Stars are likely to form with non-zero initial angular momentum, and will consequently possess a substantial gaseous protostellar disc in the early phases of their evolution. At this early stage, the disc mass is expected to be comparable to the mass of the protostar. The disc’s self-gravity therefore plays an important role in the subsequent evolution of the system, regulating the accretion of matter onto the protostar, as well as being potentially capable of forming low mass stars and massive planets by disc fragmentation. The protostellar disc may later evolve into a protoplanetary disc, providing the feedstock for planet formation. Therefore, if the current stellar populations and exoplanetary systems are to be understood, an understanding of the evolution of protostellar discs is crucial, especially their earliest self-gravitating phases. I have used various methods of numerical simulation to probe the physics of self-gravitating protostellar discs and their constituents. When constructing a model for self-gravitating protostellar discs, including detailed thermodynamics and radiative transfer is essential. I have developed two distinct numerical techniques for incorporating radiative transfer into Smoothed Particle Hydrodynamics (SPH) simulations. The first allows the modelling of frequency-averaged radiative transfer during the SPH simulation, in effect approximating radiative SPH (RSPH) with only a marginal increase in runtime (around 6%). The second takes the output from SPH simulations, and creates synthetic, wavelength-dependent telescope images and spectra of SPH systems. This allows the direct construction of observables from SPH simulations, providing, for the first time, a direct connection between the output of SPH simulations and observations. I have used these numerical methods to analyse, in detail, the local angular momentum transport induced by self-gravity in protostellar discs, testing the robustness of the “pseudo-viscous” analytical approximation for local disc stresses. I confirm that semi-analytical disc modellers are justified in using the pseudo-viscous approximation in some cases, but I also outline the limits in which non-local transport effects causes the approximation to fail. Also, I have investigated the evolution of protostellar discs when perturbed by a secondary companion, in particular identifying whether such events will in general trigger a) a disc fragmentation event, or b) a stellar outburst event. For case a), I found no significant evidence that perturbation by a companion improves the possibility of disc fragmentation in compact discs - in case b), I found that stellar outburst events do indeed occur, but they are unlikely to be seen by observers due to their rare occurrence, as well as due to self-obscuration effects.

<p> Fragmentation during encounters between protostellar disks provides a possible scenario for the formation of substellar objects such as brown dwarfs and planets. A series of simulations of protostellar disk encounters were performed to investigate the fragmentation under different encounter parameters, and to characterize the properties of any resultant fragments. It was found that the initial disk minimum Toomre Q must satisfy Qini ;S 1.1 for the fragmentation to be induced by the encounters. Fragments of substellar mass can form via disk fragmentation, shock layer fragmentation and tidal tail fragmentation, and the effectiveness of each mechanism is closely related to the initial disk configuration. The fragmentation is also constrained by the relative encounter velocity since the number of fragments decreases quickly with increasing velocity. </p> <p> In comparing to previous studies of protostellar disk encounters it was also found that resolving both the local Jeans Mass during the encounter and the disks' vertical structure are critical to prevent artificial fragmentation and give the correct picture. Heating and cooling rates were estimated in both the optically thin and thick regimes. The comparison between the two indicates that during strong impacts the heating rate increases rapidly but is still comparable to the cooling rate, so the locally isothermal equation of state used in this study is an acceptable approximation. </p> <p> 32 clumps formed in various Qini = 0.9 disk-disk encounters were taken as the sample in an analysis of fragment properties and prospects for their further evolution. The results show that the clump masses are all less than the hydrogen burning mass limit ~ 0.075M0 , so the objects are substellar. Most of the clumps are of brown dwarf mass since the formation of planetary mass clumps is suppressed due to numerical resolution. The mass distribution is broadly consistent to the observed