Quench cracks result from stresses produced during the transition from Austenite to Martensite, which involves an increase in volume.
The martensitic transformation starts at the outermost surfaces of the part being quenched. As the transformation goes deeper into the softer austenite towards center of mass, its change in volume is restricted by the martensite already created in the outer volumes of the part adjacent to the surface.
This creates internal stresses which place the surface into tension.
When enough martensite has formed to create internal stress greater than the ultimate strength (tensile strength) of the as quenched martensite at the surface, a crack results.
As-quenched Martensite is hard and brittle- it has virtually no ductility.
Here are 3 ways to recognize a quench crack:
1) The crack runs from the surface towards the center of mass in a fairly straight line. The crack will also tend to be open or spread at the OD surface.
2) Quench cracks do not have decarburization apparent, since the quenching occurs at relatively low temperatures. If there is decarb associated with a crack, that shows that the crack existed at the time the material was at temperatures hot enough to decarburize. In other words, the crack existed prior to austenitizing.
3) The fracture surfaces will exhibit a fine crystalline structure. I remember the first time I saw a quench crack, thinking, “it crystallized.” Well, the steel is already crystalline, but the fine martensitic structure revealed by the crack showed that there was absolutely no ductility in the material…
Bonus tip: if you see a build up of scale in the crack itself, that tells you that the crack was there after quenching but before tempering. During the tempering operation at tempering temperature, oxygen in the atmosphere created a scale where it could reach the iron in the crack.