the equations of the PN hydrodynamics. x − x′ Also we derive the dispersion relation for propagation Z | | (7) ′ ′ X(t, x)= G ρ∗ x − x d3x′ 1 Near zone is a three dimensional region which the character- | | istic length scale is smaller than λc, where λc is the characteristic Z ′ ′ wavelength of the gravitational radiation. ρ∗ v j U j (t, x)= G d3x′ x − x′ Z | | Post-Newtonian Jeans analysis 3 in which the primed fluid variables are evaluated at time the following form t and position x′. It should be noted that the modern ∗ 1 2 ∗ 1 2 ∗ approach clarifies these potentials in terms of near-zone ρ ∂t( v +Π)+ ρ v. ( v +Π)+ .(pv) ρ v. U =0 integrals. It means that the field point x is within the 2 ∇ 2 ∇ − ∇ near zone, but in the classic approach there is no such (13) restriction on domain of the field point. In addition, After some manipulations one can verify that, in the case in modern approach the domain of these integrals, i.e., of perfect fluids, the above equation is the same as the , is a surface of constant time which is bounded by a first law of thermodynamics. So one can recognize it as a M statement of conservation of energy for isentropic flows. sphere of radius λc, while in the context of classic approach, it dependsR ∼ on the boundary of the potential Therefore, equations (3)-(6), (9)-(12) and an equation of sources. Therefore, some potentials with non-compact state make a complete set of differential equations for sources (e.g. the superpotential X) are ill-defined and describing a PN perfect fluid. the integral representations of them are divergent. Con- sequently, the field equations (3)-(6) may lead to am- 2.1. Gravitational instability criterion in the biguous solutions in this approach, for more details see post-Newtonian limit Poisson & Will (2014). A common way to derive the Jeans mass in Newto- The matter variables in the PN approach are nian gravity is to use the virial theorem. In the PN ρ∗, p, Π, v where ρ∗ is the conserved mass density de- limit one may use the PN virial theorem, see (Roshan {fined as } 2012; Chandrasekhar 1965) for virial theorem in the PN ρ∗ = √ gγρ = √ gρu0/c (8) approximation, in order to find the PN version of the − − Jeans mass. However, here we follow another approach in which ρ is the proper density and g is the determinant by studying the linear perturbation analysis of the self- of the metric tensor. Also u0 is the zeroth component of gravitating system. the velocity four-vector. It is clear that in the Newtonian In order to have a crude estimate for PN overall ef- limit there is no distinction between ρ∗ and the proper fects on the Jeans stability, let us study the gravitational density ρ. furthermore p is the pressure, Π = ǫ/ρ∗ is the binding energy of a spherically symmetric object in this internal energy per unit mass, ǫ is the proper internal approximation. The general form of the binding energy energy density and v is the fluid’s velocity field, and is of a spherically symmetric distribution of matter in gen- defined with respect to the time coordinate t. eral relativity is given by Hobson et al. (2006) Now let us introduce the hydrodynamics equations in 2 the first PN (1PN ) approximation. In fact, the Newto- EB = (M˜ M)c (14) nian equations of hydrodynamics change because of both − ˜ special relativistic and general relativistic effects. In this where M and M are different masses defined as case the continuity equation can be written as R M˜ =4π ρ∗(r) g (r)r2dr ∗ ∗ j rr ∂tρ + ∂j (ρ v ) = 0 (9) Z0 (15) R p ∂ ∗ 2 where ∂α = ∂xα . It is obvious that PN continuity equa- M =4π ρ (r)r dr tion is similar to the standard case and the only difference 0 ∗ Z is the appearance of ρ instead of ρ. Where the energy where R is the radius of the system and g is the rr ∗ rr density ρ and ρ are related as component of the metric. In particular, one can easily show that the gravitational binding energy of a spherical v2 3U ρ = ρ∗ 1 + O(c−4) (10) symmetric and static fluid, which has constant density − 2c2 − c2 and extends to coordinate radius r = R, can be written in the form Also we assume that the fluid is in local thermodynamic 2 16 π ∗2 5 equilibrium. In this case the exact statement of the first EB = Gρ R (16) law of thermodynamics for the perfect fluids is dΠ = 15 (p/ρ2)dρ. Consequently we can write Now to find the PN corrections to the binding energy we substitute equation (10) into above equation. Finally we dΠ p dρ∗ = + O(c−2) (11) find the PN gravitational binding energy EBp dt ρ∗2 dt 2 32 π 2 5 −4 k E E + Gρ U R + O(c ) (17) where d/dt = ∂t + v ∂k is the Lagrangian time deriva- Bp ≃ BN 5 c2 tive. On the other hand, the PN Euler’s equation can be written as where EBN is the binding energy of a uniform sphere in 16 π2 2 5 j Newtonian gravity and is equal to 15 Gρ R . It is clear ∗ dv ∗ 1 1 2 p ρ = ∂jp + ρ ∂j U + v + U +Π+ ∂j p that EBp > EBN. This means that PN effects increase dt − c2 2 ρ∗ the gravitational binding energy of this system. There- ∗ h fore for a system deep in the PN limit, keeping in mind j ρ 2 j k (12) v ∂tp + (v 4U)∂jU v (3∂tU +4v ∂kU) the virial theorem, one may conclude that more random − c2 − − i h motion is required to hold the system in an equilibrium +4∂ U +4vk(∂ U ∂ U )+ ∂ Ψ + O(c−4) state. In other words it seems that PN corrections would t j k j − j k j i support the gravitational instability. We will show that Furthermore the energy equation in the PN limit takes this simple conclusion is true. 4
