The X-15-A-2 was rebuilt from ship #2 after a crash. In October 1967 Pete Knight flew this craft to Mach 6.7 as part of a flight envelope expansion project, with a dummy scramjet pod attached to the ventral fin stub. There was severe heating damage to the ventral fin stub from spike shock shed from the from dummy scramjet pod. The pod separated in flight. Inconel-X skins suffered about 7 times-normal heating rate due to shock wave impingement.
The first photo shows the vehicle about to be dropped from its mothership B-52. Note the white heat protection coating applied to vehicle, and the dummy scramjet pod attached to ventral fin stub. The white heat protection coating did not prove to be successful. This version of the X-15 had two big drop tanks of rocket propellants.
The second photo is actually 3 images combined to get 3 views of damaged ventral fin stub. It shows clearly the damage to the ventral fin, obviously of a meltback nature. The material was simply overheated by shockwave impingement. These skins were Inconel-X, with a listed max useful temperature about 1500 F, and with meltpoint well beyond 2500 F but far less than 3000 F.
The third image is another view of the damaged ventral fin, with damage to the aft fuselage also visible. The separated scramjet pod was also recovered, and is shown.
What this proves is that the “parallel nacelle” arrangement is a really bad idea when flight speeds exceed about Mach 5 or 6, varying with the toughness of the materials involved. If you let this situation go on long enough, the impinging shock wave will cut away all the adjacent structure that it strikes.
That is the fundamental reason there have been no successful hypersonic vehicles using the “parallel nacelle” arrangement. The SR-71 and similar aircraft simply did not fly fast enough to incur this kind of damage. The X-15 did, and any entry vehicle far exceeds the heating risks that the X-15 saw.
Schematically, this is illustrated in the sketch. The areas at risk are subject to direct impingement, meaning the shock wave is “headed directly toward” the surface at risk, as if that surface is “in the way” of the wave. For waves that propagate 90 degrees away from the “headed toward” direction, the effects are very much less. That is why fins do not damage the surface from which they extend. There can be a little local heating, but nothing at all like direct wave impingement.
As to when the damage occurs, that depends upon the material and upon the flight speed. Even the turbine blade superalloys fail by melting under 3000 F, and fail structurally at temperatures not much over 2000 F. For a steady state exposure, the energy balance is convective input to the material, balanced by internal convective cooling, or by external re-radiation of heat to environment, or by a combination of both. If neither cooling occurs, the material soaks out to the recovery temperature.
The convective input is driven by the difference between the recovery temperature and the material temperature. The re-radiation ability is driven by the infrared (and optical) emissivity of the material and its temperature. In the case of the X-15 flight, there was no internal cooling, and the white coating had drastically reduced the material’s emissivity (the normal surface finish was nearly black).
At Mach 6.7 and 100,000 feet on a 1962 standard day, the turbulent recovery temperature that drives convective input calculates as about 3250 F, well beyond the melting point of any superalloy, much less its maximum use temperature. With re-radiative cooling suppressed by the white coating (and no internal cooling), it should not be surprising in retrospect that the test nacelle bow shock did extensive damage to the fin stub and aft fuselage.
The only real surprise is that this X-15 survived that particular flight at all. The exposure was brief enough that these effects did not have the time to sever its tail section.