Thread: Does steel have a correctable memory?

Originally Posted by Mike Blue

Martensite formation occurs at the speed of sound (770 miles per hour at 68F at sea level, about 13548 inches per second).

So is that "crack" sound I've sometimes heard when quenching a blade is my martensite going Mach 1? Good info, Mike.

Originally Posted by Mike Blue

From what you describe, the vapor jackets are very transient and the bubbles would break down and be replaced by the liquid quenchant as fast as they form. That amount of time is not as fast as the formation of martensite. The more likely conditions that could be affected by martensite formation are the changes occuring in the grain structure as a result of the transition from austenite to martensite or pearlite. Those could easily be affected within the amount of time available and could contribute to warping. Another likely condition is that the steel is still in the metastable phase and has not completed conversion to martensite and is "more plastic" than hard yet. In that state banging off the inside of the tank or possibly even the resistance to motion afforded by a thick cold oil could bend the blade against the resistance of motion provided by your hand. It wouldn't be the motion itself, or the vapor jacket, but specific conditions in the steel though. I admit, it's a lot easier to think that it's the vapor jacket.

The speed of the quench is important in establishing the desired crystalline structure in the blade. It is the crystal's size caused by heat treatment or mechanical deformation that causes the internal stresses, not because the edge hardens quicker. For that matter, the sword blade is under a tension stress because of the size reduction of the blade's spine to pearlite from the larger austenitic crystal.

The metallurgy texts I've read advocate agitating the steel rapidly in the oil to promote rapid cooling. Does agitating the steel break up the vapor barrier, or just bring unheated oil into contact with the blade, causing more rapid cooling?

Josh

I don't think it "breaks the sound barrier" but it sure is a dreaded sound.

From what I've read about quenching, the industrial setups all use some form of agitated quenchant. That problem has been solved by a couple folks I know with the use of a simple compressor, like those used in aquariums and the end of the hose held down in the bottom of the tank to more complicated variations on that theme. Others have used a pipe and pump setup where the quenchant flows past the object quenched.

Even if you held the blade rigidly still, the expanding bubbles would collapse bringing cooler quenchant into contact with the blade and repeat. As the quenchant heats up, a convection current will develop and carry the hotter quenchant vertically up the tank and cooler quenchant will rise into it's place.

In any case, if the blade is moving to break up the vapor jacket, or the quenchant is moving to break up the jacket, the speed of cooling is faster. The whole point is to get the steel below the nose of the curve and promote the maximum conversion of martensite to harden it.

If the quench is "too fast" , or you can't get the blade out quick enough, the dreaded ping maybe one of the results. All the other factors that go into setting the steel up for maximum performance still apply. The quench is only one of the items to pay attention to.

Originally Posted by Mike Blue

Realistically, if the edge was flat and the blade honed properly, what difference, other than the aesthetic, would there be as far as the blade's ability to shave well? Now I suspect that razor manufacturers didn't much care that the spine had a warp as long as the edge was salvageable enough to hone/shave and sell.

For me, the ideal razor is one that shaves well and hones easily. If such a razor can have an edge that lays perfectly flat on a lapped hone surface on both sides of the blade but somehow the spine is "off" or aesthetically asymmetrical , I would care very little. I don't mind having a vintage blade that insists on being a "project" razor by testing me to try many different hones, widths, stroke variations each and every time I'd have to hone then touch it up, but if I had to resort to such tactics with the majority of the razors in my rotation, I would not get the same enjoyment I do now. Upon retirement when I have all the time in the world, I could see my perspective being entirely different. I'll have to get back to you in 28-30 years on that one though.

Chris L

Originally Posted by Mike Blue

A few comments:



Ahem. Stress relieving occurs in two types. The first is to relieve internal stresses generated by heavy machining of clamped pieces. The second is tempering which relieves the internal stresses in steel caused by martensite formation against areas of non-martensite. I could use your Japanese sword example because that is a perfect illustration. So, the fact is, not personal opinion, stress relief is desirable when one of the above conditions exists in the material.

Thanks for proving my point. The opion was about normalization and spheroidization. Sorry will try to go slower next time.

Martensite formation occurs at the speed of sound (770 miles per hour at 68F at sea level, about 13548 inches per second). Even in a great big bowie knife (1/4-5/16ths of an inch thick at the spine), I don't think that any wiggling about a human being could do will cause a warp to occur in the short amount of time available between the initial quench and the formation of martensite, unless some very specific conditions are controlled. About the only way to significantly alter quenching both sides of a blade to cause one side to cool faster than the other, would be to lay only one side of the blade on the surface of the quench bath.

And how did those brainiacs come up with that number? How can you possibly time that? Are they taking photographs of the change and time it that way? I would hate to be the guy with the stop watch trying to get that one right.