Now, to study the linear perturbation growth in PN tion is limit, we need to find the linearized PN hydrodynamics. dv 1 p Using these equations one can, in principle, find a disper- ρ∗ 1 = p + ρ∗ U + Π + 0 p 0 dt −∇ 1 0∇ 1 c2 0 ρ∗ ∇ 1 sion relation governing the propagating of the first order 0 perturbations. On the other hand, this dispersion rela- ρ∗ + 0 [4∂ U + Ψ ] (22) tion helps to find a criterion for stability of the system c2 t 1 ∇ 1 against small overdensities. This criterion in the Newto- nian case is the so-called Jeans stability criterion. Our By differentiating equation (19) with respect to time and purpose here is to find the relativistic corrections to it. combining the result with the divergence of equation As we will see, analysis of the stability is more compli- (22), we find cated than the Newtonian case. In order to check the ∂2ρ∗ 1 p reliability of the results, we find the dispersion relation 1 2p + ρ∗ 2U + Π + 0 2p (23) ∂t2 − ∇ 1 0∇ 1 c2 0 ρ∗ ∇ 1 using two slightly different procedures and find a same 0 dispersion relation. ρ∗ + 0 4 .∂ U + 2Ψ =0 In the first method, one can solve the linearized equa- c2 ∇ t 1 ∇ 1 tions by a perturbative procedure. To do so, we perturb all the quantities up to the first order of perturbations. Conveniently we use the Fourier expansion as Q1 = i(k·x−ωt) Then using the Fourier expansion one can solve the lin- Qae for all perturbed quantities. Finally we find earized equations and find the Fourier coefficients. Fi- 1 p nally the dispersion relation is derived. Let us first as- ω2ρ k2p + k2ρ∗U + Π + 0 k2p (24) a − a 0 a c2 0 ρ∗ a sume that the fluid is a barotropic fluid and obeys the 0 following equation of state ρ∗ + 0 4ωk.U + k2ψ + ω2U =0 c2 − a a a p = p(ρ∗) (18) note that
Furthermore we assume that every quantity Q0(x) is 1 2Ψ = 2ψ + ∂ 2X = 2ψ + ∂ U . (25) perturbed as Q(x,t) = Q0(x)+ Q1(x,t) where zero in- ∇ 1 ∇ 1 2 tt∇ 1 ∇ 1 tt 1 dex refers to the non-perturbed quantities and “1” sub- script stands for the corresponding small perturbation, Now using the equation of state, Poisson equations and Q1 the first law of thermodynamics one can easily find all i.e., Q 1. Now we substitute these perturbed quanti- 0 ≪ the Fourier coefficients, i.e., Qa, in terms of ρa. From ties in the PN hydrodynamic equations (3)-(6), (9)-(12), equation (18) we have and (18) and keep the first order terms. For simplicity 2 it is assumed here that the background fluid is static, in- pa = csρa (26) finite and homogeneous. Therefore v=0 and as a result ∗ dp U0 vanishes. Also homogeneity implies that ρ , p0, Π0 2 0 where cs is the sound speed, i.e., cs = dρ . On the and U0 are constant. In this case the linearized form of ρ0 perturbed continuity equation is other hand equation (3) leads to
∗ ∗ ∗ 4πG ∂ ρ + ρ .v + v . ρ = 0 (19) Ua = ρa (27) t 1 0∇ 1 1 ∇ 0 k2 It is important mentioning that the above mentioned as- and equation (6) gives sumptions for the background matter distribution do not ∗ satisfy the equations. More specifically, the Euler’s equa- 4πGρ0 Ua = va (28) tion (12) implies that Ψ = 0. On the other hand, k2 ∇ 0 Possion’s equations (3) and (5) require that By substituting the perturbed quantities in the men- tioned form in to equation (22), one can easily show that 2 ∗ U0 = 4πGρ0 (20) v and k are parallel. As a result by considering equation ∇ − (19) we find and ω va = ∗ ρa (29) 3p kρ0 2ψ = 4πGρ∗ U + Π + 0 (21) ∇ 0 − 0 − 0 0 ρ∗ On the other hand using equation (11) we can simply 0 relate Πa to ρa as ∗ One can see only if ρ0 = 0, these results are consistent p0 with each other. This inconsistency can be removed by Πa = ρa (30) ρ∗2 an ad hoc assumption that unperturbed potentials are 0 zero and the Poisson’s equations are only able to de- To find ψa with respect to ρa, let us rewrite equation (5) scribe the perturbed fluid. This assumption can be con- as sidered as a generalized version of the so-called Jeans 3p 3p swindle in the Newtonian case, see Binney & Tremaine k2ψ =4πG ρ∗ U + Π + a 0 ρ (31) (2008). Therefore, following the Newtonian case, we use a 0 − a a ρ∗ − ρ∗2 a 0 0 the Jeans swindle, and complete the stability analysis. 3p It is easy to verify that the linearized PN Euler’s equa- +ρ Π + 0 a 0 ρ∗ 0 Post-Newtonian Jeans analysis 5
Now using equations (26), (27) and (30) we have In order to simplify this expression let us obtain U1, ∂ttU1 ∗ 2 and ∂t U1 in terms of ρ1. We begin with ∂ttU1. In 4πGρa 4πG p0 c ∇ · ∗ s this case the first time derivative of U1 is ψa = 2 ρ0 2 2 ∗2 +3 ∗ k − k − ρ0 ρ0 h (32) 3p ∂ U (x,t)= G∂ ρ∗(x′,t)fd3x′ (42) + Π + 0 t 1 t 1 0 ρ∗ Z 0 ∂f ∂ρ′∗ i = G ρ∗ + f 1 d3x′ Finally by substituting (26)-(29) and (32) into (24) and 1 ∂t ∂t keeping terms up to O(c−2), we find the following dis- Z persion relation 1 where f = |x−x′| . Using equation (19) and divergence theorem, and keeping in mind that the surface terms are 2 2 2 ∗ 1 p0 2 2 ∗ ω = csk 4πGρ0 Π0 + csk +4πGρ0 vanished, after some manipulations we obtain − − c2 ρ∗ 0 2 2 ∗2 ′ k 32π G ρ0 ′∗ ′k (x x ) 3 ′ (33) ∂tU1 = G ρ v − d x (43) − c2k2 0 1 x x′ 3 Z | − | By ignoring the PN correction terms, i.e., terms includ- similarly the second time derivative can be written as ing c−2 coefficients, the Newtonian dispersion relation is recovered. Equation (33) is the main result of this sec- ′ ′ k ′∗ dv1k (x x ) tion. In fact, as we mentioned before, we will find the PN ∂ttU1 = G ρ − (44) 0 dt x x′ 3 Jeans stability criterion using this dispersion relation. It Z | − | is important here to note that dispersion relation (33) is (x x′)k +ρ′∗v′kv′ ∂′ − d3x′ not valid for long wave lengths. One may simply find and 0 1 j j x x′ 3 constraint on the wave number of the perturbation mode | − | by taking into account that the PN corrections in (33) It is important mentioning that only first order terms, must be smaller than the Newtonian terms. Therefore i.e O(1), can appear in terms including U1. Therefore in we expect that the first term of the above equation we use the Euler’s ∗ linearized equation and keep the first order terms. Also 2 csρ0 2 it is clear that the second term is negligible because gives k > √2 kJ (34) cρ0 raise to O(4) terms. Finally we can write where the standard Jeans wavenumber kJ is defined as 2 ′∗ cs ′ ′∗ ∂ttU1 = G ρ ρ + 4πGρ0 0 ′∗ 1 k2 = (35) − ρ0 ∇ J 2 Z h (45) cs ρ∗(y,t) (x x′) ′ G 1 d3y − d3x′ Before moving on to find the stability criterion, let us ∇ x′ y · x x′ 3 derive the dispersion relation once again using an an- Z | − | i | − | other procedure based on the modern approach of the Now let us calculate ∂t U1. Using the fact that U1 = ∗ ′ ∇ · PN theory. In this case, it is straightforward to derive ρ0 v1 3 ′ G x