1095 HT troubles

[Andrew Garrett]

I read threads like this and feel like an idiot!

I make most of my knives from 1095. Most are bars purchased from Sheffield's or Jantz. Some are from Nicholson 'Black Diamond' files.

I find 1095 to be very easy to work with and have come to dislike the heat treating of most other carbon steels.

My method (apparently FAR too simple and without adequate science invoked):

I clay coat most of my knives and stick the blade in the atmospheric forge 'edge up' while the Satanite is still wet (NEVER had a problem with that practice). I hold the tongs with an under handed grip in my right hand.

I dip the tip a few times as it gets redder faster until the rest of the edge catches up. (thank you Ed's basic bladesmithing video)

When it's even and orange, I start testing against my magnet. Once it goes nonmagnetic, I stick it back in for 3 to 5 seconds and pull it out, roll my wrist over and dunk it in my olive/veggie oil quench.

I push forward and pull back (never side to side). I'd never heard of 'thermal jacketing' until I just read it, but I think that's why I move the blade in the quench--I want it exposed to cooler oil than that which is super heated the instant the glowing steel hits it.

The oil often catches fire (kinda fun) and I usually just let it burn while the edge under the surface cools a bit more.

A paper towell and a wire brush get it ready for the vice where I give it a file test. I listen for that high pitched skitter and then it's locked in a c-clamp and placed in my pre-heated kitchen oven--again--'edge up'.

One hour at 450. Cool. Some get a second temper. Some don't.

I get excellent performance from these knives (read: no complaints). The edge holding ability is better than any I've made from 1080 or 1084, using the same HT method.

Good luck Alan.

[Kevin R. Cashen]

Originally Posted by sdcb27

Kevin,
First off let me be the first and little doubt not the last to say thank you for reposting.

Careful, those who know me would tell you that it is best not to encourage my behavior

I would start by saying that it is the respect that I have for Ed Caffrey and our friendship, would give me pause to ever post things contradictory to what he may have offered previously, his gentlemanly approach to this helps ease that action in hopes that we may understand sources that would suggest 1650F to the benefit of us and others concerned. Some folks make it really easy to disagree with them, Ed Caffrey?s level headed and kind demeanor makes it almost painful.

Originally Posted by sdcb27

To be clear your saying ALL STEEL can and will benefit to some degree from a properly controlled soak not just a hypereutectoid steel?

Indeed. And my suggestions about hypereutectoids (over .8% Carbon) was more about the temperature than the time. Soak time becomes more important with steels that have alloying which will retard full austenite solution, and with hypoeutectiods (steels with less than .8% carbon) since the carbon will need some time to diffuse into the low carbon areas. With the higher carbon stuff overheating becomes more critical as it will result in brittleness and retained austenite. With simple steels the grain growth will be delayed but once the last bit of residual iron carbide is dissolved grain growth will be very rapid, and the larger the grains the greater the high carbon embrittlement will become. In fact with over .6% carbon into solution and large grains many of the individual grains will be littered with microcracking that you will not even be aware of until the steel utterly fails.

Originally Posted by sdcb27

Again a resounding Yes please! I beleive these are the photots I have seen before, Fitz and I were talking about them last night, where your ears burning?HEHE

Here is an excerpt from a thread on another forum on the same topic, I have pasted it here instead of retyping it all:

The people who make the steel we work with thoroughly work out the best ways to get the most potential out of them. As bladesmiths we trust them to make our steel yet often seem to doubt they know anything about treating and working it. The assumption that just because a piece of steel is in the shape of a blade that none of the properties or principles apply is erroneous at best. Often the reason that later numbers on the spec sheets don?t apply is because we didn?t pay any attention to the previous numbers. We totally ignore the recommended heat and soak times and then try to say that the books are wrong when we get different results and the tempering temperatures don?t match up later on. One can?t pick and choose which recommended procedures they are going to follow or ignore and expect to have all the results match up.

