Say I start with 52100 bearing ball about two inches in diameter. I draw this out into a knife blank and then anneal and grind the bevel. Starting out with one per cent carbon, the outer portion of the blank should have been decarburized, but by how much? And how deep? The goal here will be having the cutting edge about the eutectoid point with the outer layers somewhat less.
Now, I think that this can be hit or miss to control the decarburization but does anyone have any experience and comments? Yes, I know that I should start out with the right steel but that wouldn't be nearly as much fun.

Warren

A lot will depend on your heat source that you will be using and how high a temp is used.
If you have control of your heat, and say you are using a propane forge, you can create a carburizing flame and experience little loss in carbon.
If the blade you forge is left with sacrificial steel to grind off to finish the blade, you should/could have a final blade shape with full carbon content.
And, by the way, 52100 is the "right" steel.

I don't think that 52100 is a "wrong" steel but my preference would be a little less carbon and some vanadium to control grain growth. I will be using a coal forge. Still it should be fun to make a knife or tool from a bearing ball.
Warren