x′ d x one may write the following relations by using the equation of state (18) | − | and Poisson’s equations (3)-(6) R ′∗ ′k ρ0 v1 3 ′ ∂tU1k = G∂t d x 2p = c2 2ρ∗ (36) x x′ 1 s 1 Z ∇ ∇ |′k − ′∗| 2 ∗ dv1 ρ0 3 ′ U1 = 4πGρ1 (37) = G d x (46) ∇ − x x′ 2 dt X1 =2U1 (38) Z | − | ′∗ ∇ ′k ′ ρ0 3 ′ + G v v ∂ ′ d x 1 1j k x x′ Z | − | 3p0 2ψ = 4πG ρ∗ Π + Because perturbation is linear, again the second term ∇ 1 − 1 0 ρ∗ 0 vanishes. In this case, by taking a space derivative we ∗ 3p1 ∗ p0 find +ρ0 U1 + Π1 + ∗ 3ρ1 ∗2 (39) − ρ0 − ρ0 2 ′ ′∗ ′∗ cs∂kρ1 ∂k∂tU1k = G ρ (47) 2U = 4πGρ∗v (40) 0 − ρ′∗ ∇ 1 − 0 1 Z 0 ρ∗(y,t) 1 using these differential equations together with equation +G ∂′ 1 d3y ∂ d3x′ (23) one can write k x′ y k x x′ Z | − | | − | 2 ∗ ∗ 2 ∂ ρ1 2 2 ∗ ∗ ∗ ρ0 p0 cs 2 ∗ or equivalently we can write 2 cs ρ1 4πGρ0ρ1 + 2 Π0 + ∗ ∗ ρ1 ∂t − ∇ − c ρ0 ρ0 ∇ 2 ′ ′∗ h cs ρ1 p ∂ U = G ρ′∗ ∇ (48) 4πG 0 + Π +3c2 ρ∗ (41) ∇ · t 1 − 0 − ρ′∗ ∗ 0 s 1 0 − ρ0 Z ∗ ′ ′ Gρ1(y,t) 3 (x x ) 3 ′ ∗ + d y − d x +4∂t U1 +4πGρ0U1 + ∂ttU1 =0 ∇ x′ y · x x′ 3 ∇ · Z | − | | − | i 6
Therefore it is clear form equations (45) and (48) that 3. POST-NEWTONIAN JEANS MASS ∂ U = ∂ U (49) So far we have derived the dispersion relation for prop- tt 1 − t∇ · 1 agation of the local overdensities in an infinite homoge- ∗ PN Now using Fourier’s expansion for ρ1 in the form neous fluid system. If we disregard the 1 terms, then ∗ i(k·x−ωt) the equation (33) coincides with the corresponding dis- ρ1(x,t)= ρae and using equation (49), equation (41) takes the following form persion relation in the Newtonian theory. Let us define the following parameters in order to simplify the disper- ∗ 2 2 2 ∗ ρ0 p0 2 sion relation ω c k +4πGρ + 4πG + Π0 +3c − s 0 c2 ρ∗ s 1 p 0 0 2 (50) α = 2 (Π0 + ∗ ) < 1 p0 cs 2 2 2 ∗ 2 ∗ c ρ0 + Π0 + k 3iGcsI1 +3G ρ0I2 4πG ρ0I3 =0 (54) ρ∗ ρ∗ − − c2 = (1 α)c2 0 0 sp − s where integrals Ii are defined as Gp = (1+ α)G
′ e−ik·(x−x ) where α> 0, csp is an effective sound speed and Gp is an I = k (x x′)d3x′ effective gravitational constant. As expected in the New- 1 x x′ 3 · − ZM | − | tonian regime α 0 and these parameters coincide with −ik·(x−y) ′ → e (x x ) cs and G respectively. The p index represents the con- ′ 3 ′ 3 ′ (51) I2 = ′ d y . − ′ 3 d x stants in PN theory and we use this notation in the rest M ∇ N x y x x Z Z ′| − | | − | of this paper. Generally speaking, the PN effects, effec- −ik·(x−x ) e 3 ′ tively, decrease the sound velocity and increase the grav- I3 = ′ d x M x x itational constant. Therefore, noting that sound speed Z | − | and gravitational constant are the representatives of the As previously mentioned, in this second derivation we pressure support and gravitation strength respectively, use the modern approach of PN theory. Thus we deal one may naturally expect that PN effects make the sys- with the near zone and integrals should be taken in this tem more unstable against the local perturbations. How- region. The index in this relation means that our ever, let us find the exact form of the stability criterion in calculations are restrictedM to inside the near zone. Here, this limit. Using the new variables we rewrite equation we can also see one of the ambiguities that appear in the (33) in the following simple form classic approach. In the classic approach the domain of the integrals over non-compact sources extends to infin- 32π2G2ρ2 ω2 = c2 k2 4πG ρ 0 (55) ity and consequently one may easily show that integrals sp − p 0 − k2c2 I2 and I3 in (51) will diverge. On the other hand in the It should be mentioned that since v = 0 and U = 0 (by modern approach, after integrating on a finite region of ∗ radius , we remove the terms that contain . In fact if Jeans swindle) one may infer ρ0 = ρ0. Since csp < cs, one considerR the contributions of wave zoneR to the inte- Gp > G and the third term in (55) appear with nega- grals, then there will be terms containing that cancel tive sign, the instability width will extend and include the terms arising from the near zone contributionR and we a larger interval of wavenumbers. On the other hand will not confront infinities. This cancellation in modern a perturbation mode k that is stable in both Newtonian approach provides a type of regularization of ill-defined and PN regimes, will have a smaller oscillation frequency integrals that improves the problem of divergent and am- in the PN limit. Equivalently for an unstable mode, as biguous solutions of Poisson equations. For more details expected, the growth rate is larger in the PN case. on this point we refer the reader to Pati & Will (2000). By setting ω = 0 in equation (55), we can obtain the To solve the integrals, in the spherical coordinate sys- border of stability. In this case the Jeans wavenumber in tem, without loss of generality, we impose that k is in the the PN limit can be obtained as z direction, i.e k = kzˆ wherez ˆ is the unit vector in the 2 2 2 2 p0 direction of z axis. Then integrating over all directions kJp kJ 1+ 2 cs + + Π0 (56) ≃ c ρ0 x, we obtain the following result for I2 4π 4π cos(k ) sin(k ) Where we have expanded the result and kept only the I = R 4π + i R (52) 1PN corrections. System is stable against perturbation 2 k2 − k2 k R with wavenumber k if k > kJp, or equivalently if the corresponding wavelength λ satisfies λ < λJp, where and similarly for I3 we obtain λJp = 2π/kJp. It is obvious that kJp > kJ. This ex- 4π 4π cos(k ) plicitly means that perturbations with wavelength in the I3 = R (53) interval λ <λ<λ , which are stable in the Newto- k2 k2 Jp J − nian dynamics, are unstable in the PN regime. In the It should be noted that I1 does not diverge and can be following we will show that these modes can be excited easily integrated, the result is 4πi. Based on our discus- at mediums with high temperature and low mean molec- sions on the modern approach− to PN theory, we remove ular weight. all terms that depend on . Therefore one may conclude On the other hand, we have another constraint on k in 2 2 R 2 that I2 = 16π /k and I3 = 4π/k . Substituting these equation (34) to ensure the validity of the PN approxi- results into equation (50) we find the same dispersion mation. Therefore combining these constraints, one may relation obtained in (33). conclude that the perturbation with wavenumber k is Post-Newtonian Jeans analysis 7 unstable provided that 0.030
