花輪 知幸

A new computational scheme is developed to study gas accretion from a circumbinary disk. The scheme decomposes the gas velocity into two components one of which denotes the Keplerian rotation and the other of which does the deviation from it. This scheme enables us to solve the centrifugal balance of a gas disk against gravity with better accuracy, since the former inertia force cancels the gravity. It is applied to circumbinary disk rotating around binary of which primary and secondary has mass ratio, 1.4:0.95. The gravity is reduced artificially softened only in small circular regions around the primary and secondary. The radii are 7% of the binary separation and much smaller than those in the previous grid based simulations. Seven models are constructed to study dependence on the gas temperature and the initial inner radius of the disk. The gas accretion shows both fast and slow time variations while the binary is assumed to have a circular orbit. The time variation is due to oscillation of spiral arms in the circumbinary disk. The masses of primary and secondary disks increase while oscillating appreciably. The mass accretion rate tends to be higher for the primary disk although the secondary disk has a higher accretion rate in certain periods. The accretion rates onto the two components are similar within the fluctuations in late times, i.e., after the binary rotates more than 20 times. The primary disk is perturbed intensely by the impact of gas flow so that the outer part is removed. The secondary disk is quiet in most of time on the contrary. Both the primary and secondary disks have traveling spiral waves which transfer angular momentum within them.

We have developed a new scheme to solve gaseous disks rotating around stars. The scheme evaluates centrifugal force more accurately by decomposing the gas velocity into two components, the reference velocity and deviation from it. The reference velocity is set to be the Keplerian rotation one and gravity is reduced apparently by cancelation with the centrifugal force. Thanks to the reference velocity, the softening radius of gravity can be reduced to only 7% of the binary separation. We apply this method to accretion from a circumbinary disk to unequal mass binary and find that the inflow from L2 point is riot captured directly by the secondary. We also apply this method to classical T Tau binary, V4046 Sgr. The model suggests hot spots on the circumstellar disks as a candidate for the origin of broad Balmer emission lines observed.

Studies of the structure and evolution of protoplanetary disks are important for understanding star and planet formation. Here we present the direct image of an interacting binary protoplanetary system. Both circumprimary and circumsecondary disks are resolved in the near-infrared. There is a bridge of infrared emission connecting the two disks and a long spiral arm extending from the circumprimary disk. Numerical simulations show that the bridge corresponds to gas flow and a shock wave caused by the collision of gas rotating around the primary and secondary stars. Fresh material streams along the spiral arm, consistent with the theoretical scenarios in which gas is replenished from a circummultiple reservoir.

Young circumbinary disks are expected to supply gas onto protoplanetary disks through accretion. Such mass supply has been studied based on new 2D numerical simulations in which a circular binary is surrounded by an isothermal Keplerian disk at the initial stage. The simulations show that the gas supply is variable due to oscillation of spiral waves in the circumbinary disk. The amplitude of the oscillation varies on the timescale of several tens rotation period. The gas supply rate is positively correlated with the amplitude of the oscillation. The gas is supplied through the L2 point to protoplanetary disks associated with the primary and secondary. The mass accretion rate tends to be higher for the primary disk although the secondary disk has a higher accretion rate in certain periods. The primary disk is perturbed so intensely by the impact of gas flow that the outer part is removed. Conversely, the secondary disk is quiet most of the time. Both the primary and secondary disks have travelling spiral waves which transfer angular momentum inside them. The mass supply from the circumbinary disk is higher when the temperature is higher. The inner edge of the circumbinary disk retreats gradually.

We reexamine the similarity solution for a self-gravitating isothermal gas sphere and consider the implications for star formation in turbulent clouds. For adequately chosen parameters, the similarity solution describes an accreting isothermal gas sphere bounded by a spherical shock wave. The mass and radius of the sphere increase in proportion to time, while the central density decreases in proportion to the inverse square of time. The similarity solution is specified by the accretion rate and the infall velocity. The former has a critical value for a given infall velocity: when the accretion rate lies below this value, there exists a pair of similarity solutions determined by the given combination of accretion rate and infall velocity. One of these solutions is confirmed to be unstable against spherical perturbations. The implication is that the gas sphere will collapse to initiate star formation only when the accretion rate is higher than the critical value. We also examine the stability of the similarity solution against nonspherical perturbations, which are found to be damped.

