Tempered Martensite Embrittlement

Rick Marchand

Rick Marchand

Donkey on the Edge
Moderator
Joined
Jan 6, 2005
Messages
9,680
I figured this topic needed it's own thread. It is mentioned around the forums but I have yet to see a conclusive discussion.

Hammerfall recently brought up tempering embrittlement in another thread and it really got me interested in the subject.

Some folks will complete their oven tempering cycles, then take it further by drawing the spine back even more with a torch. I have seen makers take it way past the blue spring temper to dull grey and beyond. That has to be hitting the Tempered Martensite Embrittlement realm. I can't help but think it is working against what they hope to accomplish.

Take a peek at this chart taken from the above link...

ih1006-htdr-fig1_lg.jpg


I have brought steel well into that first valley and haven't noticed any embrittlement in my own testing. The hope I have is that because I don't hear about this phenomenon very often, it really has little bearing on the knifemaking world.

What are your thoughts?
 
I may be a bit slow here, but how is it that TME, occuring in the 480°F–750°F, is supposedly irreversible, yet the chart shows toughness going up in the 705°F –1070°F range?
 
... Just stay out of the valleys, especially in cases and places where impact toughness is a concern.

It's kind of hard to wrap a brain around, but in many cases it may just be a trade off between different types of "embrittlement"...
 
Carl_First_Timer said:
I may be a bit slow here, but how is it that TME, occuring in the 480°F–750°F, is supposedly irreversible, yet the chart shows toughness going up in the 705°F –1070°F range?
Click to expand...
How I understand it, where and how you stop heating very important.(slow cool or quench) It seems that you can heat past the first valley and onto the second climb as long as you rapidly quench and do not allow it to precipitate. The first bell is really where all that martensite from your quench gets manipulated and drawn out. The second uphill is no longer dealing in hardness value... only impact strength. There is much more going on than martensite conversion and releasing trapped carbon during tempering.

At this point, I am still trying to understand it and grasping at straws. I don't even know if I should be concerned.
 
I'm only concerned, albeit very little, because of the camp sword tempered spine I'm working on.

It does state this (which makes me not so worried):
"It is important to understand that the degree of embrittlement is affected by the prior austenite grain size and hardness. So, if we are dealing with a fine-grained plain-carbon steel of low hardness, it may not experience embrittlement symptoms despite its phosphorous content whereas a more highly alloyed Cr-Ni steel used at higher hardness is more susceptible to impurity content."

And this:
"The susceptibility of a given steel to temper embrittlement depends on a number of factors including grain size, hardness, steel grade and the impurity control in the steelmaking process itself. Not all steels and not all steelmaking processes are equal."

Since most of us here try to purchase high quality steels recommended, and accepted by the knife making community to be the best for making knives, maybe we do not need to worry about this as much. Also maybe this is why certain steels are recommended and advised.
 
i would like to say do not try your luck. if your knife is tempered in its steel's embrittlement zone, the brittleness will stay there, doesn't matter if you feel it or not. and that brittleness will wait until someday you really depends your life on that blade, then it will snap and shatter leaving a big chipp off in your a$$:p.

different steel has slight or complete different embrittlement valley. the best way to avoid it is to ask the impact toughness vs tempering temperature graph from the manufacture. or check the stuff like this http://www.crucible.com/eselector/prodbyapp/plastics/crus7p.html. tough i really do not like how crucible put the test result. they never show the size of specimens nor temperature of the impact test. but something is better than nothing. and from crucible's table you can have a general outline of how to avoid the embrittlement valley.
 
If I'm reading it correctly, are they are showing impact toughness as interchangeable with ductility? I.e. more bendy/springy the higher the impact toughness, inversely showing lower Rc harness? I am thinking only in tems of simple high carbon alloys like 10xx since I have no experience with more complex alloys.


-Xander
 
Might be overthinking and/or under-testing this. Make a blade and PUNISH IT SEVERELY!.........then if it is not already broken.........BREAK IT! Having done this with through hardened 5160......as well as O-1, 1095, 1084, and various saw blade steel, RTS(rusty truck springs)etc. I am confident that a properly through hardened and tempered, and drawn back spined 5160 blade(with correct edge geometry) will survive anything short of a violent train wreck........and will still be alive after I am worn to a frazzle. Tempering embrittlement not withstanding:D
That's my .02.........might not be worth even that much:p
Darcy:)
 
DEllis said:
Might be overthinking and/or under-testing this. Make a blade and PUNISH IT SEVERELY!.........then if it is not already broken.........BREAK IT! Having done this with through hardened 5160......as well as O-1, 1095, 1084, and various saw blade steel, RTS(rusty truck springs)etc. I am confident that a properly through hardened and tempered, and drawn back spined 5160 blade(with correct edge geometry) will survive anything short of a violent train wreck........and will still be alive after I am worn to a frazzle. Tempering embrittlement not withstanding:D
That's my .02.........might not be worth even that much:p
Darcy:)
Click to expand...

+1 I've tested a fair amount of blades/steel to destruction made from the steels Darcy listed with similar results. I have no experience drawing high tempers with higher alloy steels however.
 
Darcy, what are you drawing your spines to? Colour? I have a feeling it is not as critical either or I would have noticed something in my testing as well
 
Rick Marchand said:
Darcy, what are you drawing your spines to? Colour? I have a feeling it is not as critical either or I would have noticed something in my testing as well
Click to expand...

