Jump to content

triple quench results


Recommended Posts

heres a couple of pics that show how effective a triple quench is on reducing grain size.Both are 1095 and were held a high heat for about 8 min.The top and second from the bottom were quenched and broken over the anvil.The second and bottom piece were given a triple quench in oil.

triple_quench__800_x_600_.jpg

Edited by mark stephen
Link to comment
Share on other sites

Those are pretty convincing results, I was told it gave a great grain refinement to 5160 and 52100 because of the chromium content, but it looks like it works well on 10XX steel too , Ill have to give that a try. Have you tried it on 1084?

J Anderson R

 

" Fools live to regret there words, wise men to regret there silence"- Will Henry

Link to comment
Share on other sites

Aaron, a longer soak time would grow the grain, which is something to be avoided as it increased brittleness in the steel. Normalizing, which is best done three times also, and triple quenching are used to correct this condition.

 

Mark, I find your results intresting. This was kicked around on Ed Caffrey's site and he held that he only saw an appriciable difference in grain size when there was chromium in the steel.

 

Doug Lester

HELP...I'm a twenty year old trapped in the body of an old man!!!

Link to comment
Share on other sites

I do have to ask, the grain on the sample you hardened once and broke looks realy coarse, did you make sure it didnt get overheated befor the quench? (just thought Id ask)

J Anderson R

 

" Fools live to regret there words, wise men to regret there silence"- Will Henry

Link to comment
Share on other sites

Yes they were both overheated to show how the cycling (triple quenching) effects grain size.John Verhoevens book covers this and shows examples of grain size before and after the process.1086 showed a grain size of 11 and after the cycling was a 15.In other words",ulta fine grain sizes were obtained."

Edited by mark stephen
Link to comment
Share on other sites

Soaking at austinizing temp should have no effect on grain size, unless you go over that temp?

 

Now I'm confused....

 

Have you compared the results of triple quenching to multiple normalizing cycles?

George Ezell, bladesmith

" How much useful knowledge is lost by the scattered forms in which it is ushered to the world! How many solitary students spend half their lives in making discoveries which had been perfected a century before their time, for want of a condensed exhibition of what is known."
Buffon


view some of my work

RelicForge on facebook
Link to comment
Share on other sites

I went way over the temps that i would normal quench at.I did this to grow large grain size.The idea was to take those overly large grains and see how much reduction i could get with the cycles of quenching.I was originaly going to experiment with 5 or 6 cycles to see how it effected the blades hardness after the final quench.Also ,yes the same can be done with normalizing cycles.

Edited by mark stephen
Link to comment
Share on other sites

I'm so confused there is so much conflicting info out there :(:angry::huh:

 

 

This is true...and one of the reasons you can't just take anything found on the internet for granted =/. When you see a post like this, try it for yourself! If you forge a knife and have a tag end at the tang, cut it off, forge it out, thermal cycle it and heat treat it like it's a little knife itself and test it. I've determined to try to leave a little extra on every single knife just for that purpose. I mean, there's pictures on this thread, but you don't know every single step he took in the heat treating process. Just the ones he stated. Perhaps he has other habits that you don't, etc you know? Habits he might not think to state due to the fact that they are what they are...a habit =). The only way to know how it will work for you is to try it.

 

Now, don't get me wrong lol. I'm almost completely new at this, and find myself repeating some information I haven't tried yet...because someone well respected said it. I try to catch myself though. I learned long ago in the high performance/racing automotive forums that misinformation runs the most rampant when people have good intentions. I find that bladesmithing and building high end racing engines have many of the same pitfalls, and are equally complex. We're very lucky that there's far more people willing to share their information in bladesmithing, and also not looking to make a buck off you selling you some wonder tool or part that 'you just can't do without' lol.

 

All I can say is guys like us have to keep our chins up, appreciate the internet for what it is...be thankful for posts like this that show something different than what 'others' may or may not have experienced...