We have developed a numerical simulation scheme for three-dimensional magnetohydrodynamical flow with shocks. The numerical scheme is based on the Roe-type scheme and includes additional numerical diffusion in the direction tangential to the shock front to care the carbuncle instability. The numerical viscosity is added only in the regions where the characteristics of either fast or slow wave converges, i.e., in the regions potentially dangerous to the carbuncle instability. Accordingly the numerical viscosity is as small as that of the Roe scheme except near the shock fronts. It is demonstrated from comparison with the HLLE scheme that the magnetic Reynolds number is higher in the simulations obtained with our scheme. We show application of the scheme to the magnetohydrodynamical simulations of type II supernova. It is also proved that the scheme is free from the odd-even decoupling. (C) 2008 Elsevier Inc. All rights reserved.

We show three-dimensional magnetohydrodynamical simulations of a core-collapse supernova in which the progenitor has magnetic fields inclined to the rotation axis. The simulations employed a simple empirical equation of state in which the pressure of degenerate gas is approximated by piecewise polytropes for simplicity. Energy loss due to neutrinos is not taken into account for simplicity as well. The simulations start from the stage of dynamical collapse of an iron core. The dynamical collapse halts at t-189 ms by the pressure of high-density gas, and a proto-neutron star (PNS) forms. The evolution of the PNS was followed for about 40 ms in typical models. When the initial rotation is mildly fast and the initial magnetic fields are mildly strong, bipolar jets are launched from the upper atmosphere (r similar to 60 km) of the PNS. The jets are accelerated to similar to 3; 10(4) km s(-1), which is comparable to the escape velocity at the footpoint. The jets are parallel to the initial rotation axis. Before the launch of the jets, magnetic fields are twisted by rotation of the PNS. The twisted magnetic fields form torus-shaped multilayers in which the azimuthal component changes alternately. The formation of magnetic multilayers is due to the initial condition in which the magnetic fields are inclined with respect to the rotation axis. The energy of the jet depends only weakly on the initial magnetic field assumed. When the initial magnetic fields are weaker, the time lag is longer between the PNS formation and jet ejection. It is also shown that the time lag is related to the Alfven transit time. Although the nearly spherical prompt shock propagates outward in our simulations, it is an artifact due to our simplified equation of state and neglect of neutrino loss. The morphology of twisted magnetic field and associate jet ejection are, however, not affected by the simplification.

We show three-dimensional numerical simulations on the core collapse of a rotating massive star threaded by strong magnetic fields. The initial magnetic field is nearly uniform in the core and inclined with respect to the rotation axis. We have made 17 models having different angular velocity and magnetic field of the progenitor All the models show nearly spherical prompt shock just after the core bounce. The magnetic field is amplified by rotation in a torus region around the protoneutron star The azimuthal component of magnetic field changes its sign with a semiregular interval in the torus. This is due to the inclination of initial magnetic field. After the twisted magnetic field is accumulated, the bipolar jets are launched along the initial rotation axis. When the initial magnetic field is weak, the magnetic energy is dissipated substantially by numerical diffusion before the launch of jets. © 2008 American Institute of Physics.

