Take an initially isolated virialized self-gravitating stellar population with kinetic energy and Newtonian gravitational potential energy . Instantaneous spatially uniform removal of mass, without kicks to surviving material, leaves fraction with and . Hence : the homologous whole-system energy is negative only for . Positive whole-system energy does not forbid a bound core after selective escape, and slow mass loss follows the adiabatic mass-loss expansion law instead.
First specify the dynamical estimate: assume an initially isolated virialized system, stars of different masses well mixed with the same velocity distribution, and ejecta escaping rapidly compared with a stellar crossing time, without imparting kicks to surviving stars. If is the retained stellar mass fraction, the original virial theorem gives , where is kinetic energy and is Newtonian gravitational potential energy. Immediately after loss, positions and velocities are unchanged, so
Thus impulsive disruption by stellar mass loss makes the energy of the whole remaining system nonnegative when at least half the initial mass is removed.
For the power-law initial mass function, the fraction removed is a mass fraction, not a number fraction:
For integer slopes, , , , and . Interpolating between and puts the threshold around --. Solving more accurately gives
The rearranged equation has a spurious root at introduced when the vanishing denominator was cleared; the logarithmic limit excludes it. Increasing shifts the mass distribution towards lower masses, so the top-heavy side, , has nonnegative remnant energy in the impulsive model; equality is marginal. One can see strict monotonicity from the covariance obtained by differentiating : under the normalized mass-weighted distribution.
The initial mass function alone does not guarantee complete disruption. If losses occur slowly, the adiabatic mass-loss expansion law gives expansion rather than the instantaneous half-mass threshold, and positive total energy can still leave a bound core after selective escape. The stated value is therefore the conventional global impulsive estimate, with no external confining Newtonian gravitational potential.
Direct evidence for stellar feedback includes expanding ionized shells and superbubbles around young associations, hot X-ray-emitting gas in supernova remnants, broad or split emission lines, and blueshifted absorption showing cool and warm outflows. P-Cygni profiles reveal massive-star winds, while extraplanar filaments and metal-enriched gas demonstrate transport away from star-forming disks. Indirect evidence includes the galaxy mass--metallicity relation, low baryon fractions and suppressed star formation in dwarf galaxies, chemically enriched circumgalactic gas, and correlations of outflow speed and mass loading with the star formation rate.
Rapid gas removal changes the gravitational potential before stellar and dark-matter orbits can respond adiabatically. Positions and velocities are initially unchanged, but orbital binding energies rise; orbits expand and become more eccentric, and some particles escape. Repeated burst--outflow--reaccretion cycles can irreversibly transfer energy to collisionless matter and turn a central dark-matter cusp into a core. This matters because dwarf-galaxy rotation curves are used to test dark-matter microphysics: a feedback-made core can mimic a non-cold or self-interacting dark-matter signature.
Write the initial potential energy as . The virial theorem gives . If a well-mixed fraction remains after instantaneous mass loss, the immediate kinetic and potential energies are and , so
After revirialization at , . Equating energies gives the impulsive mass-loss expansion law
The remnant is bound only for ; loss of half or more of the gravitating mass disrupts this idealized system.
For many infinitesimal, individually revirialized losses, put in the impulsive result. To first order, . Integration yields the adiabatic mass-loss expansion law
Slow loss causes finite expansion for every positive remaining mass fraction and has no sharp disruption threshold.
Finally consider an initially circular orbit of radius around a point mass . Its speed and specific angular momentum obey and . Immediately after , these remain unchanged, while the new specific energy is
Using the orbital eccentricity relation gives
It is an ellipse for , parabolic at , and unbound for smaller , in agreement with the virial argument.