Stellar winds and photoionization in a spiral arm

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 pc x 500 pc x 100 pc section of a spiral arm from a galaxy simulation. We use smoothed particle hydrodynamics to re-simulate the region at higher resolution (1 M-circle dot 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 interstellar medium 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 similar to 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 (similar to 10(2) M-circle dot) and producing fewer high-mass sinks (similar to 10(3) M-circle dot). Overall, stellar winds play at best a secondary role compared to photoionization, and on many measures, they have a negligible impact.