1 J 2 0.025 √2cs kJ
The dimensionless standard Jeans mass m is then 4 θ-value J 0.05 3/2 3 mJ = (βθ) (67) 2 0.04 And the PN dimensionless Jeans mass mJp is
2 1 γ W m = m 1 3 β + θ (68) 0.03 Jp J − γ 1 − 0 Furthermore the PN dimensionless Jeans mass-energy 0.02 -1 MJp takes the following form
-2 0.01 3γ 1 -10 -5 0 5 10 MJp = mJ 1 3β + − θ (69) − γ 1 q − It is clear that the fractional difference between Newto- Figure 2. Dashed curves show the dispersion relation in New- tonian theory. From up to down, dashed curves correspond to nian and PN Jeans masses grows with dimensionless tem- θ = 0.05, 0.04, 0.03, 0.02 and 0.01. The solid curves are the cor- peratures as θ. We have plotted these masses in Fig. 1. responding dispersion relation in the PN approximation. Further- For simplicity, and without loss of generality we assume more, the vertical dashed lines show the corresponding qmin for the above mentioned temperatures. Also, we have assumed that the that the background fluid is monoatomic and evolves adi- 5 5 adiabatic index is γ = 3 . abatically (i.e., β = γ and γ = 3 ). Dashed curve is mJ, the red curve is MJp, and the blue curve corresponds to 2.0 θ-value mJp. As expected at low temperatures θ or equivalently at low thermal velocities, i.e., v c, there is no differ- 0.05 th ≪ ence between Jeans masses in PN limit and Newtonian 1.5 regime. However after θ 0.01 departure from standard 0.04 case appears. The Newtonian∼ Jeans mass increases with θ forever. However, although PN masses also increase 1.0 with θ, they start to decrease after a specific θ. In fact 0.03 mJp has a maximum at θ = 0.048 and MJp has a max- imum at θ = 0.055. At these θ the fractional difference 0.5 0.02 between PN and Newtonian Jeans masses is more than 60%. Squared dimensionless growth rate 0.0 0.01 However one should note that the decreasing behav- 0 5 10 15 20 25 30 ior of the PN masses is not reliable. In other words at Squared dimensionless wavenumber θ =0.048 we have vth 0.3c and so the thermal velocity ∼ Figure 3. Squared growth rate S2 with respect to the squared of the particles (or the sound speed cs), at this temper- wavenumber q2 for adiabatic fluid with γ = 5 . From up to ature is of the same order of the speed of light. So the 3 down, dashed curves belong to the Newtonian case with θ = slow-motion condition is not established. One may add 0.01, 0.02, 0.03, 0.04 and 0.05. Furthermore the solid curves are higher PN corrections to the hydrodynamic equations in the corresponding curves in the PN approximation. The vertical 2 order to study such a high temperature. Therefore let us dashed lines show qmin for the above mentioned temperatures. define a critical dimensionless temperature as θc =0.048 beyond which our analysis here may not be reliable. This It is clear that, as expected, by increasing the temper- parameter for isothermal fluid is about 0.057. So our re- ature θ the difference between dispersion relations gets sults are limited to θ<θc. We have shown this boundary larger. Also by increasing θ the instability interval, i.e., ∗ ∗ with vertical line in Fig. 1. qmin
βπc2 γ 0.08 L > θ 1 β + θ (71) c 0.06 s Gρ0 − γ 1 − now since the matter density is related to the number 0.04 density n as ρ0 = µmH n, we can rewrite the above equa- Dimensionless Jeans mass 0.02 tion as 0.00 0.00 0.02 0.04 0.06 0.08 7 βθ γ Lc > 5.17 10 1 β + θ kpc (72) × sµn − γ 1 Dimensionless pressure − 1.5 On the other hand, more importantly, the typical time Γ=1.4 scale for contraction of the perturbation must be smaller Γ=2 than the dynamical time scale t of the host system. If d 1.0 Γ=3 without loss of generality we assume the typical free fall time is tff 1/√Gρ0 , then the following constraint should also∼ be satisfied for triggering the perturbation growth 0.5 µn −1/2 t > 3Myr (73) Fractional difference d 103 cm−3 It is necessary to mention that the free fall time is dif- 0.0 ferent from 1/√Gρ0 in the PN regime. Therefore one 0.00 0.02 0.04 0.06 0.08 0.10 may use teff 1/√Gǫ0 as a better estimation. However, Dimensionless Pressure since ǫ > ρ∼, if the constraint (73) is satisfied then the 0 0 Figure 4. Top panel shows the Jeans masses with respect to p corresponding constraint in the PN approximation will for a polytropic fluid for different Γ. In this panel the solid curves be satisfied. belong to the PN case and the dot dashed curves belong to the In the Sec. 4 we briefly review some high temperature standard case. The bottom panel shows the fractional difference systems and then check the required conditions for grav- between the Jeans masses. itational collapse in the PN limit and compare it with Interestingly for Γ = 2, which is a common approxima- the Newtonian case. tion for neutron stars, the Jeans wavelength does not Γ directly depend on the matter density of the system and 3.2. polytropic fluid: p = Kρ K plays a central role. As in the case of ideal fluid, one In this section we study the PN Jeans analysis for a can impose a constraint as λJp Table 1 Characteristics of the various astrophysical systems −3 System n (cm ) T (K) Size (pc) λJp (pc) 100 × ∆mJ/mJ H II regions 0.1 − 104 104 0.31 − 31 32.78 − 1.03 × 104 1.86 × 10−6 NGC 7027 6 × 104 3 × 106 ∼ 4 × 10−3 231.82 5.59 × 10−4 Cygnus loop > 105 3.5 × 106 ∼ 3 × 10−4 193.96 6.52 × 10−4 ICM 10−3 107 − 108 ∼ 106 (3.28 − 10.37) × 106 (1.86 − 18.63) × 10−3 Fermi Bubble 10−2 108 − 109 104 (3.28 − 10.36) × 106 (1.86 − 18.63) × 10−2 HMNS 1039 1010 − 1011 ∼ 10−13 (1.03 − 3.07) × 10−13 (1.86 − 18.63) NDAF 1037 1011 ∼ 10−11 30.75 × 10−11 18.63 Note.