Let?s start with 5160 again. If done the way the spec sheets recommend we will heat to 1525F and soak for 1 hour per inch of thickness. Now I will dispense with any discussion about the idea that this could cause grain growth since we have put enough nails in that coffin allow it to rest in peace. The reason we need to go higher with 5160 or 1050, than say 1095, is mostly due to the fact that is has less than .8% carbon which means will consist of 75% pearlite (finely segregated iron and carbon) grains, and 25% ferrite (low carbon iron) grains. When we heat this to the critical temperature (there are formulas to account for the alloying that will give you around 1370F for A1 in 5160), the pearlite will begin to dissolve, fairly quickly but certainly not the instantaneous change the Iron carbon equilibrium diagram would theoretically suggest. But anyhow the pearlite?s time is now over and we need to worry about that darned ferrite (low carbon iron). That is what the upper critical temperature (Ac3) line is all about. From Ac1 to Ac3 we are putting more and more carbon into that iron to make austenite out of it. When we cross Ac3 we will have accomplished this, but there will still be plenty of carbon trapped in carbides to pull into the mix if we want good hard steel (martensite) when we quench. Yes carbides! Even in 5160, because that pearlite was made up of ferrite and cementite (iron-carbide) with a good amount of chromium carbides thrown into the mix. In 1095 you can make all austenite and leave many of those carbides in place because with .95% carbon you have more than enough to spare and dissolving even more could even be a problem! But with 5160 you are right on the threshold of maximum hardness to begin with leaving too much undissolved carbide will result in very wimpy martensite. Thus you need to go to 1525F and hold it there to let the heat do its job if you want to get the most out of the steel.



This is 5160 lamellar annealed (what you get from heating to critical and stuffing in vermiculite, wood ash or your forge) and then cycled several times a low heat (just under or just at nonmagnetic) and then heated to 1414F (non magnetic) for 5 minutes and quenched. Does this heat treatment seem familiar?

Here you will see very wimpy martensite (low carbon) that is in the areas that managed to go into solution. If you look you can actually see the interface where the pearlite lamellae dissolved together and formed austenite solution. But more interestingly you will still see the ghostly remnants of the lamellae in the form of lines of individual undissolved carbides, looking like strings of black beads in the martensite.

This sample took considerable effort to break, but snapped as if it were fully hard. It cut neat groves down the side of, and ruined, a brand new Nicholson file in just two high pitched passes. Anybody how has ever done a file test would have passed the steel with flying colors. Indeed at a gathering in February some the top names in bladesmithing passed it around and skated a file on it, before I revealed that it was only 47 Rockwell ?C?. One of the greatest pitfalls plaguing bladesmithing is the use of tests that do not tell at all what we think they tell us, the file tests scratch hardness NOT penetrative hardness, thus it has a very difficult time telling us if we have unwanted pearlite (soft stuff). The file test is still useful, we just have to realize that it is not telling us what most bladesmiths think it is telling us.



Here is another image of the pearlite being broken down into spheroidal structures before going into solution.

[Kevin R. Cashen]

Oh now come on guys! You ask for the interesting stuff and then leave . I want to get you talking, not be a thread killer. Or have I finally done it and actually bored something to death

Actually I wanted to stop back in and share my thoughts on something that occurred to me about the reason some sources may recommend 1650F for a simple carbon steel. I was mulling it over and remembered the automated heat treating lines I have seen and some other industrial type applications. In order to save on time with an automated system industry can replace time with temperature if they have it all worked out to the exact decimal point. When the parts move through the furnace at a rate of a couple of minutes the soak time can be drastically shortened by upping the temperature to increase diffusion rates. This is not an undertaking to be approached lightly, obviously when you are shortening soak times to seconds by taking the steel to very high temperatures you are going to need to be in complete control of exact temperatures with very precise times every time. This is accomplished by automated systems capable of repeating the process exactly again and again, and these days there are a cood ammount of computing involved to keep track of all the parameters. Modern specs or texts may very well include heat treating recommendations for these modern and complex operations.

The safer way, and the only practical way, for the bladesmith is to follow the more general recommendations of letting time do the work and eliminating the threat of overheating. If kept around 1475F most of us simply don't have the time or the patience to wait long enough for grains to grow, but we can wait the 10 or 15 minutes it takes to get good solution. Unless of course all we have is a forge or a torch, then switching to a much simpler steel like 1080, 1075 or 1084 will make our life a little less stressful.

[mete]

That's right , don't copy us metallurgists we know what we're doing !!

[AcridSaint]

Hey Kevin, I was actually just discussing that on the phone last night with a good friend. He mentioned the exact process you're describing. We were also talking about looking at some of those knives and parts heat treated that way under a scope to see how things look.

[AlanR]

Originally Posted by Kevin R. Cashen

Oh now come on guys! You ask for the interesting stuff and then leave . I want to get you talking, not be a thread killer. Or have I finally done it and actually bored something to death

Actually, I thought those were images of the Moon's surface and you had hijacked the thread for lunar viewing!!

Glad to see this topic continue. Andrew Garrett might have cleared up the easiest part of my problem. I'm using an oven and by the time I open the door and get to quench the temp has probably fallen off quite a bit - even though it's as fast as I can do it. I remembered after his comment that the first knife I used SS Foil and it might have protected the blade from room temp some? Dunno.