initial mass function in Pleiades. The clumps have highly flattened disk-like shapes and possess large spin angular momentum, which implies that young brown dwarfs may develop disks, jets, or planetary mass companions. About one third of the fragments are unbound to the stars and likely to form free floating brown dwarfs. Orbital analyses of the clumps which are bound to the stars show that there is a lack of close brown dwarf companions ( R < 3 AU), which is consistent to the observed "brown dwarf desert". Many of the orbits are highly eccentric and intersect with other orbits, so ejection of some clumps due to gravitational scattering is likely. Also, dispersion of gas during the encounter and the high spin angular momentum of the clumps may provide mechanisms other than ejection to prevent the clumps from accreting more mass, making the simulated clumps representative of the long term substellar mass function. </p> / Thesis / Master of Science (MSc)

La formation des étoiles massives reste, à ce jour, encore mal connue à cause de l’extrême quantité d’énergie que ces étoiles dégagent, limitant en conséquence leurs masses théoriques et contredisant les observations de ce type d’étoile. Les observatoires du futur (en particulier l’observatoire spatial Herschel) vont tenter de répondre à cette problématique grâce notamment aux émissions moléculaires de l’eau. L’analyse précise et correcte de ces données, dans l’avenir, nécessite donc dès aujourd’hui un travail associant des observations et des modélisations des objets concernés. C’est dans ce but que cette thèse a consisté en l’élaboration d’une méthode de modélisation dite « globale » d’objets protostellaires massifs (proto-amas ou cœurs denses massifs). Celle-ci a permis une description physique et une étude chimique des multiples cœurs denses massifs étudiées, et a ouvert de nombreuses voies vers des aspects évolutifs. Elle a également donné des indices pour a?ner le programme d’observation en temps garanti WISH des raies moléculaires de l’eau et con?rmé le rôle clef de cette molécule pour la compréhension de la formation des étoiles massives. / Today the formation of massive stars is still not well understood due to the huge interac- tion of these objects with their environment, leading to a theoretical limit in the ?nal mass that observations contradict. The future observatories, like the Herschel Space Observatory, will try to answer some of the questions linked to this topic, particularly through the water line emissions. The correct and precise analysis of the future data is then necessary and needs a full work linking the observations and the modelling of the objects that will be studied. Hence the main goal of this PhD Thesis was to elaborate a robust and global modeling method of the massive dense cores in which high-mass stars are forming. The method leaded to a physical description and a chemical study of multiple massive dense cores, opening new views on evolution aspects. In addition it gave some tweaks on the guaranteed-time key program WISH for the water line emissions and con?rmed the key role of this molecule for a better understanding of the high-mass star formation.

A long-standing problem in low-mass star formation is the "luminosity problem," whereby protostars are underluminous compared to the accretion luminosity expected both from theoretical collapse calculations and arguments based on the minimum accretion rate necessary to form a star within the embedded phase duration. In this dissertation, I present new research on protostars and the protostellar accretion process that addresses the luminosity problem in the following ways: I report new infrared detections of a very low luminosity protostar in Taurus and use all existing data ranging from the infrared through millimeter wavelengths to constrain radiative transfer models and determine physical properties of the source. I argue that the derived source luminosity is lower than that expected based on the properties of a previously detected molecular outflow driven by this source and suggest that this discrepancy can be resolved by variable rather than constant mass accretion. I report the discovery of a new protostar that is also driving a molecular outflow. Following a similar modeling procedure as above, I show that this source has an even lower luminosity that is once again inconsistent with that expected based on the properties of its outflow, again suggesting variable mass accretion. I present the results of a complete search for all protostars with luminosities less than or equal to that of our Sun in a new infrared survey of nearby star-forming regions. I identify 50 protostars with such luminosities. Only a small fraction (15-25%) of dense cores thought to be starless (not yet collapsing to form stars) in fact harbor low luminosity protostars. The distribution of luminosities of these 50 protostars is inconsistent with a constant protostellar mass accretion rate. I present a set of evolutionary models that start with existing models following the inside-out collapse of singular isothermal spheres and add isotropic scattering off dust grains, a circumstellar disk, two-dimensional envelope structure, mass-loss and the opening of outflow cavities, and a simple treatment of episodic mass accretion. I conclude that episodic mass accretion is both necessary and sufficient to resolve the luminosity problem. / text