— n and T are the number density and temperature respectively. The characteristic size of the system is shown in the fourth column. For each system, we have calculated the PN Jeans wavelength in the fifth column. Furthermore the fractional difference ∆mJ/mJ has been shown for all systems in the last column. Note that for all cases we have set µ = 0.615. It is worthwhile to point out that we choose thickness of the shell as a characteristic size for NGC 7027 and Cygnus loop. dashed lines for different Γ. It is clear that beyond pmax than the typical size of these stars, i.e., R 15 km, and the PN Jeans mass starts to decrease while the Newto- consequently in the Newtonian gravity the≃ local gravita- nian Jeans mass increases forever. As in the ideal gas tional collapse is not allowed. However in the PN limit case, we can not rely on our results for p > pmax. It is the Jeans wavelength is 34% smaller. In other words we interesting that pmax increases by increasing the Γ pa- have λJp = 24.11 km. However, we emphasize again that rameter. More specifically, it is maximum at Γ = 2.36 in order to find a more precise result one should add sev- and then decreases with Γ. The fractional difference be- eral physical effects which are important in the dynamics tween Jeans masses with respect to p has been illustrated of NS. in the bottom panel of Fig. 4. As expected the fractional 4. APPLICATION TO HIGH TEMPERATURE difference ∆m /m is also sensitive to Γ. Interestingly, J J ENVIRONMENTS at a same p, it is minimum when Γ = 2. This polytopic index is widely used for neutron stars. To compare the new Jeans mass with the Newtonian As an example for this section let us calculate PN Jeans case for an ideal fluid, which have been calculated in Sec. mass for a neutron star (NS). The exact form of equa- 3.1, let us first introduce some relatively high temper- tion of state for NSs is still not known. However the poly- ature environments and finally calculate the fractional tropic equation of state is widely used to model/simulate difference ∆mJ/mJ, where ∆mJ = mJ mJp, between this type of stars. Therefore, in what follows we assume Jeans masses for each system. We have− summarized the a polytropic equation of state with Γ = 2 and ignore the results in the Tables 1 and 2. effects of strong magnetic fields and differential rotation i) HII regions are one of the most easily observed ob- of material inside a NS. Thus, we find a crud estimation jects in the galaxies. The sources of the ionization in for the relativistic Jeans mass inside a NS, and for more these regions are cosmic-ray protons, intense ultra vio- reliable results one has to take into account other effects late emission from O and early B-type stars which have like magnetic fields and may be higher PN corrections. formed inside the molecular cloud, and shock waves from It is easy to show that for a NS the PN Jeans mass supernovae regions. They are composed primarily of hy- start to decrease for dimensionless pressures larger than drogen. The temperatures of these regions are of the 4 pmax = 0.042. By substituting typical value of K order of 10 K and their size varies from a light-year to 0.014m5 kg−1 s−2 (Lattimer & Prakash 2001) into equa-≃ several hundred light years. For example some of them tion (78) we find the critical density ρcrit 2.7 are maybe only a light-year or less across (ultra-compact 1017kg m−3. Therefore it is easy to verify that the≃ New-× HII regions) and some bigger regions (giant HII regions) tonian gravitational potential divided by c2 for a NS with are several hundred light-years across (Anderson et al. M =1.4M , ρ (2.7 5.4) 1017 kg m−3, and the typ- 2009). Corresponding masses of HII regions are of the ⊙ 2 4 ≃ 2− × 2 order of 10 10 M⊙, and their typical number density ical size 15 km is U/c MG/Rc 0.14. Furthermore − 4 −3 for this system we have≃c2/c2 0.17≃ and Π /c2 0.08. in the bright parts is of the order of 0.1 10 cm . It s ≃ 0 ≃ is easy to show that Newtonian Jeans wavelength− and As we have already mentioned these dimensionless quan- 4 tities measure deviations from Newtonian gravity. It is Jeans mass for HII regions are 1.04 32.8 10 pc and 2.8 106 8.9 108M respectively.− × clear that the PN corrections in this systems are signifi- × − × ⊙ cant. ii) Planetary nebulae (PNe) can be considered as high Now let us apply our PN Jeans criterion to a NS. temperature environments. A white dwarf that is sur- In other words, we check the local stability of inside rounded by gas clouds can be assumed to be a plane- of a NS in the PN approximation. We use the follow- tary nebula. This kind of hot bubble can be a product ing typical values for physical properties of NS: ρ of evolution of the central star from post-asymptotic gi- ant branch star to a white dwarf. In this phase very 5.4 1017 kgm−3, p = Kρ2 where K 0.014m5 kg−1 s−≃2. × ≃ fast and energetic winds are made by star (with speed In this case it is easy to show that the standard Jeans −1 −7 −1 wavelength is λ = 36.32 km. This wavelength is larger 1000kms and mass loss rates & 10 M⊙yr ). J When∼ such a fast wind receives to the gas cloud, there Post-Newtonian Jeans analysis 11 will be a shock and the gas is superheated enough to emit oretical models have been presented to explain the for- in X-ray (for example see Zhekov & Perinotto 1996). mation of the Fermi Bubbles. For example in Mou et al. NGC 7027 is a young and dense planetary nebula. It (2014) it has been shown that winds from the “post” contains a distribution of hot molecular hydrogen and hot accretion flow of the Sgr A∗ have caused inflation of ionized gas that are investigated by results of Hubble the Fermi Bubble. The temperature of the gas within Space Telescope and Near-Infrared camera and Multi Ob- the Bubbles is in the range of 108 109K. Moreover − −2 −3 ject Spectrometer program. Density in the shell is nH+ = the number density of the Bubbles is about 10 cm 6 104cm−3 and its thickness is about 0′′.9 (1.2 1016cm) (Su et al. 2010). Therefore the Newtonian∼ Jeans wave- (Latter× et al. 2000). Plasma temperature of the× ionized length is equal to 3.3 106 1.0 107 pc and the Jeans 6 15 × −16 × shell is estimated to be about 3 10 K (Kastner et al. mass is 2.8 10 8.9 10 M⊙. 