We present simulations of a 500 pc2 region, containing gas of mass 4 × 106 M⊙, extracted from an entire spiral galaxy simulation, scaled up in resolution, including photoionizing feedback from stars of mass >18 M⊙. Our region is evolved for 10 Myr and shows clustered star formation along the arm generating ≈ 5000 cluster sink particles ≈ 5 per cent of which contain at least one of the ≈ 4000 stars of mass >18 M⊙. Photoionization has a noticeable effect on the gas in the region, producing ionized cavities and leading to dense features at the edge of the H ii regions. Compared to the no-feedback case, photoionization produces a larger total mass of clouds and clumps, with around twice as many such objects, which are individually smaller and more broken up. After this we see a rapid decrease in the total mass in clouds and the number of clouds. Unlike studies of isolated clouds, our simulations follow the long-range effects of ionization, with some already dense gas, becoming compressed from multiple sides by neighbouring H ii regions. This causes star formation that is both accelerated and partially displaced throughout the spiral arm with up to 30 per cent of our cluster sink particle mass forming at distances >5 pc from sites of sink formation in the absence of feedback. At later times, the star formation rate decreases to below that of the no-feedback case.

We have performed simulations of cluster formation along two regions of a spiral arm taken from a global Milky Way simulation, including photoionising feedback. One region is characterised by strongly converging flows, the other represents a more typical spiral arm region. We find that more massive clusters are able to form on shorter timescales for the region with strongly converging flows. Mergers between clusters are frequent in the case of the strongly converging flows and enable the formation of massive clusters. We compare equivalent clusters formed in simulations with and without ionisation. Photoionisation does not prevent massive cluster formation, but can be seen to limit the masses of the clusters. On average the mass is reduced by around 20 per cent, but we see a large spread from ionisation having minimal difference to leading to a 50 per cent reduction in mass. Photoionisation is also able to clear out the gas in the vicinity of the clusters on Myr timescales, which can produce clusters with larger radii that are surrounded by more massive stellar haloes. We find that the ionising feedback has more impact in our second region which is less dense and has less strongly converging flows.

We present Ekster, a new method for simulating star clusters from birth in a live galaxy simulation that combines the smoothed-particle hydrodynamics (SPH) method Phantom with the N-body method PeTar. With Ekster, it becomes possible to simulate individual stars in a simulation with only moderately high resolution for the gas, allowing us to study whole sections of a galaxy rather than be restricted to individual clouds. We use this method to simulate star and star cluster formation in spiral arms, investigating massive GMCs and spiral arm regions with lower mass clouds, from two galaxy models with different spiral potentials. After selecting these regions from pre-run galaxy simulations, we re-sample the particles to obtain a higher resolution. We then re-simulate these regions for 3 Myr to study where and how star clusters form. We analyse the early evolution of the embedded star clusters in these regions. We find that the massive GMC regions, which are more common with stronger spiral arms, form more massive clusters than the sections of spiral arms containing lower mass clouds. Clusters form both by accreting gas and by merging with other proto-clusters, the latter happening more frequently in the denser GMC regions.

We present hydrodynamic and magnetohydrodynamic (MHD) simulations of sub galactic regions including photoionising and supernova feedack. We aim to improve the initial conditions of our region extraction models by including an initial population of stars. We also investigate the reliability of extracting regions in simulations, and show that with a good choice of region, results are comparable with using a larger region for the duration of our simulations. Simulations of star formation on molecular cloud scales typically start with a turbulent cloud of gas, from which stars form and then undergo feedback. In reality, a typical cloud or region within a galaxy may already include, or reside near some population of stars containing massive stars undergoing feedback. We find the main role of a prior population is triggering star formation, and contributing to gas dynamics. Early time supernova from the initial population are important in triggering new star formation and driving gas motions on larger scales above 100 pc, whilst the ionising feedback contribution from the initial population has less impact, since many members of the initial population have cleared out gas around them in the prior model. In terms of overall star formation rates though, the initial population has a relatively small effect, and the feedback does not for example suppress subsequent star formation. We find that MHD has a relatively larger impact than initial conditions, reducing the star formation rate by a factor of 3 at later times.