-AlanR

[AcridSaint]

Hey Alan, the foil might have added some mass and kept circulating air off of the steel, but for 1095 I wouldn't use foil. It could add an odd dynamic to your oil quenching.

I use an oven for 10 series steels and don't have a problem with it. You do want to be as quick as possible and keep your quench nearby. However, as I've mentioned previously, I stick with 1084 and 1080 for the most part.

[jjh]

Just wish I was able to follow it all I keep reading and understand more each time. D**m my feeble brain

[sdcb27]

And remeber 1095 curve is .6 of a second exstremly fast. I would bet there is the issue. Even IF one was over hot at 1650 it would still harden to some degree IF you and your quenchant were quick enough. I have never used a oven but presume (hypothetically) you open the door , may or maynot fumble around getting the blade, pull it out , maybe close the door, turn to quench. plunge it and THEN you still have to get beat the curve all within that time frame.
When ever i personally have bunged up hardening a blade its been because of me,i.e. trying to do to much at once, not focusing becuase someone is talking to me. not pointing the quench tank magnetic north ect. Just Joshing ya Kevin

[son_of_bluegras]

I do have a couple of questions, it just takes a while for my mind to mull over what enters it. (some say I have a mind like a steel trap... it mangles whatever gets caught )
For one, I am curious about the steel that skated a file at 47 Rc hardness. How well would that perform as a knife? Does the ability to resist a file indicate a steel that will take an edge and hold it well enough or would the edge just bend/fold at the edge resulting in something that looks like a knife but doesn't cut worth a d***?
And for those of us stuck using primitive methods and equipment, what sort of test can be done to ensure we are getting what we want? I do destructive testing on my knives occacionally and use the others so if someone wants on I can pass it on with confidence that it will serve the purpose. But I would like to do the best I can with what I have available and I know about the file test and can break the occasional blade blank after heat treat, is there something else I should be doing without investing a lot in equipment?

I am still mulling over some of the other things you have brought up (many from past threads so I've been at it a while), so one day other questions may follow.

ron

[cdent]

Would anyone comment on how quickly heat is lost (degrees/sec?) in still air of say a .035" edge? And, how much slower would the heat loss be of a say .210" spine near the ricasso?

In other words, what would the temp of an edge be that started at say 1475* and took one second to get in the quench?

Thanks, Craig

[Kevin R. Cashen]

Originally Posted by cdent

Would anyone comment on how quickly heat is lost (degrees/sec?) in still air of say a .035" edge? And, how much slower would the heat loss be of a say .210" spine near the ricasso?

In other words, what would the temp of an edge be that started at say 1475* and took one second to get in the quench?

Thanks, Craig

When the steel is heated above 1335F it begins to go into solution (austenite), this is not an instantaneous thing but a gradual increasing process with temperature and time. When held at 1475F the previous structures will be dissolved and replaced by austenite until eventually you have an internal makeup that is entirely austenite with whatever leftover carbides are permitted.

Cooling is the same way in reverse but not exactly at the same temperatures. Austenite will begin to transform to pearlite as the carbon comes back out of solution but this will be a very slow process to begin with and then get increasingly faster until around 1000F where it will be much harder to avoid. So for most of our purposes we have around 400 degrees to play with before we really need to get worried, the blade will be getting rather dark in color when this point is reached.

So you can imagine our cooling as a line on an X/Y axis with time being horizontal and temperature being vertical. Our cooling will not be a straight line but a curve, and the time from the oven to the quench being more horizontal than veridical due to the slow cooling nature of air. That line will progress towards pearlite much slower until the hit the quench, then the line will rapidly drop to more vertical and we will be on a collision with pearlite at 1000F if the quench is not fast enough to steer us to the left of it in time.

Many makers get very hung up on time from oven to quench with something like 1095 but time in the air is much more abundant than time in the quench. When using a substandard quench that may cool slower or form heavy vapor, we may want to blame the pearlite formed on our fumbling too long in the air when we may need to really consider another quenchant.

[Andrew Garrett]

Ron,

I guess I missed where someone wrote that they skated a file at 47RC. I simply do not believe that. Either the file was annealed, or the testing was flawed. No offense to the author.

The file test has little to do with edge retention in and of itself. It's just an easy way to determine if the steel achieved good hardness after the quench.

The theory is, that a file, by it's nature, must be hard. I use Nicholson files for testing as they are made from 1095. During their creation, they, themselves, are quenched and then lightly tempered to relieve stress rather than soften the steel. They are among the hardest tools at a common hardware store.

