Abstract
Star and planet formation is a hierarchical, multi-stage process governed by complex dynamics. The earliest phase, spanning from the collapse of a molecular cloud core to the emergence of a protostar and protostellar disk, establishes the initial conditions for the subsequent evolution of the star-disk system and eventual planet formation. However, the detailed evolution of this early stage and the interactions among key ingredients—such as gas, dust, and magnetic fields—remain poorly understood. Observational challenges arise due to the small size of the protostar, obscuration by the surrounding envelope, and the rapid timescales of early evolution. As a result, numerical simulations are essential for uncovering the fundamental physical processes shaping this stage. This thesis provides theoretical and numerical insights into the early stages of star formation from three distinct projects, each focusing on a key aspect of the problem.
The first project investigates dust dynamics and grain growth. Observations suggest that grain growth begins early in the star formation process, yet the underlying mechanisms driving this growth remain unclear. Using numerical models that incorporate gas dynamics, grain dynamics, and grain growth, we demonstrate that grain growth is slow in laminar protostellar disks, indicating the need for additional mechanisms to accelerate the process. Furthermore, we derive a simple analytic estimate for the grain growth timescale, providing a useful framework for interpreting observational constraints.
The second project examines the collapse of a dense molecular cloud core leading to the formation of stars and disks. Using non-ideal magnetohydrodynamic (MHD) models, we show that the evolution of the protostellar envelope is governed by the formation of ``gravo-magneto-sheetlets''—thin, dense structures that arise from the interplay between turbulence, gravity, and magnetic fields. These sheetlets play a crucial role in delivering mass, angular momentum, and magnetic flux to the forming protostar and protostellar disk, with the efficiency of this process regulated by non-ideal MHD effects, which redistribute magnetic flux relative to the gas. If a molecular cloud core possesses sufficient mass, initial angular momentum, and strong non-ideal MHD effects, it can undergo "fragmentation", forming multiple stars. After the first protostar emerges, a "Dense Rotation-Dominated (DROD) structure" develops between the gravo-magneto-sheetlets and the newly formed star. The dynamics within this DROD can give rise to companion stars, provided that the structure is sufficiently demagnetized. This "DROD-fragmentation mechanism" offers a novel pathway for the formation of close, 100-au scale multiple systems with misaligned protostellar disks, providing new insights into the origins of compact stellar multiples.
The third project of this thesis explores the launching mechanisms of jets and outflows from Young Stellar Objects (YSOs). Using a combination of 2D axisymmetric and 3D global non-ideal MHD models, we demonstrate that a circumstellar disk alone—if threaded by a vertical open magnetic field—can launch both a jet and a disk wind driven primarily by toroidal magnetic pressure. While this mechanism shares some similarities with the traditional magneto-centrifugal and magnetic-tower models in terms of outflow properties, it operates through a distinct process that questions some observational interpretations. Additionally, disk-driven outflows tend to develop asymmetries between the two hemispheres due to asymmetry in magnetic field geometry and configuration. In contrast, simulations that include the stellar magnetosphere reveal that magnetosphere-disk interactions generate more powerful and highly collimated bipolar jets. The jet is launched through a ``load–fire–reload'' cycle, in which differential rotation between the star and disk winds up toroidal magnetic fields ("load"), the resulting pressure gradient accelerates gas vertically into a jet ("fire"), and magnetic reconnection resets the field configuration ("reload"). In this case, the field geometry naturally favors the formation of bipolar outflows by channeling rotational energy from both the star and disk into the polar region.