 

And use the information given to spur ideas for our own experimentation and testing =).

 

Back to topic:

 

Thank you for posting those pictures up! I really, really need more experience recognizing grain sizing in steel. Pictures like these really help me to see the differences.

 

Also, just out of curiosity...is triple quenching something useful only for through hardened blades? I can't imagine how one would go about triple quenching something with a clayed blade.

 

Cris

Slow is smooth, smooth is steady, steady is fast, fast is deadly... Erik R.

http://www.facebook.com/scorpionforge
Link to comment
Share on other sites

You could triple quench your blade ,normalize and go to your stock removal.Then you wont be damaging the clay.

 

Ohhh ok. I understand now. By 'triple quench' I assumed, due to something else I'd read on this forum...that we were speaking of the hardening process. Tim Lively's video also says to triple quench when hardening.

 

Now that my head's on straight...back to the discussion =).

 

And, thanks!

 

Cris

Slow is smooth, smooth is steady, steady is fast, fast is deadly... Erik R.

http://www.facebook.com/scorpionforge
Link to comment
Share on other sites

According to Verhoeven, temperature and time are controlling factors in grain growth with temperature being the most critical. If you had a forge that you could control to hold temperature just above austinizing temperature for a given steel then soaking for a reasonably long period of time would have little effect on grain growth, but there is a limit. However, most forges don't have that level of control, so long soaks take the steel way over the austinizing temperature and causes excessive grain growth.

 

When steel goes from austinite to pearlite the new cristals form on the borders of the austinite crystals and are smaller. Each time that this is done, considering that the steel is only just allowed to shift to it's austinitic state between cooling cycles and not cause grain growth, the pearlite crystals will become progressively smaller. That is why we normalize for three cycles and/or quench for three cycles.

 

Another thing that normalizing the steel does after forging is to allow the crystaline structure of the steel to reform and relieve the stress caused by distortion of the crystals caused by forging blows to the steel.

 

Doug Lester

HELP...I'm a twenty year old trapped in the body of an old man!!!

Link to comment
Share on other sites

According to Verhoeven, temperature and time are controlling factors in grain growth with temperature being the most critical. If you had a forge that you could control to hold temperature just above austinizing temperature for a given steel then soaking for a reasonably long period of time would have little effect on grain growth, but there is a limit. However, most forges don't have that level of control, so long soaks take the steel way over the austinizing temperature and causes excessive grain growth.

 

When steel goes from austinite to pearlite the new cristals form on the borders of the austinite crystals and are smaller. Each time that this is done, considering that the steel is only just allowed to shift to it's austinitic state between cooling cycles and not cause grain growth, the pearlite crystals will become progressively smaller. That is why we normalize for three cycles and/or quench for three cycles.

 

Another thing that normalizing the steel does after forging is to allow the crystaline structure of the steel to reform and relieve the stress caused by distortion of the crystals caused by forging blows to the steel.

 

Doug Lester

Ahhhhhh, now I'm finally starting to get it!

 

So just to recapitulate:

- At high forging temperatures: grain grows (by crystals combining I presume?), the higher the temp, the faster the growth

- At just austenitic temperature: crystals don't grow significantly

- Transition austenitic into pearlite (or martensite when quenching), new crystals form in the original crystals and are therefore smaller.

 

By letting the steel pass through phases multiple times, without getting the temperature so high that the crystals can grow significantly, the grain size will be reduced each step.

 

So basically, the reduction of grain size is possible due to the fact that steel can transform from one crystal form to another. How does the speed of the transition have an influence on this? My feeling would say that a fast transition would result in a smaller grain, as the new crystals have less time to grow. And would cooling the steel down to room temperature after the recrystalisation have any effect? My feeling would be no, as the crystals have already formed at that the transition. Am I on the right track here? And what happens if the pearlite or martensite goes to austenite, does that reduce the grain size as well?

