Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Manifestations of Dynamo Driven Large-Scale Magnetic Field in Accretion Disks of Compact Objects G. D. CHAGELISHVILI, R.G . CHANISHVILT, J.G. LOMINADZE, AND Z.A. SOKHADZE Abastumani Astrophysical Observatory ABSTRACT Many observations, as well as the possible theoretical explanations of these observations, indicate that the existence of mean (large-scale) magnetic fields in the majority of astrophysical objects is determined by the turbulent dynamo action. The magnetic field generation sources in the turbulent dynamo mechanism are the differential rotation of the medium and the gyrotropic character of the turbulence, existing in this medium (Moflatt 1978; Vainstein et al. 198~, Krause and Radler 1980~. In the accretion disks of the compact objects the matter is moving due to the Kepler law, i.e. the rotation has a strong differential character. That is why one can suppose that in the convect~vely active regions of these objects the especially favorable conditions for the large-scale magnetic field generation are realized. In fact, the generation of the large-scale magnetic field in turbulent dynamo theory depends on the value of the so- called dynamo number (Moflatt 1978), determined by the parameters of the medium and turbulence. In the case of the accretion disk, these parameters have such values that when there is a thermal convection, the large-scale magnetic field is generated without any difficulties. The maximum value reached by the magnetic field is determined by nonlinear phenomena, i.e. by suppression of the sources of generation by the magnetic field itself (in the case of the accretion disk it is a suppression of the helical character of turbulence (Chagelishvili et al. 19863. The turbulent dynamo nonlinear theory developed by us (Chagelishvili et al. 1986, 1988) shows that in the compact objects of accretion disks, the generated large-scale magnetic field (when the generation takes place) has a practically toroidal configuration. 46
HIGH-ENERGY ASTROPHYSICS 47 Its energy density can be much higher than turbulent pulsations energy density, and it becomes comparable with the thermal energy density of the medium. On the basis of these constantations the manifestations to which the large-scale magnetic field can lead at the accretion onto black holes and gravimagnetic rotators respectively are presented. In particular, it is shown below that at the accretion onto the black holes, the dynamical activity of the strong toroidal, large-scale magnetic field as a result of the Parker instability development, in the disk, max- imum energy release region, can create such formations (hot, optically thin coronaes3 which can explain, for example, the Cyg X-1 spectrum and radiation specifications in the low state of this source. It is also shown below that at the disk accretion of the magnetized plasma onto an aligned grav~nagnet~c rotator, the existence of two magnetic fields of different origin in the system leads to the asymmetric accretion of the matter. The accretion mainly takes place selectively onto one of the magnetic poles depending on the co-rotation and Alfven radii ratio. BIMODAL ACCRETION ON CYGNUS X-1 Turbulence in accretion disks may be caused by the differential char- acter of matter rotation on Keplerian orbits and/or by the existence of a superadiabatic pressure gradient across the disk (when it does exist). The kinds of turbulence corresponding to these two factors are called shear and convective turbulence, respectively. Both kinds of turbulence can create an anomalous nscositr, transporting angular momentum outwards through the disk and thus causing an accretion. However, there is a basic difference between these kinds of turbulence, which account for the bimodal character of accretion in Cyg X-1. This difference is the following: shear turbulence is mainly two~imensional and has no gyrotropic character, while convec- tive turbulence is especially three-dimensional and of gyrotropic character. Therefore, convective turbulence in a differentially rotating accretion disk generates helicity and leads to amplification of large-scale magnetic fields (Chagelishvili e! al. 1986~. Thus, unlike shear turbulence, convective tur- bulence may make an accretion disk magnetically active, but only if there exists a large-scale magnetic seed field which is not too small. On this property of convective turbulence along the bimodal ac~edon, the model for Cyg X-1 (Chagelishvili et al. 1986, 1988) is founded. It predicts, for a "low" state, the formation of hot magnetic arcs: the very inhomogeneities which are able to create MV for this source. The model assumes (Chagelishvili et al. 1986, 1988) a variation of the . accretion rate M in a certain interval (M1, M2) and the existence of some critical rate MA in it (Here MCr has nothing in common with the critical Eddington accretion rate). Shear turbulence definitely exists in the whole