We present simulations of the formation and evolution of clusters in spiral arms. The simulations follow two different spiral arm regions, and the total gas mass is varied to produce a range of different mass clusters. We find that including photoionizing feedback produces the observed cluster mass radius relation, increasing the radii of clusters compared to without feedback. Supernovae have little impact on cluster properties. We find that in our high density, high gas mass simulations, star formation is less affected by feedback, as star formation occurs rapidly before feedback has much impact. In our lowest gas density simulation, the resulting clusters are completely different (e.g. the number of clusters and their masses) to the case with no feedback. The star formation rate is also significantly suppressed. The fraction of stars in clusters in this model decreases with time flattening at about 20 per cent. In our lowest gas simulation model, we see the formation of a star forming group with properties similar to an OB association, in particular similar to Orion Ia. We suggest that low densities, and stronger initial dynamics are conducive to forming associations rather than clusters. In all models cluster formation is complex with clusters merging and splitting. The most massive clusters which form have tended to undergo more mergers.

We explore the interplay between supernovae and the ionizing radiation of their progenitors in star forming regions. The relative contributions of these stellar feedback processes are not well understood, particularly on scales greater than a single star forming cloud. We focus predominantly on how they affect the interstellar medium. We re-simulate a 500 pc2 region from previous work that included photoionization and add supernovae. Over the course of 10 Myr more than 500 supernovae occur in the region. The supernovae remnants cool very quickly in the absence of earlier photoionization, but form much larger and more spherical hot bubbles when photoionization is present. Overall, the photoionization has a significantly greater effect on gas morphology and the sites of star formation. However, the two processes are comparable when looking at their effect on velocity dispersion. When combined, the two feedback processes increase the velocity dispersions by more than the sum of their parts, particularly on scales above 5 pc.

The role of different stellar feedback mechanisms in giant molecular clouds is not well understood. This is especially true for regions with many interacting clouds as would be found in a galactic spiral arm. In this paper, building on previous work by Bending et al., we extract a 500 × 500 × 100 pc section of a spiral arm from a galaxy simulation. We use smoothed particle hydrodynamics (SPH) to re-simulate the region at higher resolution (1 M⊙ per particle). We present a method for momentum-driven stellar winds from main sequence massive stars, and include this with photoionization, self-gravity, a galactic potential, and ISM heating/cooling. We also include cluster-sink particles with accretion radii of 0.78 pc to track star/cluster formation. The feedback methods are as robust as previous models on individual cloud scales (e.g. Dale et al.). We find that photoionization dominates the disruption of the spiral arm section, with stellar winds only producing small cavities (at most ∼ 30 pc). Stellar winds do not affect the resulting cloud statistics or the integrated star formation rate/efficiency, unlike ionization, which produces more stars, and more clouds of higher density and higher velocity dispersion compared to the control run without feedback. Winds do affect the sink properties, distributing star formation over more low-mass sinks (∼102 M⊙) and producing fewer high-mass sinks (∼103 M⊙). Overall, stellar winds play at best a secondary role compared to photoionization, and on many measures, they have a negligible impact.

It remains unclear how galactic environment affects star formation and stellar cluster properties. This is difficult to address in Milky Way-mass galaxy simulations because of limited resolution and less accurate feedback compared to cloud-scale models. We carry out zoom-in simulations to re-simulate 100–300pc regions of a Milky Way-like galaxy using smoothed particle hydrodynamics, including finer resolution (0.4M⊙ per particle), cluster-sink particles, ray-traced photoionization from O stars, H2/CO chemistry, and ISM heating/cooling. We select ∼106M⊙ cloud complexes from a galactic bar, inner spiral arm, outer arm, and inter-arm region (in order of galactocentric radius), retaining the original galactic potentials. The surface densities of star formation rate and neutral gas follow $\Sigma _\mathrm{SFR}\propto \Sigma _\mathrm{gas}^{1.3}$, with the bar lying higher up the relation than the other regions. However, the inter-arm region forms stars 2–3x less efficiently than the arm models at the same Σgas. The bar produces the most massive cluster, the inner arm the second, and the inter-arm the third. Almost all clusters in the bar and inner arm are small (radii <5pc), while 30-50 per cent of clusters in the outer arm and inter-arm have larger radii more like associations. Bar and inner arm clusters rotate at least twice as fast, on average, than clusters in the outer arm and inter-arm regions. The degree of spatial clustering also decreases from bar to inter-arm. Our results indicate that young massive clusters, potentially progenitors of globular clusters, may preferentially form near the bar/inner arm compared to outer arm/inter-arm regions.