Jeroen Zuiderwijk

Facebook page: https://www.facebook.com/barbarianmetalworking

Link to comment
Share on other sites

As your cycling(quenching)your forming the smaller martensite microstructures.The best way to understand is to try it out at the forge.Take a steel you use and heat it to a temp you would forge weld with (as you would with pattern welding)and hold it there for a few min.(again this would be like the time/temp a billet is exsposed to)and quench it and break it.You will see large grains.Simple.Now take another piece and repeat the heating to welding temps and do 1 quench ,again you now have a large grain structure but instead of breaking it to exspose the grain ,take the blade up to the lower temp. you would normaly quench a blade .Quench the blade and repeat this 2 or 3 time.You will see a big difference in the steels structure.You can also thermal cycle this way several times without the quench.This will also reduce your grain size.You dont have to learn all the terms or even understand the process of whats happening to the steel for it to work.The steel does not care.

Edited by mark stephen
Link to comment
Share on other sites

Jeroen, thanks for sending me back to the books. Whether going from pearlite/martinsite to austinite or the other way around, crystalization starts at the boundries of the existing crystals and the new crystals will be smaller. Also the faster the phase change that smaller the crystals will be. The problem presented by the rate of transformation is that the thin edge of a blade will heat faster and gain a highter temperature than the spine of the blade. The rate of growth in grain size increased quickly with temperature. A graph in Verhoeven's book (p 76 fig. 8.3) illustrates that the average grain size at 2hrs @ 1600 degrees F is the same at 6min @ 1800 degrees F. Trying to bring the temperature of the spine up quickly could then results the overheating of the edge of the blade with an increase in grain size. To get around this one would have to optimize the transfer of heat that the lowest possible temperature, such as in a molten salt or molten metal bath. Going in the reverce direction, form austinite to pearlite/martinsite supports multiple quenches. Multiple quenches should produce finer grain, at least in theory, than multiple normalizations due to the quicker rate of transormation.

 

If anyone sees an error in what I posted, please post back. Heaven knows that I frequently need my work check over. Please note that the above only discusses the effect of temperature and the rate of temperature change on grain size along with beginning grain size. There are other thinks that influence it also, such as alloying ellements.

 

Doug Lester

Edited by Doug Lester

HELP...I'm a twenty year old trapped in the body of an old man!!!

Link to comment
Share on other sites

This may not be related, but I'm trying to get my head around the whole thing.

 

Isn't a martensitic transformation a very expansive transformation (thus the curve in a Japanese style blade etc)? Or is it based simply on the atomic structure, meaning how many parts carbon to iron etc, not the actual size of the crystal?

 

I hope that question makes sense. I think I had a good visual as to what you guys are saying, but when Jeroen brought up martensite it sort of clashed in my head lol.

 

Cris

Slow is smooth, smooth is steady, steady is fast, fast is deadly... Erik R.

http://www.facebook.com/scorpionforge
Link to comment
Share on other sites

Chris, just about everything that we do in forging and heat treating has to do with crystal structure. The two basic iron crystals are ferite and austinite. Ferite is the "cool" form (below 1418 degrees F.), it's shape is a cube with an iron molecule at each corner and in the middle of the cube. It is a body centric crystal. Austinite is the "hot" form (above 1674 degrees F.) Again it is a cube but it has an iron molecule at each corner and one in the center of each face with nothing in the middle of the structure. It is called a face centered crystal.

 

Ferite can hold almost no carbon within it's matrix, less than 2 points. Austinite has more room inside it's matrix and can hold 77 points of carbon. When austinite is cooled quickly the carbon does not have enough time to migrate out of the austinite crystal and is trapped when the crystal goes from face centered to body centered. That holds the iron molecules farther appart in the cube. That form is martinsite and it's a larger crystal than ferite with the same number of iron atoms. The atomic bonds between the iron atoms are under stress in this form and the more carbon caught in the matrix the more stress there is. With enough carbon caught in the matrix the resulting steel is very hard but very brittle. Tempering allows some of that disolved carbon to escape and decrease the strain on those atomic bonds. The carbon that is not caught in the body centered crystal of the steel combines with iron atoms to form cementite, iron carbide, or combines with alloying elements to form carbides with that element.

