Context : Massive stars influence their environment via stellar winds , ionising radiation and supernova explosions .
This is signified by observed interstellar bubbles .
Such “ feedback ” is an important factor for galaxy evolution theory and galactic wind models .
The efficiency of the energy injection into the interstellar medium via bubbles and superbubbles is uncertain , and is usually treated as a free parameter for galaxy scale effects .
In particular , since many stars are born in groups it is interesting to study the dependence of the effective energy injection on the concentration of the stars .

Aims : We aim to reproduce observations of superbubbles , their relation to the energy injection of the parent stars and to understand their effective energy input into the interstellar medium ( ISM ) , as a function of the spatial configuration of the group of parent stars .

Methods : We study the evolution of isolated and merging interstellar bubbles of three stars ( 25 , 32 and 60 M _ { \odot } ) in a homogeneous background medium with a density of 10 m _ { \mathrm { p } } cm ^ { -3 } via 3D-hydrodynamic simulations with standard ISM thermodynamics ( optically thin radiative cooling and photo-electric heating ) and time dependent energy and mass input according to stellar evolutionary tracks .
We vary the position of the three stars relative to each other to compare the energy response for cases of isolated , merging and initially cospatial bubbles .

Results : Due to mainly the Vishniac instability , our simulated bubbles develop thick shells and filamentary internal structures in column density .
The shell widths reach tens of per cent of the outer bubble radius , which compares favourably to observations .
More energy is retained in the ISM for more closely packed groups , by up to a factor of three and typically a factor of two for intermediate times after the first supernova .
Once the superbubble is established , different positions of the contained stars make only a minor difference to the energy tracks .
For our case of three massive stars , the energy deposition varies only very little for distances up to about 30 pc between the stars .
Energy injected by supernovae is entirely dissipated in a superbubble on a timescale of about 1 Myr , which increases slightly with the superbubble size at the time of the explosion .

Conclusions : The Vishniac instability may be responsible for the broadening of the shells of interstellar bubbles .
Massive star winds are significant energetically due to their – in the long run – more efficient , steady energy injection and because they evacuate the space around the massive stars .
For larger scale simulations , the feedback effect of close groups of stars or clusters may be subsumed into one effective energy input with insignificant loss of energy accuracy .