Triple quenching?

[Einar]

I have seen that some smiths quench their blades up to three times, aparently to reduce the grain size, If I understood it correctly. But isnt this also achieved when normalising multiple times before a single quench? Is there a difference between normalising three times and then quenching once, or quenching multiple times?

[Einar]

Sorry, I tried the search function on the forum and came up with nothing, so I made this thread, but now I tried google, and it found multiple threads on the subject on this forum. I'm off to do some reading

[justin carnecchia]

If I'm not mistaken, and I certainly may be, triple quenching is used for mid-alloyed steels like 5160 or 52100 to try to bring carbon and the other carbides into solution for quenching. Steels are typically soaked at a given time and temperature prior to quenching in order to fully dissolve the carbides into solution. This requires precise control of temperature and the typical forge that a bladesmith uses won't work. So basically triple quenching is an attempt to get everything into solution using "backyard" methods, like a torch.

I don't work with either 5160 or 52100, but I have read of reports that triple quenching increases performance. I don't believe there is any advantage to doing it with 10 series or W series steels.

[Josh Brannen]

Try puting this is the search bar-> site:"bladesmithsforum.com" your search here

 

This is something I saw Dave Stephens point out a while ago and works awesome. There is a ton of info on here about that if you can find it. The built-in search engine isn't so great.

[Jerrod Miller]

As Einar found out this has indeed been covered, but since this thread will show up in future searches (with Google as Josh suggests, the forum search engine is indeed lacking) here is a quick summary:

 

Thermal cycling refines grain structure. The more rapid the thermal change (cooling especially) the more drastic the refining. The refinement comes from thermally induced stresses. Too much stress can lead to cracking. Most find that normalizing several times (3-5 is "generally agreed upon") is sufficient. Carbon gets tied up in carbides, both iron carbide and "other metal" carbide, such as chrome carbide, vanadium carbide, etc. Time and temperature is what dissolves the carbides. The hotter you get, the faster they dissolve. The hotter you get the more grain growth you get too, thus requiring grain refinement. Rapidly cooling (anything but an ultra slow anneal) will help in dissolved carbon forming smaller carbides rather than larger carbides. Smaller is more desirable.

 

General guide: Normalize from a "pretty hot temperature" to back. The hot temperature will dissolve your carbides and the normalize (rather than anneal) will help form smaller carbides. Then normalize at least a couple more times from a lower temperature. Your carbides will already be "smaller" from the previous normalize, this will continue to refine them a little, but really help in general grain refinement. Then do your quench cycle.

[Einar]

Thank you for the feedback, guys.