 

The reason that a sword curves that is differentially hardeded is that the martinsite crystal alonge the hardened edged are larger than the ferite and cementite crystals that combine to produce pearlite and free ferite crystals in the spine of the blade. The difference in the size of the crystals causes a strain between the spine and edge of the blade causes the edge to curve towards the spine. That is unless too much stess if formed in the blade in which case it breaks

 

Doug Lester

Edited by Doug Lester

HELP...I'm a twenty year old trapped in the body of an old man!!!

Link to comment
Share on other sites

Chris, just about everything that we do in forging and heat treating has to do with crystal structure. The two basic iron crystals are ferite and austinite. Ferite is the "cool" form (below 1418 degrees F.), it's shape is a cube with an iron molecule at each corner and in the middle of the cube. It is a body centric crystal. Austinite is the "hot" form (above 1674 degrees F.) Again it is a cube but it has an iron molecule at each corner and one in the center of each face with nothing in the middle of the structure. It is called a face centered crystal.

 

Ferite can hold almost no carbon within it's matrix, less than 2 points. Austinite has more room inside it's matrix and can hold 77 points of carbon. When austinite is cooled quickly the carbon does not have enough time to migrate out of the austinite crystal and is trapped when the crystal goes from face centered to body centered. That holds the iron molecules farther appart in the cube. That form is martinsite and it's a larger crystal than ferite with the same number of iron atoms. The atomic bonds between the iron atoms are under stress in this form and the more carbon caught in the matrix the more stress there is. With enough carbon caught in the matrix the resulting steel is very hard but very brittle. Tempering allows some of that disolved carbon to escape and decrease the strain on those atomic bonds. The carbon that is not caught in the body centered crystal of the steel combines with iron atoms to form cementite, iron carbide, or combines with alloying elements to form carbides with that element.

 

The reason that a sword curves that is differentially hardeded is that the martinsite crystal alonge the hardened edged are larger than the ferite and cementite crystals that combine to produce pearlite and free ferite crystals in the spine of the blade. The difference in the size of the crystals causes a strain between the spine and edge of the blade causes the edge to curve towards the spine. That is unless too much stess if formed in the blade in which case it breaks

 

Doug Lester

 

You said it better than I did by far =).

 

What I'm wondering then I guess, is if the martensite crystals are larger than the ferrite and cementite, and we're quenching (forming martensite/retained austenite, and some pearlite/ferrite/cementite depending on the effeciency of the quench correct?) to reduce grain size, how does the smaller grain propogate off of the larger crystals? Maybe a clarification on the relationship between the crystal structures and what we're calling 'grain size' or even the grains themselves would allow me to see what the purpose is. We have atoms, crystals (of various types) formed by various arrangements of those atoms, and then...grains formed of face centered (efficient) or body centered (less effecient) crystal stacks?

 

Thanks for bearing with me here, I've sort of off tracked this topic into newbie land I know...but I think I have a good grasp on the original poster's concept...I'm just missing one key part =). Maybe the detour will help some of the other newer guys reading understand a bit better as well.

 

Cris

Slow is smooth, smooth is steady, steady is fast, fast is deadly... Erik R.

http://www.facebook.com/scorpionforge
Link to comment
Share on other sites

This may just confuse the issue more but, for what it's worth.

A crystallite is a domain of solid-state matter that has the same structure as a single crystal. Metallurgists often refer to crystallites as "grains".
More detailed information here: Wikipedia

 

~Bruce~

“All work is empty save when there is love, for work is love made visible.” Kahlil Gibran

"It is easier to fight for one's principles than to live up to them." - Alfred Adler

Link to comment
Share on other sites

  • 2 weeks later...
How does the smaller grain propogate off of the larger crystals?

