The driving mechanisms of low-and high-velocity outflows in star formation processes are studied using threedimensional resistive MHD simulations. Starting with a Bonnor-Ebert isothermal cloud rotating in a uniform magnetic field, we calculate cloud evolution from the molecular cloud core (n c ¼ 10 4 cm À3

The evolution of the magnetic field and angular momentum in the collapsing cloud core is studied using three-dimensional resistive MHD nested grid simulations. Starting with a Bonnor-Ebert isothermal cloud rotating in a uniform magnetic field, we calculate the cloud evolution from the molecular cloud core (n=10^4 cm^-3) to the stellar core (n \simeq 10^22 cm^-3). The magnetic field strengths at the center of the clouds converge to a certain value as the clouds collapse, when the clouds have the same angular momenta but different strengths of the magnetic fields at the initial state. For 10^12 cm^-3 < n < 10^16 cm^-3, Ohmic dissipation considerably removes the magnetic field from the collapsing cloud core, and the magnetic field lines, which are strongly twisted for n <10^12 cm^-3, is de-collimated. The magnetic field lines are twisted and amplified again for nc > 10^16 cm^-3, because the magnetic field is recoupled with the warm gas. Finally, protostars at their formation epoch have 0.1-1kG of the magnetic fields, which are comparable to observations. The magnetic field strength of protostar slightly depends on the angular momentum of the host cloud. The protostar formed from the slowly rotating cloud core has a stronger magnetic field. The evolution of the angular momentum is closely related to the evolution of the magnetic field. The angular momentum in the collapsing cloud is removed by the magnetic effect. The formed protostars have 0.1-2 days of the rotation period at their formation epoch, which are slightly shorter than the observation. This indicates that the further removal mechanism of the angular momentum such as interaction between the protostar and disk, wind gas or jet is important in further evolution of the protostar.Comment: 39 pages,11 figures, Submitted to ApJ, For high resolution figures see http://www2.scphys.kyoto-u.ac.jp/~machidam/protostar/proto/ms.pd

We report the first three-dimensional radiation magnetohydrodynamic (RMHD) simulations of protostellar collapse with and without Ohmic dissipation.We take into account many physical processes required to study star formation processes, including a realistic equation of state. We follow the evolution from molecular cloud cores until protostellar cores are formed with sufficiently high resolutions without introducing a sink particle. The physical processes involved in the simulations and adopted numerical methods are described in detail.We can calculate only about one year after the formation of the protostellar cores with our direct three-dimensional RMHD simulations because of the extremely short timescale in the deep interior of the formed protostellar cores, but successfully describe the early phase of star formation processes. The thermal evolution and the structure of the first and second (protostellar) cores are consistent with previous one-dimensional simulations using full radiation transfer, but differ considerably from preceding multi-dimensional studies with the barotropic approximation. The protostellar cores evolve virtually spherically symmetric in the ideal MHD models because of efficient angular momentum transport by magnetic fields, but Ohmic dissipation enables the formation of the circumstellar disks in the vicinity of the protostellar cores as in previous MHD studies with the barotropic approximation. The formed disks are still small (less than 0.35 AU) because we simulate only the earliest evolution. We also confirm that two different types of outflows are naturally launched by magnetic fields from the first cores and protostellar cores in the resistive MHD models.

The formation and evolution of the circumstellar disk in unmagnetized molecular clouds is investigated using three-dimensional hydrodynamic simulations from the prestellar core until the end of the main accretion phase. In collapsing cloud cores, the first (adiabatic) core with a size of ∼ 10 AU forms prior to the formation of the protostar. At its formation, the first core has a thick disk-like structure, and is mainly supported by the thermal pressure. After the protostar formation, it decreases the thickness gradually, and becomes supported by the centrifugal force. We found that the first core is a precursor of the circumstellar disk. This indicates that the circumstellar disk is formed before the protostar formation with a size of ∼ 10 AU, which means that no protoplanetary disk smaller than < 10 AU exists. Reflecting the thermodynamics of the collapsing gas, at the protostar formation epoch, the first core (or the circumstellar disk) has a mass of ∼ 0.01 − 0.1 M ⊙ , while the protostar has a mass of ∼ 10 −3 M ⊙ . Thus, just after the protostar formation, the circumstellar disk is about 10 − 100 times more massive than the protostar. Even in the main accretion phase that lasts for ∼ 10 5 yr, the circumstellar disk mass dominates the protostellar mass. Such a massive disk is unstable to gravitational instability, and tends to show fragmentation. Our calculations indicate that the planet or brown-dwarf mass object may form in the circumstellar disk in the main accretion phase. In addition, the mass accretion rate onto the protostar shows strong time variability that is caused by the perturbation of proto-planets and/or the spiral arms in the circumstellar disk. Such variability provides a useful signature for detecting the planet-sized companion in the circumstellar disk around very young protostars.

We investigate gas accretion on to a protoplanet, by considering the thermal effect of gas in three‐dimensional hydrodynamical simulations, in which the wide region from a protoplanetary gas disc to a Jovian radius planet is resolved using the nested grid method. We estimate the mass accretion rate and growth time‐scale of gas giant planets. The mass accretion rate increases with protoplanet mass for Mp < Mcri, while it becomes saturated or decreases for Mp > Mcri, where Mcri≡ 0.036 MJup(ap/1 au)0.75, and MJup and ap are the Jovian mass and the orbital radius, respectively. This accretion rate is typically two orders of magnitude smaller than that in two‐dimensional simulations. The growth time‐scale of a gas giant planet or the time‐scale of the gas accretion on to the protoplanet is about 105 yr, that is two orders of magnitude shorter than the growth time‐scale of the solid core. The thermal effects barely affect the mass accretion rate because the gravitational energy dominates the thermal energy around the protoplanet. The mass accretion rate obtained in our local simulations agrees quantitatively well with those obtained in global simulations with coarser spatial resolution. The mass accretion rate is mainly determined by the protoplanet mass and the property of the protoplanetary disc. We find that the mass accretion rate is correctly calculated when the Hill or Bondi radius is sufficiently resolved. Using the oligarchic growth of protoplanets, we discuss the formation time‐scale of gas giant planets.