48 AMERICAN AND SOY7ET PERSPECTIVES accretion disk (Chagelishvili et al. 1988), but convective turbulence is only inherent to the inner t radiation-dominated region of the accretion disk and is always absent In the outer region. Thermal convection is also absent in the middle region if M > Mcr' and for M < MCr the parameters of the region are such that it becomes unstable against thermal convection; that is, the middle region becomes connectively active. This is the very circumstance that makes the accretion disk magnetically active and switches Cyg X-1 to the "low" state. But how does it actually happen? Originally a magnetic field is camed to the disk by the matter coming from the optical component of the binary system. A part of this field, being large-scale, may be considered as a seed field for the processes described by the turbulent dynamo equations. In the outer region where helical turbulence is absent (there is no thermal convection), the large-scale magnetic field decreases because of the turbulent diffusion arising from the shear turbulence. If thermal convection in the middle region is still absent (M > Mcr), the decrease of the large-scale field transported through by the matter is not able to supply the inner region of the disk with a sufflcientl`,r strong large-scale magnetic field. As a result, in spite of the fact that the large-scale magnetic field is generated in the inner, convectivetr active region of the disk, estimations show (Chagelishvili e! al. 1986) that it does not have time to increase up to perceptible values, and accretion goes on without a large-scale magnetic field, mainly in accordance with the standard model Thus, we can say that when M > M=, Cyg X-1 is in a "high" state. When M < Mar, thermal convection appears in the middle region of the disk and the generation of a large-scale magnetic field has already begun here long before the matter comes to the inner region. It should be emphasized that the generated large-scale magnetic field is virtually azimuthal (Chagelishvili e! al. 19863. Under such conditions, the magnetic forces become stronger in the region of the main energy release of the disk and have a real influence on the matter dynamics. Namely they give rise to a Parker instabilit~rather oblonged parts of some magnetic tubes emerge out of the main volume of the disk, forming arcs of decreased density above it. At the same time the greater part of the magnetic tube matter sinks toward the central plane, forming "clots" situated between the arcs (see Figure 1~. The study of Parker instability in the inner region of an accretion disk made by Chagelishvili ~ al. (1988) shows that in the process ~ It is well known that depending on the nature of opanDr (complon scauenng Ke ,, or free-free transitions Kff ~ and on the ratio of the gas pressure to the radiation pressure Pg/Pr), there may be three regions in an accretion disk (Shakura and Sunyaev 1973~: an outer region (ems << Kf I; P,/Pr ~ 1), a middle region (ReS ~ Kff; Pg/Pr ~ 1), and an inner region (Ken ~ Kff; Pg/Pr << 1).
HIGH-ENERGY ASTROPHYSICS A_ _: ;Z' :'\- 49 FIGURE 1 Schematic geometry of the disl: inner region in the cross section r = const at the Parker instability development. Shaded regions present clustem of concentrated gas. Solid Attires with arrows are magnetic lines folded from the initial azimuthal magnetic field. of instability development, mainly with maximum growth rate perturbations that have no Z ~ -Z reflection symmetry are amplified (see Figure 1). The action of Parker instability described above must be considered as a quasiperiodical process. We assume that magnetic flux tubes emerging in the region of the main energy release are heated, forming a hot, optically thin corona. The latter is made up of several magnetic arcs which are connected with the main disk and are sweeping inward in the process of matter accretion. Then the magnetic flux tubes emerge once again and the Cycle continues. At any subsequent time, the magnetic flux tube may emerge at a different distance from the black hole. Thus we propose that a number of these magnetic arcs of decreased density form, due to heating, the hot, optically thin corona so necessary for explaining the power-law spectrum of Cyg X-1 in the "low" state. Such a spectrum is generated in the arcs by comptoni~tion of soft X-ray photons emitted by dense clusters of relatives cold plasma between and under the arcs in the vicinity of the equatorial plane of the disk (see Figure 1~. The corona made up of these magnetic arcs will cover only a part of the colder "core" which, in fact, is indirectly confirmed by some observations of C;yg X-1 (Sunyaev and lluemper 1979~. We hold that macroscopic magnetic arcs rapidly rotating in the region of the main energy release are the very "hot spots" necessary to explain millisecond variation phenomena in a "low" state.
