We present Phantom, a fast, parallel, modular and low-memory smoothed particle hydrodynamics and magnetohydrodynamics code developed over the last decade for astrophysical applications in three dimensions. The code has been developed with a focus on stellar, galactic, planetary and high energy astrophysics and has already been used widely for studies of accretion discs and turbulence, from the birth of planets to how black holes accrete. Here we describe and test the core algorithms as well as modules for magnetohydrodynamics, self-gravity, sink particles, dust-gas mixtures, H2 chemistry, physical viscosity, external forces including numerous galactic potentials, Lense-Thirring precession, Poynting-Robertson drag and stochastic turbulent driving. Phantom is hereby made publicly available.

We present an investigation into the morphological features of the Milky Way. We use smoothed particle hydrodynamics (SPH) to simulate the interstellar medium (ISM) in the Milky Way under the effect of a number of different gravitational potentials representing spiral arms and bars, assuming the Milky Way is grand design in nature. The gas is subject to ISM cooling and chemistry, enabling us to track the evolution of molecular gas. We use a 3D radiative transfer code to simulate the emission from the SPH output, allowing for the construction of synthetic longitude-velocity (l-v ) emission maps as viewed from the Earth. By comparing these maps with the observed emission in CO from the Milky Way, we infer the arm/bar geometry that provides a best fit to our Galaxy. We find that it is possible to reproduce nearly all features of the l-v diagram in CO emission. There is no model, however, that satisfactorily reproduces all of the features simultaneously. Models with 2 arms cannot reproduce all the observed arm features, while 4 armed models produce too bright local emission in the inner Galaxy. Our best-fitting models favour a bar pattern speed within 50-60km s −1 kpc −1 and an arm pattern speed of approximately 20km s −1 kpc −1 , with a bar orientation of approximately 45 • and arm pitch angle between 10 • -15 • .

Aims. This work aims to improve the current understanding of the interaction between H  regions and turbulent molecular clouds. We propose a new method to determine the age of a large sample of OB associations by investigating the development of their associated H  regions in the surrounding turbulent medium.Methods. Using analytical solutions, one-dimensional (1D), and three-dimensional (3D) simulations, we constrained the expansion of the ionized bubble depending on the turbulence level of the parent molecular cloud. A grid of 1D simulations was then computed in order to build isochrone curves for H  regions in a pressure-size diagram. This grid of models allowed us to date a large sample of OB associations that we obtained from the H  Region Discovery Survey (HRDS). Results. Analytical solutions and numerical simulations showed that the expansion of H  regions is slowed down by the turbulence up to the point where the pressure of the ionized gas is in a quasi-equilibrium with the turbulent ram pressure. Based on this result, we built a grid of 1D models of the expansion of H  regions in a profile based on Larson's laws. We take the 3D turbulence into account with an effective 1D temperature profile. The ages estimated by the isochrones of this grid agree well with literature values of well known regions such as Rosette, RCW 36, RCW 79, and M 16. We thus propose that this method can be used to find ages of young OB associations through the Galaxy and also in nearby extra-galactic sources.

The arm structure of the Milky Way remains somewhat of an unknown, with observational studies hindered by our location within the Galactic disc. In the work presented here we use smoothed particle hydrodynamics (SPH) and radiative transfer to create synthetic longitude-velocity observations. Our aim is to reverse-engineer a top down map of the Galaxy by comparing synthetic longitude-velocity maps to those observed. We set up a system of N -body particles to represent the disc and bulge, allowing for dynamic creation of spiral features. Interstellar gas, and the molecular content, is evolved alongside the stellar system. A 3D-radiative transfer code is then used to compare the models to observational data. The resulting models display arm features that are a good reproduction of many of the observed emission structures of the Milky Way. These arms however are dynamic and transient, allowing for a wide range of morphologies not possible with standard density wave theory. The best fitting models are a much better match than previous work using fixed potentials. They favour a 4-armed model with a pitch angle of approximately 20• , though with a pattern speed that decreases with increasing Galactic radius. Inner bars are lacking however, which appear required to fully reproduce the central molecular zone.

We present simulations of the gaseous and stellar material in several different galaxy mass models under the influence of different tidal fly-bys to assess the changes in their bar and spiral morphology. Five different mass models are chosen to represent the variety of rotation curves seen in nature. We find a multitude of different spiral and bar structures can be created, with their properties dependent on the strength of the interaction. We calculate pattern speeds, spiral wind-up rates, bar lengths, and angular momentum exchange to quantify the changes in disc morphology in each scenario. The wind-up rates of the tidal spirals follow the 2:1 resonance very closely for the flat and dark matter dominated rotation curves, whereas the more baryon dominated curves tend to wind-up faster, influenced by their inner bars. Clear spurs are seen in most of the tidal spirals, most noticeable in the flat rotation curve models. Bars formed both in isolation and interactions agree well with those seen in real galaxies, with a mixture of "fast" and "slow" rotators. We find no strong correlation between bar length or pattern speed and the interaction strength. Bar formation is, however, accelerated/induced in four out of five of our models. We close by briefly comparing the morphology of our models to real galaxies, easily finding analogues for nearly all simulations presenter here, showing passages of small companions can easily reproduce an ensemble of observed morphologies.