I think you are referring to casehardening low carbon mild steel using the pack method.
The piece is packed in a carbonaceous material - bone black, charred material - in a container. The container is sealed with fireclay, etc. The sealed container is then heated in a forge or furnace to a bright red, and held at that temperature. Carbon will migrate into the surface of the steel while the container is in the forge. If the container is allowed to cool slowly, the piece will not be hardened, although carbon content in the surface will be dramatically increased; if it subsequntly heated and quenched, the surface will be hard. If, while the container is glowing, it is removed from the furnace, broken open and the part dumped into water, the part will be surface hard. Variations on this method can be used to produce coloured casehardening. Pack casehardening can result in a thicker hard layer than Kasenite can produce.
If the surface is too hard, the flint can skate over it without producing good sparks. If the hardness is drawn back slightly, as Mooncoon has suggested, better sparks may result.
On occasion, casehardening a high carbon steel casting can be worthwhile, because the surface of the casting can become decarbonized. The piece must be tempered though, because otherwise it will be brittle hard all the way through.
It is possible to selectively draw the temper in a frizzen, so that the striking face is hard, but the pivot lug is not brittle. This reduces the risk of breakage.