2001). In this case, one can∼ show× that the Newtonian vi) As another× high− temperature× system, let us men- Jeans wavelength is 0.23kpc and the Newtonian Jeans tion the hyper massive neutron star (HMNS) which is 9 mass is 5.94 10 M⊙. a result of merger of a neuron star binary. It under- iii) The Cygnus× loop is a large soft-shelled supernova goes collapse to a rotating BH and the final result is a remnant in the constellation Cygnus, an emission neb- BH embedded in a massive, hot accretion torus contain- ula at a distance of 440 pc, and the angular dimensions ing 1-10% of the total mass of the system. Such a of 2◦.8 3◦.5 which corresponds to linear dimensions system∼ is considered as a proposal for generating short- 21.5 27pc× (Blair et al. 1999). A thin plasma with tem- duration gamma-ray bursts. The initial temperature of perature× of T 3.5 106K is suggested as the best-fit the HMNS is 5 1010K, and can even rise to 5 1011K model in Kimura≈ et al.× (2013). Furthermore this system (Rezzolla & Zanotti× 2013). The maximum density× is −4 14 −3 has a thickness which is not greater than 3 10 pc ρmax = 9 10 g cm . Furthermore, the life time of (Boss & Keiser 2010) and its molecular hydrogen∼ × column the surrounding× torus is 10 ms and has a temperature 21 −2 of order & 1012K (Shibata∼ et al. 2006). Furthermore its density is about NH2 0.1 10 cm (Uyanıker et al. 2001). Considering these≃ values,× Newtonian Jeans wave- density lies between 1012 and 1013g cm−3. The vertical length and Jeans mass for this system will be 0.19kpc and horizontal sizes of the torus are about 20km and 9 ∼ and 5.8 10 M⊙ respectively. 60 km respectively (Rezzolla & Zanotti 2013). The iv) Rich× clusters of galaxies are luminous sources of X- ∼Newtonian Jeans wavelength and Jeans mass for the 43 45 −1 HMNS are 3.086 10.184km and 9 10−3 0.28 M rays, with X-ray luminosities of LX = 10 10 erg s . − × − ⊙ These regions may have a dimension of− several Mpc respectively. and are called as intracluster medium (ICM). ICM vii) Gamma-ray Bursts (GRBs) are the most energetic has an average particle density n 10−3cm−3(Sarazin phenomena in cosmological scales. In order to explain 1986). Moreover the spectroscopy∼ in X-ray indicates the energy source of GRBs several model have been pro- that this emission almost is coming from the thermal posed. The most popular model for generation of short bremsstrahlung of hot plasma with a temperature of GRBs is mergers of compact objects (Narayan et al. 107 108K (Mo et al. 2010). Using mentioned num- 1992). On the other hand it is commonly believed that ber density− and temperature, the Newtonian Jeans wave- the collapse of a single massive stars can be responsi- length and Jeans mass can be calculated as 3.3 106 ble for long GRBs (Woosley 1993). For both cases, a 7 14 15 × − dense and extremely high temperature disk is expected. 1.0 10 pc and 2.80 10 8.86 10 M⊙ respectively. It× is worth mentioning× that− since× the characteristic In these hyper-accreting systems where the accretion −3 −1 ˙ −1 length scale of galaxy clusters is large and their density rate is extremely high (10 M⊙s M 10M⊙s ), ≤ ≤ is low, one may expect that the cosmic cosmological con- the neutrino radiation is dominant cooling mechanism. stant Λ plays significant role in the binding energy and Therefore these hyper-accreting disks are usually named consequently on the global stability of the system. The neutrino-dominated accretion flows (NDAFs). (e.g., effect of Λ on the binding energy of polytropic configura- Popham et al. 1999; Kohri & Mineshige 2002; Yi et al. tions in general relativity has been recently discussed by 2017). For this system temperature and density profile Stuchl´ıket al. (2016). In fact, when the density of this can be found in (Kohri & Mineshige 2002). One can see system is low and its characteristic size is large enough, at r = 4 rs (where rs is the Schwarzschild radius) when 20 −2 the very small repulsive cosmological constant, i.e., dark the surface density is 10 g cm , and MBH = 3M⊙, energy or vacuum energy, can sufficiently increase the the temperature and density will be 1.1 1011K and Jeans mass, see also Kiessling (2003). On the other hand ρ =2.3 1013g cm−3 respectively. Moreover× this model × as we mentioned, the Jeans wavelength for this system is does not work at r & 40 rs. For simplicity we assume comparable to the size of the system. Therefore, one may Lc = 40rs as a characteristic size of this system. In this concern about the effects of the cosmological constant. case the Newtonian Jeans wavelength and Jeans mass However existence of Λ will lead to a same contribution to are 101.22 km and 2.8 M⊙ respectively. both Newtonian and PN Jeans masses and consequently Now let us calculate the PN Jeans mass for the above will decrease the fractional difference ∆mJ/mJ. How- mentioned systems. To do so, it is convenient to rewrite ever, this fraction is already small for ICM while we are the Jeans masses, equations (67) and (68), with respect looking for systems where ∆mJ/mJ is large. Therefore to temperature T and not the dimensionless temperature we do not take into account cosmological constant effects. θ. Also it can be straightforwardly scaled to cases where 4 4 −3 v) Using the Fermi Large Area Telescope two giant temperature is 10 K and mass is M⊙ and n 10 cm . gamma-ray bubbles that extend nearly 10 kpc in diame- Furthermore to have a clear form of the standard∼ Jeans ter have been discovered. Their positions are above and mass we use ρ0 = µmH n. In this case the Jeans mass below the galactic plane (Cheng et al. 2011). Several the- in the Newtonian theory, i.e., equation (67), takes the 12 Table 2 Properties of NDAFs at different radii 36 −3 10 MBH r(rs) n(10 cm ) T(10 K) mJ(M⊙) λJ(km) ∆mJ/mJ 4 24.44 12.73 1.99 64.11 0.21 5M⊙ 10 7.52 5.80 1.10 78.02 0.09 40 1.27 1.77 0.45 105 0.03 4 32.89 15.52 2.30 61.02 0.25 10M⊙ 10 10.13 7.07 1.28 74.25 0.11 40 1.70 2.15 0.52 99.94 0.04 Note. — The fractional difference between standard and the PN Jeans masses for NDAF calculated in different radii for different masses of the central BH. We have assumed that µ = 0.70. following form In other words in this case one may doubt the validity of 3 the PN analysis. 