50 AMERICAN AND SOVIET PERSPECTIVES ASYMMETRIC DISK ACCRETION ONTO MAGNETIZED ROTATING COMPACT STARS The phenomena, occurnog at disk accretion onto magnetized rotating compact stars in all their diversity, have been studied since the early 1970s. They have been collected and classified in the books of Shapiro and Teukosly (1983) and Lipunov (1987~. The completeness of the research is limited as they considered only the non-magnetized disks, i.e., the cases of the accretion disks without Me magnetic field, which can really influence the interaction processes of the accredng gas and compact star's magnetic field. First of all it is relevant to the processes taldug place on the Even surface. In fact, as the researchers say, under certain conditions (the existence of the convective turbulence of hyrotropic nature in the larger part of the disk) the case where the disk has the large-scale magnetic field generated as a result of the turbulent dynamo action is realized. The field obtained thus is mainly of toroidal configuration (Chagelishvili e' al. 1986) and is capable of mounting to the meaning where the magnetic pressure of the disk becomes comparable with its thermal pressure (Chagelishvili e' al. 1986~. In other words, it becomes comparable with the compact star magnetic pressure on the Alden surface. It is easy to understand that this circumstance will radically change the Secreting gas penetration ways and means into the star magnetic pole. To be more specific, the strong toroidal magnetic field of a disk will prevent the interchange instability and thus will exclude the scenario of plasma penetration into the magnetosphere developed in papers by Ghosh and Lamb (1979a,b). The peculiarities of the accretion at a magnetized disk are revealed in the most simple and, consequently, the most easily obsene~d case which we are going to consider: a compact star, whose rotation and magnetic axes coincide (let us direct these axes along z-coordinate). We shall idealize a compact magnetized star as a gravimagnetic rotator with three character- istics: mass, magnetic dipole moment, and rotational moment Such an object can seine as the model for absolutely different astophysical objects: neutron stars, white dwarfs, magnetic stars and spinars-the supermassive stars. Let us consider that this magnetized thin accretion disk is in the equatorial plane of the star. Moreover, let us suppose the equilibrium rotation of the compact star to be almost achieved (Lipunov 1987~: on the Alden surface the Secreting plasma almost corotates with the field lines of the star's magnetic field. This occurs when the Alfven (RA) and corotation (Rc) radii are close to each other: RA ~ RC. AS iS known (Lipunov 1987) the similar corelation between these radii is quite widespread and that is why when considering it, one can hope that we encompass the most interesting cases Besides, it is especially easy to describe the field lines
HIGH-ENERGY ASTROPHYSICS 51 reconnection process of the rotator and the accretion disk magnetic fields while obse~v~g the above-mentioned ratio. Let us note that the change of the dipole magnetic field direction into the opposite one, in the case of non-magnetized plasma disk accretion in the align rotator, does not influence the accretion physics: the physical processes are symmetrical relatives to the equatorial plane. But in our case the existence of two magnetic fields of different origin, the accretion disk toroidal field (Be) and the rotator dipole field (BD), in the system leads to the occurrence of asymmetry (relative to z-axis) in the direction of the physical phenomena, and this asymmetry is determined by the B.<,, and BD mutual orientation. Namely, the manifestation of this asymmetry is such a circumstance when the plasma accretion will take place mainly to one of the rotator magnetic poles and not to both. The accretion will be asymmetrical in the case discussed by us and that, by all means, should adequately describe the observational manifestations of such systems. Concretely, the reason for this asymmetry is the following: the star and the accretion disk magnetic fields are crossed. This fact promotes their reconnection on the Eleven surface and the latter promotes the formation of the "channels" (see Figure 4~. The accreting plasma "slips" along the "channels" mainly to one of the two magnetic poles of the rotator. Let us discuss the reconnection process in detail. The reconnection of the magnetic field lines takes place as a result of the tearing-mode instability development (Furth et al. 1963; Lee and Fu 1985; Southwood et al. 1986) in solar situations. In our case the dissipative effects are insured by the disk turbulence (Chagelishvili et al. 1986) that leads to the Reynolds magnetic numbers Rm ~ 102 . 