We have reexamined gas accretion onto YSO binary from the circum-binary disk on the basis of 21) numerical simulations with very high spatial resolution. The binary was assumed to have either a circular or an elliptic orbit. At the initial stage, the circum-binary disk has an inner edge at Tin >= 1.75a, where a denotes the mean separation of the binary. The disk is assumed to rotate with the Keplerian velocity. We have confirmed that gas accretes from the circum-binary disk through the L-2 point into the Roche lobe. The gas accretes mainly onto the primary after circulating half around the secondary. This means that the accretion decreases the mass ratio. The accretion through gap is due to some pairs of two-armed spiral shock waves. While a pair of them corotates with the binary, the other pairs rotate more slowly, e.g., at ohm = ohm(*)/4 in most models, where ohm(*) denotes the mean angular velocity. Whenever the slowly rotating shock waves get accross the L2 point, the gas flow increases. Thus the accretion rate of the binary changes with the frequency, nu = (3/2) (ohm(*)/2 pi), in the case of a circular orbit. When the orbit is eccentric, the accretion rate changes mainly with the binary orbit but not purely periodically.

We show 3D MHD simulations in which a massive star collapses to form a proto-neutron star having magnetic field inclined to its rotation axis. The simulations take account of equation of state of high density gas as well as self-gravity and magnetic field. We solved the MHD equations with a Roe-type approximate Riemann solver. High spatial resolution and wide dynamic range are realized with nested grids; the whole computation box is covered with a coarse grid, while a smaller region around the center is covered with a finer grid. A typical model is computed with the nested grid consisting of 64(3) x 8 cells. Our simulations show that the magnetic field is amplified by the collapse and rotation of the proto-neutron star. We discuss both the numerical techniques employed and astrophysical implications.

We discuss the evolution of the magnetic flux density and angular velocity in a molecular cloud core, on the basis of three-dimensional numerical simulations, in which a rotating magnetized cloud fragments and collapses to form a very dense optically thick core of > 5 x 10(10) cm(-3). As the density increases towards the formation of the optically thick core, the magnetic flux density and angular velocity converge towards a single relationship between the two quantities. If the core is magnetically dominated its magnetic flux density approaches 1.5(n/5 x 10(10) cm(-3))(1/2) mG, while if the core is rotationally dominated the angular velocity approaches 2.57 x 10(-3) (n/5 x 10(10) cm(-3))(1/2) yr(-1), where n is the density of the gas. We also find that the ratio of the angular velocity to the magnetic flux density remains nearly constant until the density exceeds 5 x 10(10) cm(-3). Fragmentation of the very dense core and emergence of outflows from fragments will be shown in the subsequent paper.

We investigate self-gravitational collapse of magnetized molecular cloud cores and formation of the outflow. We employ a nested grid in order to resolve fine structures of protostar and outflow generation, of which size is as small as 1 AU, and to follow the whole structure of the molecular cloud core, of which radius reaches 0.1-1 pc, simultaneously. The nested grid allows us to follow the evolution of the cloud core with the high dynamic range of 10(5)-10(6) in the spatial resolution. In this paper, we introduce implementations of the self-gravitational MHD nested grid code and show applications to early stages in star formation: gravitational collapse of cloud core, "first core" formation, and bipolar outflow ejection. In both cases of single and binary star formation, magnetic fields play important role in the outflow formation. The outflow region has extremely low beta regions of beta = 10(-6) - 10(-3), and our code shows no sign of numerical instability even in these low-beta regions. (C) 2004 Elsevier B.V. All rights reserved.

We have reexamined accretion in a protobinary system with two dimensional numerical simulations. We consider protostars which rotate around the center of the mass with circular orbits. The accreting gas is assumed to flow in the orbital plane. It is injected from a circle whose radius is 5 times larger than the orbital separation of the binary. The injected gas has constant surface density, infall velocity, and specific angular momentum. The accretion depends on the specific angular momentum of the injected gas, j(inf). When j(inf) is small, the binary accretes the gas mainly through two channels: one through the Lagrangian point L-2 and the other through L-3. When j(inf) is large, the binary accretes the gas only through the L-2 point. The primary accretes more than the secondary in both cases, although the L-2 point is closer to the secondary. After flowing through the L-2 point, the gas flows half around the secondary and through the L-1 point to the primary. Only a small amount of gas flows-back to the secondary and the rest forms a circumstellar ring around the primary. The accretion decreases the mass ratio, q = M-2/M-1, where M-1 and M-2 denote the masses of the primary and secondary, respectively. The accretion rate increases with time. When j(inf) is large, it is negligibly small in the first few-rotation periods.