Gasausströmungen, oft in der Form hoch kollimierter Jets, sind ein allgegenwärtiges Phänomen bei der Geburt neuer Sterne. Emission von stossangeregtem molekularem Wasserstoff bei Wellenlängen im nahen Infrarotbereich ist ein Merkmal ihrer Existenz und auch in eingebetteten, im Optischen obskurierten Ausströmungen generell gut zu beobachten. In dieser Arbeit werden die Resultate einer von Auswahleffekten freien, empfindlichen, grossflächigen Suche nach solchen Ausströmungen von Protosternen in der v=1-0 S(1) Linie molekularen Wasserstoffs bei einer Wellenlänge von 2.12 µm vorgestellt. Die Durchmusterung umfasst eine Fläche von etwa einem Quadratgrad in der Orion A Riesenmolekülwolke. Weitere Daten aus einem grossen Wellenlängenbereich werden benutzt, um die Quellen der Ausströmungen zu identifizieren. Das Ziel dieser Arbeit ist es, eine Stichprobe von Ausströmungen zu bekommen, die so weit wie möglich frei von Auswahleffekten ist, um die typischen Eigenschaften protostellarer Ausströmungen und deren Entwicklung festzustellen, sowie um die Rückwirkung der Ausströmungen auf die umgebende Wolke zu untersuchen.<br /> Das erste Ergebnis ist, dass Ausströmungen in Sternentstehungsgebieten tatsächlich sehr häufig sind: mehr als 70 Jet-Kandidaten werden identifiziert. Die meisten zeigen eine sehr irreguläre Morphologie anstelle regulärer oder symmetrischer Strukturen. Dies ist auf das turbulente, klumpige Medium zurückzuführen, in das sich die Jets hineinbewegen. Die Ausrichtung der Jets ist zufällig verteilt. Insbesondere gibt es keine bevorzugte Ausrichtung der Jets parallel zum grossräumigen Magnetfeld in der Wolke. Das legt nahe, dass die Rotations- und Symmetrieachse in einem protostellaren System durch zufällige, turbulente Bewegung in der Wolke bestimmt wird. <br /> <br /> Mögliche Ausströmungsquellen werden für 49 Jets identifiziert; für diese wird der Entwicklungsstand und die bolometrische Leuchtkraft abgeschätzt. Die Jetlänge und die H2 Leuchtkraft entwickeln sich gemeinsam mit der Ausströmungsquelle. Von null startend, dehnen sich die Jets schnell bis auf eine Länge von einigen Parsec aus und werden dann langsam wieder kürzer. Sie sind zuerst sehr leuchtkräftig, die H2 Helligkeit nimmt aber im Lauf der protostellaren Entwicklung ab. Die Längen- und H2 Leuchtkraftentwicklung lässt sich im Wesentlichen durch eine zuerst sehr hohe, dann niedriger werdende Massenausflussrate erklären, die auf eine zuerst sehr hohe, dann niedriger werdende Gasakkretionsrate auf den Protostern schliessen lässt (Akkretion und Ejektion sind eng verknüpft!). Die Längenabnahme der Jets erfordert eine ständig wirkende Abbremsung der Jets. Ein einfaches Modell einer simultanen Entwicklung eines Protosterns, seiner zirkumstellaren Umgebung und seiner Ausströmung (Smith 2000) kann die gemessenen H2- und bolometrischen Leuchtkräfte der Jets und ihrer Quellen reproduzieren, unter der Annahme, dass die starke Akkretionsaktivität zu Beginn der protostellaren Entwicklung mit einer überproportional hohen Massenausflussrate verbunden ist.<br /> <br /> Im Durchmusterungsgebiet sind 125 dichte Molekülwolkenkerne bekannt (Tatematsu et al. 1993). Jets (bzw. Sterne) entstehen in ruhigen Wolkenkernen, d.h. solchen mit einem niedrigen Verhältnis von interner kinetischer Energie zu gravitativer potentieller Energie; dies sind die Wolkenkerne höherer Masse. Die Wolkenkerne mit Jets haben im Mittel grössere Linienbreiten als die ohne Jets. Dies ist darauf zurückzuführen, dass sie bevorzugt in den massereicheren Wolkenkernen zu finden sind, welche generell eine grössere Linienbreite haben. Es gibt keinen Hinweis auf stärkere interne Bewegungen in Wolkenkernen mit Jets, die durch eine Wechselwirkung der Jets mit den Wolkenkernen erzeugt sein könnte. Es gibt, wie von der Theorie vorausgesagt, eine Beziehung zwischen der Linienbreite der Wolkenkerne und der H2 Leuchtkraft der Jets, wenn Jets von Klasse 0 und Klasse I Protosternen separat betrachtet werden; dabei sind Klasse 0 Jets leuchtkräftiger als Klasse I Jets, was ebenfalls auf eine zeitabhängige Akkretionsrate mit einer frühzeitigen Spitze und einem darauffolgenden Abklingen hinweist.