103. The mathematic methods of the tearing mode instability research are quite non-trivial. However, the information necessary for the qualitative description of the field lines behavior during the reconnections on the Alfven surface can be extracted from the papers (Furth et al. 1963; Lee and Fu 1985; Southwood et al. 19863. In accordance with them the field line dynamics can be schematically presented as follows. Figure 2 presents the align rotator ~D) and the accretion disk (B¢) field lines on the Alfven surface separately from each other. The bold lines present the regions, which are magnified in Figure 3 and where the reconnection physics is described. The reconnection of the crossed field lines occurs at the nodal points (Figure 3a) during the characteristic tome (r ~ Rmi/2/Q) where Q is the star rotation angular velocity (Furth et al. 1963), which is less than the accretion time. The further dynamics of the field lines, resulting in paths formation, is presented in Figures 3(b) and 3(c). Figure 4 shows one of the field tubes, formed as a result of the above processes. The configuration of the given field lines in the equatorial region is such that the matter, included and frozen into it, receives the impulse in the z-direction and "slips" to the upper magnetic pole if the Kepler
52 Q AMERICAN AND SOVIET PERSPECTIVES BD l ~ Bv FIGURE 2 The align rotator ~D) and the accretion disk (13~) field lines on the Albren surface separately from each other. The bold lines present the regions, which in Figure 3 are magnified and where the reconnection physics is dubbed. ~ -> ~ -) ~ a b c FIGURE 3 1-he c~ field lines reconnection dynamics rotation velocity of the matter in a disk is higher than BRA on the Alden radius. In other words, it occurs when RC > RA tie. when RC/RA-1 > O). NaturaLly, the "slip" of me matter to the lower magnetic pole will take place at RC < RA observance (see Figure 4~. We want to reiterate that we discuss the case when RA and RC are close to each other, as only in this case the Kepler rotation of the matter does not change the teanng-mode Instability development, described in papers by Furth e' al. (1963~; Lee and Fu (1985~; and Southwood e! al. (1986~. This paper has carried out the qualitative analysis of the accretion onto the align gravimagnetic rotator. But it is easy to understand that the main character of the accretion- the asymmetry- will be preserved at the disk accretion of the magnetized plasma (Bf ~ 0) onto the oblique gravimagnetic rotator as well. Finally we enumerate those observational manifestations to which the above-described accretion can lead due to the asymmetry: · if such a system is capable of generating jets, Men they will be observed only in one direction; · even the slightest change of the accretion rate can lead to the RC/RA-1 sign change. That will lead to the fact that the other pole will become accreting. Eking into account that the Alfven radius may
HIGH-ENERGY ASTROPHYSICS - - - Q~ l ~ _ ~- FIGURE 4 One of the field tubes, formed as a result of the connection passes. 53 be hundreds of times bigger than the radius of the star itself, the above- mentioned will significantly effect the radiation of the given source directed towards the observer. It may become, for example, the explanation of the variability of some transients; · even a weak quasispheric component always exists in the accreting flow, even at disk accretion on compact objects. That is why in the case considered by us, at least a weak flow of matter is always directed at the non-accreting magnetic pole. Supposing the neutron star to be a compact object, if the pressure of the electromagnetic radiation and the one of the relativistic particles ejected from the polar cap are sufficient high, then the accreting matter from the above-mentioned quasispheric component is "swept" out of the capture region or out of the light cylinder limits. Then the non-accreting magnetic pole will be an ejector, similar to the single magnetized neutron star. REFERENCES Chagelishv~li, G.D ., ERG. Chanishnli, and J. G. Lam inadze. 1986. Pages 563-568. J. G. Proc. Of Joint Varenna-Abastumani Intern. School and Workshop. Sukhumi, USSR (ESA SP-253~. Chagelishvili, GD., J.G Lominadze, and Zip Sokhadze. 1986. Pages 523 529. Proceed. Of Joint Varenna-Abastumani Intern. School and Workshop. Sukhumi, USSR (ESA SP-251~. Chagelishivili, G.D., G. D. Chan Vile, and R.G. Lominadze. 1 988. Advanced Space Research. 8~2~:216. Chagelish~ili, G.D., G.D. Chanishvili, and J.G. Lominadze. 1~8. Ap. Space Sci. 141:361. Furth, H.P., J. Killen, and M.N. Rosenbluth. 1963. Phys. F1. 6: 459. Ghosh, P., and Fed Lamb. 1979. Ap. J. 2~256. Ghosh, P., and F.K Lamb. 1979. Ap. J. 234:296. Krause, F., and RH. Radler. 1980. Mean field magnetohyd~dynamics and dynamo theory, Pergamon Press. Lee, L.C, and Z.F. Fu. Geophysical Research Lett. 1985. 12: 105. Iipunov, V.M. 1987. Astrophysics of neutron stem. Nauka, Moscow.
54 o~11, H.~ 1~8. ~a~edc Beld ~nemllon in elect Enduring Quit. Om~dge ~~ Pa. Sba~, N.I., and ~^ Suit. 1~. Ago Ha. ~:~7. Shape, S.L, and S.^ Nuked 19~. Black holed Wile Bat "d neuron saw. Abe ~1~ and ~= ~ ~ Sou , D), at Sowed, ~.W Dunlop, ARC ~ie~d=~i=, and R.R Gnat 1~. Pit. Sit. ad. :. SO 1~ ~.