We observed heads of two molecular pillars in the Eagle Nebula using the Nobeyama Millimeter Array with a spatial resolution of 3, 5, and 3 in the 13CO (J p 1–0) line, C18O (J p 1–0) line, and 2.7 mm continuum, respectively. We found six 13CO subclumps and four C18O cores. The 13CO clouds are elongated so as to have a head-tail structure, with the heads orientated toward the O star exciting M16. The elongation is likely to be due to radiation or wind from the O star. The cloud surface appears to be compressed, as indicated by strong 13CO emission at the cloud rim. The shapes of the 13CO clouds are quite similar to those of the dark cloud observed in the near-infrared. Three out of the four C18O cores are located within one of the 13CO clouds. One of the three cores, located near the tip of the 13CO cloud, is associated with a 2.7 mm continuum emission peak and is most likely a protostar. It is not associated with a known near-infrared source. The other two cores are located farther from the O star and are most likely starless cores. Thus, these C18O cores are aligned in order of age, with the more evolved objects closer to the O star. This linear sequence suggests the propagation of star formation activity, i.e., sequential star formation, driven by the O star. A similar sequence of a young stellar object and a C18O core was found in the other head of the molecular pillar.

We observed heads of two molecular pillars in the Eagle Nebula using the Nobeyama Millimeter Array with a spatial resolution of 3", 5", and 3" in the (CO)-C-13 (J = 1-0) line, (CO)-O-18 (J = 1-0) line, and 2.7 mm continuum, respectively. We found six (CO)-C-13 subclumps and four (CO)-O-18 cores. The (CO)-C-13 clouds are elongated so as to have a head-tail structure, with the heads orientated toward the O star exciting M16. The elongation is likely to be due to radiation or wind from the O star. The cloud surface appears to be compressed, as indicated by strong (CO)-C-13 emission at the cloud rim. The shapes of the (CO)-C-13 clouds are quite similar to those of the dark cloud observed in the near-infrared. Three out of the four (CO)-O-18 cores are located within one of the (CO)-C-13 clouds. One of the three cores, located near the tip of the (CO)-C-13 cloud, is associated with a 2.7 mm continuum emission peak and is most likely a protostar. It is not associated with a known near-infrared source. The other two cores are located farther from the O star and are most likely starless cores. Thus, these (CO)-O-18 cores are aligned in order of age, with the more evolved objects closer to the O star. This linear sequence suggests the propagation of star formation activity, i.e., sequential star formation, driven by the O star. A similar sequence of a young stellar object and a (CO)-O-18 core was found in the other head of the molecular pillar.

We show three-dimensional numerical simulations of a dynamically collapsing cloud core. We adopted a nested grid to resolve the first core embedded in a molecular cloud core. The first core is elongated and filamentary in a typical model. The elongation is due to the bar mode instability. The degree of the elongation depends on the amplitude of the initial non-axisymmetric perturbation. In most models the first core is surrounded by an infalling disk-like envelope. The very elongated, filamentary first core fragments. The less elongated first core does not fragment and becomes disk shape. The critical elongation for fragmentation is about (a(l) - a(s))/(a(l) + a(s)) = 0.5, where a(l) and a, denote the lengths of dense region along the long and short axes in the midplane. The disk-like first cores are subjected to bar instability again, and seem to fragment by fission at the later stages.

We present a second order accurate numerical scheme for ideal magnetohydrodynamical equations. Our numerical scheme is an upwind scheme based on the flux vector splitting and an extension of Roe'S scheme. The second order accuracy is achieved by MUSCL, the Monotone Upstream-centered Scheme for Conservation Laws approach. Our numerical scheme ensures the total variation diminishing (TVD) as far as Courant number is smaller than unity.