Logic then suggests that if you bring a 1095 blade to critical, quench it, and let it cool, it will be harder than the file since it has not enjoyed any tempering yet--while the file has.

So, we clean up the blade a bit so we are not decieved by the file taking off the black stuff and mistake it for cutting the steel. Then we vice it and lightly test the edge all along it's length to insure that the blade is harder than the file. If you push down hard enough, it will actually dull the file since the blade is harder. If you get that high pitched skittering sound, you win. Temper it as desired, finish it out (while keeping the steel cool), and call it a knife!

I cannot imagine any blade which properly resists being cut by a good file, being prone to rolling the edge. If sharpened properly with good geometry, it should be fine.

[AcridSaint]

It was Kevin who wrote that after my first post (and before it). As he pointed out, scratch resistance isn't the same as penetration. A rockwell tester pushes down and deforms the steel, a file just scratches the surface.

Next time you get some decent scale buildup on a knife, skate a file across the scale. It's no good for a knife edge, but it's awfully "hard" stuff.

[Kevin R. Cashen]

Originally Posted by son_of_bluegras

I do have a couple of questions, it just takes a while for my mind to mull over what enters it. (some say I have a mind like a steel trap... it mangles whatever gets caught )
For one, I am curious about the steel that skated a file at 47 Rc hardness. How well would that perform as a knife? Does the ability to resist a file indicate a steel that will take an edge and hold it well enough or would the edge just bend/fold at the edge resulting in something that looks like a knife but doesn't cut worth a d***?
And for those of us stuck using primitive methods and equipment, what sort of test can be done to ensure we are getting what we want? I do destructive testing on my knives occacionally and use the others so if someone wants on I can pass it on with confidence that it will serve the purpose. But I would like to do the best I can with what I have available and I know about the file test and can break the occasional blade blank after heat treat, is there something else I should be doing without investing a lot in equipment?

I am still mulling over some of the other things you have brought up (many from past threads so I've been at it a while), so one day other questions may follow.

ron

The more I study and test these things the more I am convinced that people simply do not use their knives the way they think they will, and the most of the uber-performance stuff is little more than more sales pitches. From common heat treating practices among bladesmiths I am pretty sure there are a whole lot of blades out there in similar condition as that sample, and yet we rarely get less than rave reviews and claims about the hand forged blade.

Edge holding is mostly dependant upon abrasion resistance with some other strength factors coming into play with edge geometry. A blade that is entirely martensitic (no leftover pearlite) with very fine and evenly dispersed carbides will beat that 47HRC steel hands down in both categories. Files measure scratch hardness not penetrative hardness, and there is absolutely nothing wrong with scratch hardness or the file test (it is still used widely in industry), we just need t be very ware of what we are actually measuring with any test or that test could be more detrimentally deceiving than useful.

On scratch hardness vs. penetrative hardness- imagine a pile of playdoh with large pieces of glass mixed into it. Running big file over it will only result in the file skating on the glass, giving the impression that the whole pile is harder than the file. But if you then push a pointed metal rod into the Playdoh it will simply push the glass aside and give the impression that the entire pile is as soft as Playdoh; however the truth is somewhere between these two readings.

If we had some health issues and went if for a thorough physical, we would be rather unhappy if all the doctor did was tell us to stick out our tongue and go ?ahhh? before proclaiming us fine and sending us home. With everything that can happen in the human body one very simple test would never be enough for us to believe everything is fine. We knifemakers too often do the equivalent in proclaiming our process perfected based on the erroneously concluding that one rudimentary can be definitive.

Add to all of this that games can be played with how things cut. A homogenously strong edge will obviously hold up the longest; however a blade that is file hard with patches of soft stuff will wear down differently than the fully hardened one. Instead of uniformly smoothing over, the soft chunks will tend to wear faster and tear out leaving jagged pieces of the hard stuff sticking up to tear aggressively into soft an fibrous materials, giving the impression of increasing cutting power. But this only holds true on certain materials and certain kinds of cuts, when chopping harder materials that soft stuff will lower the overall penetrative hardness and decrease the life of the edge, increase the chance of it rolling or even chipping.

Back to my original point, most of us could just say good enough and why bother with getting rid of that pearlite if most the consumers seem really happy with it as is. But a certain percentage of makers will be driven to madness knowing that pearlite is there and they were not in total control of their heat treatment in order to get every bit out of that steel. ?good enough? or ?there is always room for improvement?, on a daily basis, each of us has to decide which mindset to go with.