<br /> <br /> Schliesslich wird die Rückwirkung der Jetpopulation auf eine Molekülwolke unter der Annahme strikter Vorwärtsimpulserhaltung betrachtet. Die Jets können auf der Skala einer ganzen Riesenmolekülwolke und auf den Skalen von Molekülwolkenkernen nicht genügend Impuls liefern, um die abklingende Turbulenz wieder anzuregen. Auf der mittleren Skala von molekularen Klumpen, mit einer Grösse von einigen parsec und Massen von einigen hundert Sonnenmassen liefern die Jets jedoch genügend Impuls in hinreichend kurzer Zeit, um die Turbulenz “am Leben zu erhalten” und können damit helfen, einen Klumpen gegen seinen Kollaps zu stabilisieren. / The presence of outflows, often in the form of well-collimated jets, is a phenomenon commonly associated with the birth of young stars. Emission from shock-excited molecular hydrogen at near-infrared wavelengths is one of the signposts of the presence of such an outflow, and generally can be observed even if the flow is obscured at optical wavelengths. In this thesis, I present the results of an unbiased, sensitive, wide-field search for flows from protostellar objects in the H2 v=1-0 S(1) line at a wavelength of 2.12 µm, covering a 1 square degree area of the Orion A giant molecular cloud. Further data covering a wide wavelength range are used to search for the driving sources of the flows. The aim of this work is to obtain a sample of outflows which is free from biases as far as possible, to derive the typical properties of the outflows, to search for evolutionary trends, and to examine the impact of outflows on the ambient cloud.<br /> The first result from this survey is that outflows are indeed common in star forming regions: more than 70 candidate jets are identified. Most of them have a fairly ill-defined morphology rather than a regular or symmetric structure, which is interpreted to be due to the turbulent, clumpy ambient medium into which the jets are propagating. The jets are randomly oriented. In particular, no alignment of the jets with the large scale ambient magnetic field is found, suggesting that the spin and symmetry axis in a protostellar object is determined by random, turbulent motions in the cloud. <br /> <br /> Candidate driving sources are identified for 49 jets, and their evolutionary stage and bolometric luminosity is estimated. The jet lengths and H2 luminosities evolve as a function of the age of the driving source: the jets grow quickly from zero length to a size of a few parsec and then slowly shorten again. The jets are very luminous early on and fade during the protostellar evolution. The evolution in length and H2 luminosity is attributed to an early phase of strong accretion, which subsequently decreases. The shortening of the jets with time requires the presence of a continuous deceleration of the jets. A simple model of the simultaneous evolution of a protostar, its circumstellar environment, and its outflow (Smith 2000) can reproduce the measured values of H2 luminosity and driving source luminosity under the assumption of a strong accretion plus high ejection efficiency phase early in the protostellar evolution.<br /> <br /> Tatematsu et al. (1993) found 125 dense cloud cores in the survey area. The jet driving sources are found to have formed predominantly in quiet cores with a low ratio of internal kinetic energy to gravitational potential energy; these are the cores with higher masses. The cores which are associated with jets have on average larger linewidths than cores without jets. This is due to the preferred presence of jets in more massive cores, which generally have larger linewidths. There is no evidence for additional internal motions excited by the interaction of the jets with the cores. The jet H2 luminosity and the core linewidth (as predicted by theory) are related, if Class 0 and Class I jets are considered separately; the relation lies at higher values of the H2 luminosity for the Class 0 jets than for Class I jets. This also suggests a time evolution of the accretion rate, with a strong peak early on and a subsequent decay.<br /> <br /> Finally, the impact of a protostellar jet population on a molecular cloud is considered. Under the conservative assumption of strict forward momentum conservation, the jets appear to fail to provide sufficient momentum to replenish decaying turbulence on the scales of a giant molecular cloud and on the scales of molecular cloud cores. At the intermediate scales of molecular clumps with sizes of a few parsec and masses of a few hundred solar masses, the jets provide enough momentum in a short enough time to potentially replenish turbulence and thus might help to stabilize the clump against further collapse.

