Basic normalization questions

[joe pierre]

Just want to get some things straight about at what stage in the process it's best to normalize. For the record, I'm working with Aldo's 1075.

 

I've read that one should normalize shortly after the forging is complete in order to relieve stresses as soon as possible. But does this mean right after? If so, then doesn't simply cooling at room temperature from forging temp constitute the first thermal cycling (I'm guessing not because forging temp is too high for a normalization cycle).

 

Or do you first have to remove heavy scale prior to normalization? If you do that -- cool the blade and then grind -- then why not go ahead and get all the serious grinding out of the way and normalize at a later stage, prior to heat treating? And if you can do that, then can you just quench on the 4th cycle? Here I'm guessing that you'd have to remove scale again before doing the final HT -- that's how I've been doing it so far.

 

Finally, for your standard 3-cycle normalization, how important is trying to reduce the temperature each time? If the 2nd or 3rd cycle runs hotter than the first, have you basically rendered the first cycle useless? If it's really important to use descending temps during each cycle, what temps are recommended for 1075? Seems like controlling temp in that narrow range (above critical, but not too hot, and the decreasing for two more cycles) would be quite a challenge unless you have very fine control of your forge.

Edited March 23, 2013 by joe pierre

[Freya W. Ward]

Personally, my heat treatment method for all simple steels is to anneal after forging, I bring it up to critical, then bury it in vermiculite and let it stay for roughly 6 hours. Then once I've done all my grinding, I normalize only once, then it's straight to the quench. I usually use preheated canola oil. Doing this I almost never see any warping on larger blades, and they come out of the quench glass hard.

 

If I was to triple normalize however, yes I would do it right after forging to make sure I have no sections that are harder due to rapid cooling. Then once that's done descale to begin grinding.

Now about the temperature, I'm not sure exactly what other people do, but it's my understanding that you want the steel to form austenite in order to quench/normalize, and it won't do it at temperatures below critical. But I could certainly be wrong.

[Alan Longmire]

It's always a good idea to do at least one or two normalizations immediately after you finish forging, and doesn't take much time since the forge is still hot. I consider that a stress relief cycle. Do it on the as-forged piece, don't worry about scale. Just letting it cool after the last round of forging is not the same, as it leaves large/uneven grain and unequal stresses. The whole idea behind normalizing is to refine and equalize the grain, as well as to evenly distribute carbides in a hypereutectic steel.

 

I only do the vermiculite anneal thing on 5160, plain carbon steel just doesn't seem to need it unless you're engraving it.

 

After grinding and hand sanding (if any) prior to hardening, that's when the full normalization takes place. That's also your last chance for any hot-stamping of makers' marks, etc. The reason for descending heats is to keep the grain small. This is not as big a deal on alloys with vanadium and chromium, since those elements act to keep the grain small at elevated temperature anyway. With Aldo's 1075 on the other hand you have as close to a pure iron/carbon alloy as is possible unless you make it yourself. The only thing controlling grain size here is heat.

 

So: Descending heat normalization in a non-temperature-controlled forge... This is where your eye and darkness come together. I do this in a closed-end pipe in a coal forge at home, you can do it in a pipe or in an open gas forge as well provided you are not in bright light. The darker the better. Two things control grain size in simple steel, time and temperature. You can hold it at 1325 for ten minutes and it will transform. You can take it to 1450 for two minutes and it will transform. The longer you hold it at the transformation point or above the more the grain will tend to grow.

 

You know I am always going on about watching for decalescence and recalescence in the steel. It's that bright line - swirling shadows thing you see during transformation. You are going to use this to judge when to pull the blade out of the forge. On the first normalizing heat, bring it up slowly (as in it should take maybe five minutes) until you see the shadows start to swirl inside the steel. Hold at that temperature until the shadows are completely gone, then pull the blade and let cool to black. It does not need to get all the way to room temperature, anywhere below around 700 degrees F is fine. Do not adjust your forge in the meantime. When the blade is cooled enough, put it back in the forge. This time, when you see the shadows form just let it sit until almost all the swirling is gone, then pull it out and let it cool. On the final cycle, pull it out the second you see the shadows.

 

Bingo: descending-heat normalization, and you don't even need to know what the temperature was because that's not important. You were controlling the amount and time of the crystalline phase change by watching it happen. This is heavy quantum phenomena here ( electrons jumping energy levels, atoms rearranging themselves within crystalline lattices, photons emitted when the energy levels change, and so on, but don't worry about that ) but anyone can do it as long as it's not too brightly lit where you do it.

 

Final note: Since you're using that particular steel I am assuming you're going for a hamon. It's best to do the normalizing before applying the clay if you are using clay. The other groovy thing about multiple normalizations is that you can refine the grain size so much that you will reduce hardenability to the point where you get an automatic hamon. It won't be as wide, or as active, but it can happen. With 1095 and the W series steels, you can even refine the grain so much it won't harden at all because you can't quench it fast enough to miss the nose of the curve. In the unlikely event that happens to you, just reheat to critical (watch the shadows, not a magnet!) and hold as best you can for a few minutes to regrow the grain size.

 

If you are doing this with a sword-sized blade in a charcoal trench forge, use your best guess. If it's dark enough you can still see the transformation, but it's harder to do.

[joe pierre]

holy cow Alan, that's a great reply. this one is going in the notebook.

 

this forum is so awesome -- ask a question and get a reply in a few hours from guys who really know their craft. amazing how much you can learn and no way we could have done anything like this when we were young and there was no internet.

 

anyway, i went ahead and did the normalization late last night. i'm going to have to work on conjuring up those decalescence shadows, especially when just striving for an even heat is a bit of a challenge, but at least i have something to aspire to.

[Doug Lester]

Alan hit it right on the head. Another thing that I'm going to try after I get my last blades finished up is to use Tempel Stiks (temperature sensitive markers) to judge the heat by.

 

Also, you don't have to worry about cleaning off the fire scale; it shouldn't be that heavy. If it does worry you, cleaning it off with a wire brush should be fine.

 

The slow cooling in vermiculite or allowing the steel to cool off with the forge is more of an anneal. With that you can have a problem with carbides clumping and making things like drilling more of a problem.

 

Doug

[Alan Longmire]

Ah, yes, even heat... for that you sometimes just have to stroke the blade through the hot spot to keep things even. It's annoying, but Randal Graham used to heat treat katanas by stroking them through a ten-inch vertical forge. Be patient and careful and it's amazing what you can do.

[Thom Board]

I'd just like to echo Joe's thanks for your reply Alan. Any chance this could get pinned?