From what you describe, the vapor jackets are very transient and the bubbles would break down and be replaced by the liquid quenchant as fast as they form. That amount of time is not as fast as the formation of martensite. The more likely conditions that could be affected by martensite formation are the changes occuring in the grain structure as a result of the transition from austenite to martensite or pearlite. Those could easily be affected within the amount of time available and could contribute to warping. Another likely condition is that the steel is still in the metastable phase and has not completed conversion to martensite and is "more plastic" than hard yet. In that state banging off the inside of the tank or possibly even the resistance to motion afforded by a thick cold oil could bend the blade against the resistance of motion provided by your hand. It wouldn't be the motion itself, or the vapor jacket, but specific conditions in the steel though. I admit, it's a lot easier to think that it's the vapor jacket.

Actually it's called a vapor barrier because it causes a barrier. That is why boiler tubes will destroy themeselves from improper cooling if a barrier builds up between the tube wall and the cooling water. The barrier is self sustaining and the only thing the operators can do is correct there actions that caused it. To much heat, to much flow, not enough flow, wrong chemistry, etc...

You were okay up to the point of whacking it with a wooden mallet. The copper block treatment is to relieve the sori, the formal curve that makes it a Japanese sword. That technique is not used to straighten a warped blade. I've seen 1/2 steel plate straightened with a rosebud torch and a little hammer. Good smiths have good techniques, but not all of them work in all conditions.

So how does a wooden mallet make it wrong? Oh because you didn't say it.

Do I sense some abrasion here? (and I mean the bad kind, not the kind that makes things shiny)

I think eventually if enough abrasive slips in here, we might just get to the pearls.

Just so there isn't any confusion, here is your original statement.

Originally Posted by jimmyseymour

I saw this done in my bladesmithing class for the ABS. Heating the spine while keeping the blade in the water to straighten out a warped blade. Michael Connor's from winters was teaching the class. I didn't get to see the whole process but no pressure was placed on the blade. It was gently heated and it straightened itself. The guy who did it said he could feel it straighten.

A blade could also warp because it was not stress relieved properly, or at all before harding. If using an oil quench if you move the blade around to much you could also cause a warp.

Heating a spine to correct a warp, implies post-heat treatment. And I clearly specified the conditions that required stress relieving a piece. Post heat treatment is neither a normalized or spherodized state. But you did mention stress relief. There is nothing in your original post about heavy machining cold material that would require stress relief. Nothing in the original post mentioned stress relieving the blade before it was heat treated. Straightening a warped blade in this way is not tempering the blade either. So this doesn't qualify as a stress relieving cycle as well. Sadly, it does not serve to "prove your point."

And how did those brainiacs come up with that number? How can you possibly time that? Are they taking photographs of the change and time it that way? I would hate to be the guy with the stop watch trying to get that one right.

You need to get out and read some better metallurgical and heat treatment books.

Actually it's called a vapor barrier because it causes a barrier. That is why boiler tubes will destroy themeselves from improper cooling if a barrier builds up between the tube wall and the cooling water. The barrier is self sustaining and the only thing the operators can do is correct there actions that caused it. To much heat, to much flow, not enough flow, wrong chemistry, etc...

A knife blade is considerably smaller than a big industrial tube and does not hold within it, a trapped atmosphere.

I think the next time you quench a blade, if you watch carefully, the bubbles grow then pop then regrow and many of them will release from the surface and rise to the top of the quench tank in the convection current caused by the warming of the surrounding water/oil/whatever being warmed by the hot blade, pulling more cooling oil into the proximity of the blade. The fact that the convection current develops shows that the blade is releasing heat into the quenchant, despite the fact that a vapor barrier exists or not. If your theory about vapor barriers being an interference in quenching things was a little more solid, then we'd not be able to harden any kind of steel.

So how does a wooden mallet make it wrong? Oh because you didn't say it.

I didn't say it was wrong. This has nothing to do with whether it's me saying it or not. You misunderstand two entirely separate techniques and call them the same. That's where you got it wrong. I don't mix apples and oranges and try to tell everyone they are grapes. I don't get pissed off about being wrong. I learn from my mistakes and move on.

I'm trying to help you win but you won't see that for a long time I suspect.

Originally Posted by floppyshoes

Do I sense some abrasion here? (and I mean the bad kind, not the kind that makes things shiny)

Just a little. After the private message I got from Mike a year ago just about anything he say's about one of my post's really gets under my skin. So with that said I'm off this thread since we aren't going to see eye to eye simply because of personal differences, and it's distracting everyone.

Originally Posted by jimmyseymour

So with that said I'm off this thread since we aren't going to see eye to eye simply because of personal differences, and it's distracting everyone.

Thank you for offering to do this. No offense my me saying that, I just had hoped that this thread I started didn't continue down the path it clearly seems it had potential to do.

Chris L

I apologize publicly for any unintended, perceived offense that I've rendered, whether publicly or privately sourced.

I'll just as publicly correct any factual errors that I commit.