Rick, I draw them to blue........usually twice.
Darcy:)
 
So you are still just rounding the peak of the first curve(500F-540F). I am talking about getting into flat grey and beyond (700F +) I wouldn't expect to notice TME at those temperatures.
 
Last edited:
Rick, I have read and been told that you can temper some of the high alloy steels, stainless included, either above or below that embrittlement range. Some say that the blades turn out better if you use the low temps, but with the caveat that you MUST use a more complex heat treat regimen, including cryo, in order to get all of the benefit. I have heard at least one very knowledgeable guy say that if you don't have cryo, don't even think about tempering at the low temperatures. As for our favorite common plain carbon steels, tempering at those high temps works great too.........if you are trying to get the steel soft again so that you can grind it.:D
 
This whole subject is very complex.Tai Goo's 'different types of embrittlement ' is only part of the problem. Confusion by symantics, purity, steel making methods ,steel types, heating to color and calling it a specific temperature [ bad practice] ,the type of impact test , etc , etc and so forth !! Do your own tests on one type of steel and pick the best compromise.
I'll stay out of this one.
 
mete said:
This whole subject is very complex.Tai Goo's 'different types of embrittlement ' is only part of the problem. Confusion by symantics, purity, steel making methods ,steel types, heating to color and calling it a specific temperature [ bad practice] ,the type of impact test , etc , etc and so forth !! Do your own tests on one type of steel and pick the best compromise.
I'll stay out of this one.
Click to expand...

I agree,... only part of the problem, very complex, can't be resolved with just a few paragraphs or without getting our heads twisted into knots.

I'm out.
 
I've been questioning the usefulness of drawing back spines ever since I saw Kevin Cashen demonstrate the flexibility of a fully hardened razor blade.. Rockwell way over 60... bent into a circle and not break. Obviously different geometries and thickness... but at the very least it showed that fully hard doesn't mean brittle. My spines are usually softer anyway since I typically use clay during the heat treat.. but even if I fully hardened... I guess I'm no longer sold on the fact that I need to do any kind of differential tempering. My own testing seems to support this idea.. but should I reconsider this attitude? My steels are usually Aldo 1075, W2, 1084 and scrap-o-gane.
 
i brought this up doesn't mean to confuse anyone. really, to make a good knife you don't really need hundrads hours spent infront of TEM nor 3 tons of specimen smashed in types charpy test to figure out excat reason cause this embrittlment. if you are interest in this subject and inssisted to do so, its up to you and be good luck. the very simple, money and time saving way of avoid the embrittlement is: demands the data sheet from the manufacture where you brought the steel. on the sheet it will show how it is made aka var esr etc...and bounch of graphs you can get advise from. it is a hard and complex subject of course. but to be honest, its not hard to draw an outline in your head which and where in your knife making progress need to be careful with.
 
I've been reading up on this, and you'll all have to accept my apology for the thread necromancy. It didn't seem worth a new thread.

One thing that wasn't mentioned here was time. TME takes about an hour to form. That's why going through 500 isn't an issue, nor is drawing a temper.

TME is a trough around 500F or so. There are other types of embrittlement, like TE, which happen at different temps. They have different causes and solutions.


If anyone else wants to revive this and keep going, I think it is a fascinating topic - and well worth getting more information on.
 
Just to give my raw vision of the phenomenon:
We follow the chart from the origin:
1- we have untempered martensite, full strenght and very brittle
2- we start climbing in toughness until we maximixe the strenght/toughness ratio; we have been just relaxing the martensite, allowing a bit of it's carbon to squeeze out under very fine carbides, fairly dispersed, we have not lost the most of the martensite hardness.
3- continuing the heat ramp, the fines carbides oozing from the martensite starts to collect and become relevant in the grain boundaries, and although the martensite still retains enough of it's hardness, the pathway to fracture is represented by carbides increasing in grain boundaries.
4- if we continue the heat ramp we start to lose brittleness on the martensite side, simply because we are loosing hardness, and now the softer matrix "dampens" the stresses in the carbide's paths. The blade is so ductile now that the matrix itself gives it toughness and the carbides size is still small enough to have a fair coerency with the matrix.
5- now the carbides start to become a high volume fraction of the matrix, and even if the martensite is going to be softer at these temperatures there is also an increase of the strain due to the high carbides volume. That's going toward the secondary "bump" of the hardness of high alloy steels.
6- finally the carbide become spheroidized and we have a glob of wet clay containing some scattered marbles sorrounded by a very soft matrix. Very nice for machining or cold rolling.

As stated above, is needed time at those temperatures for the carbon to take those different assets, so we can "jump" to the desidered temperature and develope that specific configuration.

I hope to have this thing correct and not just mudding the waters :)
 
I wouldn't say your description is wrong. It captures all the tempering phases.

I think you could simply say that TME is when the tempering temperature and time is especially hospitable to the formation of a certain type of cemetite (Widmanstatten needles or plates), and that cemetite in that amount makes the crystal structure more brittle.

What I haven't seen anywhere is an explanation that shows if you can avoid TME by simply using shorter tempering periods, but still get the full benefit of a temper. For example, use four 30 minute tempers at 500F instead of two 60 min. I don't know if you can reset the clock on this type of cemetite formation, or if you'll just go back to growing the needles when you get back to temp.
 
You must log in or register to reply here.
Back
Top