<p>We use numerical simulations to examine the fragmentation of protostellar discs via gravitational instability (GI), a proposed formation mechanism for gas-giant planets and brown dwarfs. To accurately model heating and cooling, we have implemented radiative transfer (RT) in the TreeSPH code Gasoline, using the flux-limited diffusion approximation coupled to photosphere boundary cooling. We present 3D radiation hydrodynamics simulations of discs that are gravitationally unstable in the inner 40 AU; these discs do not fragment because the cooling times are too long. In prior work, one of these discs was found to fragment; however, we demonstrate that this resulted from an over-estimate of the photosphere cooling rate. Fragmentation via GI does not appear to be a viable formation mechanism in the inner 40 AU.</p> <p>We also present simulations of GI in the outer regions of discs, near 100 AU, where we find GI to be a viable formation mechanism. We give a detailed framework that explains the link between cooling and fragmentation: spiral arms grow on a scale determined by the linear gravitational instability, have a characteristic width determined by the balance of heating and cooling, and fragment if this width is less than twice their Hill radius. This framework is consistent with the fragmentation and initial fragment masses observed in our simulations. We apply the framework to discs modelled with the commonly-used beta-prescription cooling and calculate the critical cooling rate for the first time, with results that are consistent with previous estimates measured from numerical experiments.</p> <p>RT is fundamentally important in the star formation process. Non-ionizing radiation heats the gas and prevents small-scale fragmentation. Ionizing radiation from massive stars is an important feedback mechanism and may disrupt giant molecular clouds. We present methods and tests for our implementation of ionizing radiation, using the Optically-Thin Variable Eddington Tensor method.</p> / Doctor of Philosophy (PhD)

<p>Protostellar discs are generally thought to drive molecular outflows and jets observed in star forming regions, but there has been some debate as to how they form. The details of the driving and collimation of outflows help determine how much mass is cleared out and how much energy is fed back into the surroundings. Recently it has been argued that the magnetic brake is so strong that early protostellar disks cannot form.</p> <p>We have performed 3D ideal magnetohydrodynamic (MHD) simulations of collapsing Bonnor–Ebert spheres, employing sink particles within an AMR grid and using a cooling function to model radiative cooling of the gas. This allows us to follow the formation and early evolution of the accretion disc (2−8)×10⁴ years further into the Class 0 phase of its evolution. We form a rotationally dominated disc with a radius of 100 AU embedded inside a transient, unstable, flattened, rotating structure extending out to 2000 AU. The inner disc becomes unstable to a warping instability due to the magnetic structure of the outflow, warping 30 deg with respect to the rotation–axis by the end of the simulation. The disc is unstable to a Parker instability and sheds magnetic loops, degrading the orientation of the mean threading field. This reduces and locally reverses the magnetic braking torque of the large scale field back upon the disc. The reduction of magnetic braking allows a nearly Keplerian disc to form and may be the key way in which low mass stellar systems produce rotationally dominated discs. We discuss the relevance of our disc misalignment concerning the formation of mis–aligned hot Jupiters.</p> <p>Protostellar outflows are implicated in clearing mass from collapsing cores, and limiting the final mass of newly formed stars. The details of the driving and collimation of outflows help determine how much mass is cleared out and how much energy is fed back into the surroundings. The simulations generate outflows which are precessing, kinked, contain internal shocks and extend to a scale of 0.1 pc end–to–end. Our disc–wind theory describes magneto–centrifugal driving throughout the outflow bubble. The bulk properties of the outflow agree well with observations. The outflow has two components, a larger low speed wind (v_r < 1.5 km/s) dominated by a toroidal magnetic field Bφ, and an inner centrifugally driven jet dominated by Bp with speeds up to 20 km/s. The ratio of mass flux from the disk surface com- pared to accretion in the disk is measured to be M_out/M_in ∼ 0.1 from the inner component, whereas in the outer component M_out/M_in ∼ 1.0. The jet is misaligned and precesses as the disc warps by 30 deg with respect to the z–axis. We measure star formation efficiencies of ε_core = 0.63 (and growing), higher than theoretical predictions of ε_core = 0.29−0.39 and observations ε_core = 0.33.