2 1 492 βT n − 2 Eventually, for the last system (i.e., NDAF) the dy- m 103M (81) J ∼ µ2 × ⊙ 104K 104cm−3 namical time scale is introduced as the accretion time (for more detail see Kohri & Mineshige (2002)). In this Furthermore, the new Jeans mass given by equation (68) model at r =4rs, for BH mass M =3M⊙, and viscosity takes the following form parameter α =0.1 the accretion time is estimated about −2 −3 3 1.3 10 s which is greater than t 10 s. Thus, 2 1 ff 492 βT n − 2 ∼ × ∼ m 103M (82) in principle, this location can be locally unstable. We Jp ∼ µ2 × ⊙ 104K 104cm−3 have also obtained the temperature and number density 275 γ T at some different radii for two different central BHs and 1 10−11 β + reported the results in the Table 2. It is clear from this × − µ × γ 1 104K − table that as expected the plasma temperature is higher In the Table 1, we have summarized some of the physical at smaller radii, and consequently the fractional differ- properties of the introduced high temperature astrophys- ence is larger. This fraction for M = 10 M⊙ BH is larger ical systems. Also, the Jeans mass mJ, the Jeans wave than a BH with mass M =5 M⊙ in a same radius. This PN length λJ, and the fractional difference ∆mJ/mJ are cal- means that at least in the 1 approximation, such a sys- culated using equations (81) and (82). One can see that tem at the mentioned radii is probably subjected to local the fractional difference for adiabatic fluid have the fol- fragmentation. This fact can significantly influence the lowing form dynamics of the accretion. In fact one of the main mech- anisms for transforming angular momentum throughout 2 mJ mJp 275 −11 γ T the accretion disk is viscosity. On the other hand lo- − = 10 4 (83) mJ µ × γ 1 10 K cal fragmentations can produce inhomogeneous pattern − which in principle can exert nonzero torques to the back- As we mentioned before, this fraction increases linearly ground disk and consequently trigger the accretion flow. with temperature T . It is clear from Table 1 that, as It is important to mention that some features of GRB expected, for systems at very high temperatures (i.e., flares are consistent with a viscous accretion disk model 11 T & 10 K) the difference between mJ and mJp is signif- which is locally fragmented via gravitational instability icant. More specifically this difference reaches to approx- (Perna et al. 2005). From this point of view it seems imately 19% for HMNS and NDAF. In theses cases the that high resolution hydrodynamic simulations may con- PN Jeans wavelength is also comparable to the charac- firm the existence of such local fragmentations in the teristic size of the system. For other cases the difference accretion disks around BHs. is not tangible. Therefore let us focus on these two cases. It is also necessary to mention that for rotating disks It is necessary to reiterate that to have a gravitational it is more appropriate to use the so-called Toomre crite- collapse, the dynamical time scale of the host system rion instead of the Jeans mass criterion. In this case it should be longer than the typical time scale for the con- is required to find the PN version of the Toomre crite- traction of the perturbation, see Eq. (73). We know that rion (Kazemi et al. 2017). For rotating systems the axial a HMNS collapses to a BH with a torus or a supermassive and radial directions are not equivalent from the point of neutron star (Rezzolla & Zanotti 2013) after 1 ms 1s. view of the local gravitational instability. More specifi- − According to equation (73) we see the dynamical time of cally the perturbation can propagate in both radial and this system is greater than its typical free fall time (where axial directions. This fact makes the stability analysis is about 10−4s). So, the PN analysis shows that this complicated for the non-axisymmetric perturbations. It system can∼ be threatened by gravitational instability. It is well-known that there is no analytic criterion for the is easy to verify that for the torus the fractional differ- stability of non-axisymmetric perturbations in a differ- ence between Jeans masses is even larger and reaches entially rotating disks. In other words Toomre criterion 186%. However this large deviation is not reliable in the deals with the radially propagating axisymmetric pertur- sense that the temperature of the torus, i.e., 1012 K, is bations. larger than the critical temperature defined in Eq. (63). Post-Newtonian Jeans analysis 13 5. CONCLUSION Boss, A. P., & Keiser, S. A. 2010, ApJ, 717, L1 Burke, W. L. 1971, JMP, 12, 401 In this paper, we have studied the Jeans instability Capozziello, S., De Laurentis, M., De Martino, I., Formisano, M., in 1PN limit. In other words we have studied the first & Odintsov, S. 2012, PhRvD, 85, 044022 Chandrasekhar, S. 1965, ApJ, 142, 1488 relativistic corrections to the Jeans instability. However Chandrasekhar, S. 1967, ApJ, 148, 621 we have idealized our analysis in the sense that we have Chandrasekhar, S. 1969, ApJ, 158, 45 ignored many physical effects, like magnetic fields, vis- Chandrasekhar, S., & Esposito, F. P. 1970, ApJ, 160, 153 Chandrasekhar, S., & Nutku, Y. 1969, ApJ, 158, 55 cosity, radiation pressure and so on, which in principle Cheng, K.-S., Chernyshov, D., Dogiel, V., Ko, C.-M., & Ip, W.