 

Cris

 

Chris,

 

The link B. Norris put up says it all... among other things edged around in this thread, new grains start formation (neucleate) at strain points and the strain is in the grain boundries. Martensite has a huge amount of strain, therefore a huge number of neucleation points. But a person doesn't need to quench to reduce grain size, normalizing works well enough.

 

It's a process thing. I forge a blade, then I normalize it, then grind (and remove the decarb), then stress relieve, then austenitize, then quench, then temper. Or I forge, normalize, spheroidize anneal, grind, stress relieve, quench, temper. Normalizing equalizes grain size if done well (even, equal heating and "fast" cooling to ambient temperature in still air). Multiple normalizing sequences make grain size smaller, and equally so. Three, but sometimes four, normalizing cycles reaches the point of diminishing returns. If I quench to reduce grain size before grinding, I have to anneal. In Bruce's Wiki link, cooling slowly lets grains be large. If I do multiple quenches after grinding to reduce grain size I run against the problematic quenching failures... three times... and I've spent more total time getting the blade to this point... and right at the end I've introduced more decarburization than otherwise (a smallish point maybe, but it needs to be removed).

------------------------------------------------------

 

I was at a knife making class two weekends ago and did a normalizing demonstration. Four pieces of 3/16" x 3/4" 1080, each with a number and a shallow saw kerf on the wide side. All four brought to welding heat (2000+F) and soaked 5 minutes, Number one quenched in water and the other three air cooled. Then the three air cooled pieces run above AC3 a little, #2 quenched in water and the other two air cooled, etc. In then end, after breaking (and it was really easy to do), I had four examples of grain size ranging from coarse to very fine. Pieces three and four looked nearly the same to the naked eye... no "grains" evident but different in color. Looking at those two with a 5X hand loope showed a definate difference. With the stub of piece #4, I normalized again to change the structure to perlite. This piece was extremely easy to grind... didn't need annealing.

 

Mike

Edited by Mike Krall
Link to comment
Share on other sites

But a person doesn't need to quench to reduce grain size, normalizing works well enough.

Triple quenching is useful when a person is working in an atmospheric forge without good temperature controls. Using salt pots and electric furnaces pretty much mitigates the need for triple quenching. However, if you are using nothing but your eyes to judge temperature, it is nice to have a little extra insurance in getting the grain size as small as possible. This is just my opinion but, if I'm sending a knife out the door and taking money for it, I want to be as sure as I possibly can of the quality of the blade. Triple quenching ensures that, even if the steel did get overheated and grow some grain while being normalized or forged.

 

~Bruce~

“All work is empty save when there is love, for work is love made visible.” Kahlil Gibran

"It is easier to fight for one's principles than to live up to them." - Alfred Adler

Link to comment
Share on other sites

Bruce,

 

I don't want to go wandering off here but you mentioned grain growth. Here is a thing from "Metallurgy" by B.J. Moniz (a college beginning metallurgy text; ISBN 0-8269-3509-5, for those who want to look it up in the used market Used Book Search... it's dirt cheap).

 

"The grain-coarsening temperature of fully killed steel is 925C to 980C (1700F to 1800F) and is dependent on a variety of factors. For example, it is dependent on the amount the steel has been cold worked. One advantage of fully killed steel is the resistance to grain coarsening during high-temperature heat treating operations (for example, carburizing). These operations are performed at temperatures as highas 925C to 955C (1700F to 1750F)." An chart shows minimal grain growth (7.5 to 7.25 through a temperature range of 725C to 980C (1337F to 1800F) then rapid grain growth (7.2 to -1 ~ ASTM numbers) between 980C to 1055C (1800F to 1931F).

 

I don't know much about steel manufacturing. I assume tool steels and 52100 are "fully killed" (totally deoxegenated?) because of their use. Are they? Are the other common forging knife steels (5160,10xxseries) also "fully killed"?

 

Mike

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...