</p> <p>These new results reported in this thesis, show that disks can form in strongly magnetized media, in agreement with the observations - and that outflows are not as efficient in clearing away collapsing gas as has been assumed in various theoretical models. Both of these results have important implications for disk formation, and the origin of the IMF, as described in this work.</p> / Doctor of Philosophy (PhD)

In this thesis, I study numerical and semi-analytical models of self-gravitating protostellar discs, with the aim of furthering our understanding of the role of disc-self gravity in planet formation. At the time of writing, the ALMA era of observational astronomy is upon us. Therefore, I place my research into this context with synthetic images of both numerical and semi-analytical models. I begin with an examination into the apparent lack of convergence, with increasing resolution, of the fragmentation boundary in Smoothed Particle Hydrodynamics (SPH) simulations of a protostellar disc. I run a suite of SPH with different numerical implementations, and find that even very similar implementations can fundamentally change the final answer. I analyse a suite of SPH simulations that fragment to form gravitationally bound objects, with the motivation of informing future population synthesis model development. I find that fragment-fragment and fragment-disc interaction dominates the orbital evolution of the system even at very early times, and any attempt to produce a population of objects from the gravitational instability process must include these interactions. Before a disc fragments, it will go through a self-gravitating phase. If the disc cools globally on a timescale such that it is balanced by heating due to gravitational stresses, the disc will be in a state of quasi-equilibrium. So long as the disc mass is sufficiently low, and spirals are sufficiently tightly wound, then angular momentum transport can be described by the local approximation, for which there is an analytical description. Using this analytical description, I develop an existing 1D model into 3D, and examine a wide range of parameter space for which disc self-gravity produces significant non-axisymmetry. Using radiative transfer calculations coupled with synthetic observations, I determine that there is a very narrow range of parameter space in which a disc will have sufficiently large gravitational stresses so as to produce detectable spirals, but the stresses not be so large as to cause the disc to fragment. By developing a simple analytical prescription for dust, I show that this region of parameter space can be broadened considerably. However, it requires grains that are large enough to become trapped by pressure maxima in the disc, so I conclude that if self-gravitating spiral arms are detected in the continuum, it is likely that at least some grain growth has taken place.

<p>Stars form in clusters amidst complex and coupled physical phenomena. Among the most important of these is radiative feedback, which heats the surrounding gas to suppress the formation of many low-mass stars. In simulations of star formation, pre-main-sequence modeling has often been neglected and stars are assumed to have the radii and luminosities of zero-age main sequence stars. We challenge this approach by developing and integrating a one-zone protostellar evolution model for FLASH and using it to regulate the radiation output of forming stars. The impact of accurate pre-main-sequence models is less ionizing radiation and less heating during the early stages of star formation. For stars modeled in isolation, the effect of protostellar modeling resulted in ultracompact HII regions that formed slower than in the ZAMS case, but also responded to transitions in the star itself. The HII region was seen to collapse and subsequently be rebuilt as the star underwent a swelling of its radius in response to changes in stellar structure and nuclear burning. This is an important effect that has been missed in previous simulations. It implies that observed variations in HII regions may signal changes in the stars themselves, if these variation can be disentangled from other environmental effects seen in the chaotic cluster environment.</p> / Master of Science (MSc)