-H. influence the instabilities. Taking into account these ef- 2011, ApJ, 731, L17 Damour, T., & Taylor, J. H. 1991, ApJ, 366, 501 fects would lead to more complete and of course to much Demianski, M., & Ivanov, P. 1997, A&A, 324, 829 complicated analysis. Elmegreen, B. 1995, MNRAS, 275, 944 For this purpose, we have introduced 1PN corrections Elmegreen, B. G. 1987, ApJ, 312, 626 Elmegreen, B. G. 2011, ApJ, 737, 10 to the fluid equations. Finally linearizing the governing Epstein, R. 1977, ApJ, 216, 92 equations of the fluid system, we derived the gravita- Faber, J. A., & Rasio, F. A. 2000, PhRvD, 62, 064012 tional instability criterion. In fact we found the new Ferrari, A., Trussoni, E., & Zaninetti, L. 1978, A&A, 64, 43 Freundlich, J., Jog, C. J., & Combes, F. 2014, A&A, 564, A7 Jeans mass limit mJp (58). The main theoretical result Gammie, C. F. 1996, ApJ, 462, 725 of this work is that the new Jeans mass is smaller than Hardee, P. E., & Norman, M. L. 1988, ApJ, 334, 70 Hobson, M. P., Efstathiou, G. P., & Lasenby, A. N. 2006, General the standard one. One may simply conclude that pres- relativity: an introduction for physicists (Cambridge University sure and internal energy in the relativistic situations can Press) in principle strengthen the gravitational force. Hosseinirad, M., Naficy, K., Abbassi, S., & Roshan, M. 2017, MNRAS, 465, 1645; arXiv:1611.00139 [astro-ph.GA] In the section 3.1 we applied the ideal fluid equation Hulse, R. A., & Taylor, J. H. 1975, ApJ, 195, L51 of state to derive a stability criterion as a function of Jog, C. J. 1996, MNRAS, 278, 209 temperature. In this system the PN Jeans mass is smaller Jog, C. J. 2013, MNRAS, 434, L56 Jog, C. J., & Solomon, P. 1984, ApJ, 276, 114 than the Newtonian mass. Kastner, J. H., Vrtilek, S. D., & Soker, N. 2001, ApJ, 550, L189 In the section 3.2 we studied the Jeans instability in Kato, S. 1972, PASJ, 24, 61 a polytropic fluid. Similar to the perfect fluid case the Kazemi, A., Nazari, E., & Roshan, M. 2017, Toomre criterion in post-newtonian theory, work in progress PN Jeans mass is smaller than the Newtonian mass. We Kiessling, M. K.-H. 2003, Adv. Appl. Math., 31, 132 have shown that there is a critical density, i.e., ρcrit, be- Kim, W.-T., & Ostriker, E. C. 2001, ApJ, 559, 70 Kimura, M., Tsunemi, H., Tomida, H., Sugizaki, M., Ueno, S., yond which our calculations will not be reliable. In other Hanayama, T., Yoshidome, K., & Sasaki, M. 2013, PASJ, 65, 14 words higher order PN corrections are required. As an Kohri, K., & Mineshige, S. 2002, ApJ, 577, 311 application, we found Jeans wavelength for a NS. In the Lai, D., & Wiseman, A. G. 1996, PhRvD, 54, 3958 Latter, W. B., Dayal, A., Bieging, J. H., Meakin, C., Hora, J. L., PN regime the Jeans wavelength is about 34% smaller Kelly, D. M., & Tielens, A. 2000, ApJ, 539, 783 than the Newtonian one. More specifically, in this case Lattimer, J., & Prakash, M. 2001, ApJ, 550, 426 although the standard Jeans wavelength is larger than Mo, H., Van den Bosch, F., & White, S. 2010, Galaxy formation and evolution (Cambridge University Press) the radius of the star, by adding the PN corrections in Mou, G., Yuan, F., Bu, D., Sun, M., & Su, M. 2014, ApJ, 790, 109 this systems, the PN Jeans wavelength might be compa- Narayan, R., Paczynski, B., & Piran, T. 1992, ApJ, 395, L83 Pati, M. E., & Will, C. M. 2000, PhRvD, 62, 124015 rable to the size of the system. Perna, R., Armitage, P. J., & Zhang, B. 2005, ApJ, 636, L29 In the section 4, we have calculated the new Jeans Poisson, E., & Will, C. M. 2014, Gravity: Newtonian, mass for some different astrophysical systems for which Post-Newtonian, Relativistic (Cambridge University Press) Popham, R., Woosley, S. E., & Fryer, C. 1999, ApJ, 518, 356 the ideal fluid approximation can be applied. The New- Rafikov, R. R. 2001, MNRAS, 323, 445 tonian and PN Jeans masses are completely similar at Rezzolla, L., & Zanotti, O. 2013, Relativistic hydrodynamics low temperature. However by increasing the temperature (Oxford University Press) Romeo, A. B. 1992, MNRAS, 256, 307 the fractional difference increases and reaches to about Romeo, A. B., & Wiegert, J. 2011, MNRAS, 416, 1191 19% in HMNS and to more than 100% for its torus. In Roshan, M. 2012, CQGra, 29, 215001; arXiv:1208.3184 [gr-qc] Roshan, M., & Abbassi, S. 2014, PhRvD, 90, 044010; the case of NDAFs, we investigated different masses for arXiv:1407.6431 [astro-ph.GA] the central BH and showed that local and small disk Roshan, M., & Abbassi, S. 2015a, Ap&SS, 358, 1; perturbation at small radii around BH can be seriously arXiv:1506.00942 [astro-ph.GA] Roshan, M., & Abbassi, S. 2015b, ApJ, 802, 9; arXiv:1501.04715 threatened by gravitational instabilities. [astro-ph.GA] Sarazin, C. L. 1986, RvMP, 58, 1 Shadmehri, M., & Khajenabi, F. 2012, MNRAS, 421, 841 It is a pleasure to thank Nicolas Chamel for helpful Shibata, M., Duez, M. D., Liu, Y. T., Shapiro, S. L., & Stephens, comments and discussions. M. Roshan thanks Aspen B. C. 2006, PhRvL, 96, 031102 Siegel, D. M., Ciolfi, R., Harte, A. I., & Rezzolla, L. 2013, center for physics where a part of this work was carried PhRvD, 87, 121302 out. Also we thank the anonymous referee for useful Stuchl´ık, Z., Hled´ık, S., & Novotn`y, J. 2016, PhRvD, 94, 103513 Su, M., Slatyer, T. R., & Finkbeiner, D. P. 2010, ApJ, 724, 1044 comments. Toomre, A. 1964, ApJ, 139, 1217 Uyanıker, B., F¨urst, E., Reich, W., Aschenbach, B., & REFERENCES Wielebinski, R. 2001, A&A, 371, 675 Vandervoort, P. O. 1970, ApJ, 161, 87 Anderson, L., Bania, T., Jackson, J., Clemens, D., Heyer, M., Wang, B., & Silk, J. 1994, ApJ, 427, 759 Simon, R., Shah, R., & Rathborne, J. 2009, ApJS, 181, 255 Wielgus, M., Fragile, P. C., Wang, Z., & Wilson, J. 2015, Bertin, G., & Romeo, A. B. 1988, A&A, 195, 105 MNRAS, 447, 3593 Binney, J., & Tremaine, S. 2008, Galactic dynamics (Princeton Will, C. 1994, in Proceedings of the Eighth Nishinomiya-Yukawa university press) Memorial Symposium on Relativistic Cosmology, ed. M. Sasaki Blair, W. P., Sankrit, R., Raymond, J. C., & Long, K. S. 1999, (Tokyo: Universal Academy Press) AJ, 118, 942 Will, C. M. 1987, in 300 years of gravitation, ed. K. Thorne, Blanchet, L. 2006, LRR, 9, 4 S. Hawking, & W. Israel (Cambridge, England: Cambridge Blandford, R., & Pringle, J. 1976, MNRAS, 176, 443 University Press), 80 Blandford, R., & Teukolsky, S. A. 1976, ApJ, 205, 580 Will, C. M. 2011, PNAS, 108, 5938 14 Will, C. M. 2014, LRR, 17, 4 Zhang, W., MacFadyen, A., & Wang, P. 2009, ApJ, 692, L40 Woosley, S. 1993, ApJ, 405, 273 Zhekov, S., & Perinotto, M. 1996, A&A, 309, 648 Yi, T., Gu, W.-M., Yuan, F., Liu, T., & Mu, H.-J. 2017, arXiv:1701.07573