Gravitational waves (GWs) propagate a _change_ in the gravitational field. Consequently, GWs are only shed by accelerating matter where the change in motion i

Gravitational waves (GWs) propagate a _change_ in the gravitational field. Consequently, GWs are only shed by accelerating matter where the change in motion is not rotationally or spherically symmetric. Spherically symmetric gravitational collapses (of dust or gas, for example) are certainly systems of accelerated matter, but do not shed gravitational waves. Rotating spherical bodies also do not shed GWs. To first order, this means the rotation of stars and planets do not shed GWs, although a planet revolving around the star in an essentially Kepler orbit will shed GWs, as will a pair of stars in close orbit around each other. GWs solve the apparent removal of momentum-energy from a system where there is no removal of mass-energy by non-gravitational radiation. For most systems the energy loss is negligible. That's why the first GW detection involved a late stage inspiral of two compact massive objects with no way to ditch their angular momentum electromagnetically or via neutrino (or other dark matter) emission. GWs themselves have spin 2 symmetry; in theories with gravitons, those are spin-2 gauge bosons, typically massless (because they are long range), and you get large numbers of them forming the classical GWs, just as you will find large numbers of photons forming a bright flicker of light, although you can detect tiny numbers of photons because electromagnetism is relatively strong -- tiny numbers of gravitons are not going to be detectable by humanity any time soon. It's not curvature (and certainly not gravitational forces) heating up accretion disks. The matter in the disks move on intersecting geodesics and occasionally collide. Such collisions tend to ditch the particles' momentum into emitted photons. Hawking radiation is an effect of accelerated observers seeing more (and more energetic) particles than non-accelerated ones, and gravitational collapses producing an acceleration between past and future observers. Where a past observer sees a tiny number of low-frequency particles in the "vacuum" outside the collapsing object, a future observer (after the horizon has formed) will see a larger number of higher frequency ones. The acceleration is curvature dependent, and curvature just outside the horizon depends on black hole mass -- more massive black holes have less curvature just outside the horizon (which is further away from the gravitational singularity or whatever is at the centre of a black hole). These particles near but outside the horizon can escape, which takes mass-energy away from the dynamical spacetime just outside the horizon. That in turn causes the horizon to retract and that exposes stronger curvature. That in turn produces a stronger acceleration for further future observers, and the process continues -- that's black hole evaporation. Hawking radiation is always emitted, for SMBHs that only grow by incidental interception of CMB radiation, and even for stellar mass black holes. Infalling matter simply replaces (and then some) the outgoing "fossil" infrared radiation from the space just outside the black hole as the horizon was forming.