edge holding and toughness on...

[Phillip Patton]

Kevin,
I know you have an impact tester. Have you done any comparison testing with it? I'm thinking something like 1095, 1084, O1, and W2, all at 60 RC. I'd be very interested to know the results. (Hint hint, nudge nudge. )

 

[KLOW]

Thanks to everyone who chimed in. I appreciate your replies.

So let me see if I got this right:

Edge holding: Toughness:
1. O-1 1. 1084
2. 52100 2. 52100
3. 1095 3. 1095
4. 1084 4. O-1

This look about right?

Cheers, Ken

 

[scottickes]

If you want those kind of numbers, you are going to have to go into the metallurgy world. Try Heat Treater's Guide or steel manufacturer web sites. I'm not thinking of more now, but the numbers are there.

If information from individuals will work for you, Kevin Cashen's information is as educated and experiential as you will find... like you could bank on it... and imagine that in this day and age...

Oh, in the industry numbers you are not likely to find data on 52100 as a cutting steel. It was, after all, a steel designed to roll... why any one would use a steel designed to roll as a steel to cut is beyond me...

Sam, can I come stay with you for a while 'til this blows over?

Mike

52100 wasn't made to roll actually. It was made to carry load and resist wear. So it must be a very tough steel and a very wear resistant steel, to be able to do it's job as a bearing steel.

What properties are of most value in a knife steel? Toughness and resistance to wear!

 

[Kevin R. Cashen]

I would like to apologize for using very general words that are misleading in their inaccuracy. In our material, steel, we can have various strengths. Strength would be the materials ability to resist the deformation under the given load type. We can have tensile strength, compressive strength, flexure strength, shear strength, fatigue strength, and what is often referred to as toughness can be called impact strength.

What most bladesmiths often confuse with toughness is actually ductility, which is the opposite of many of the strengths. Bending a blade demonstrates it’s ductility, or ability to deform instead of failing or resisting the force applied. On the other hand if the blade were to flex and return to true upon removal of the load it would have good flexure strength and tensile/compressive strengths but ductility could be irrelevant.

A material can have good tensile, flexure, or compressive strength and still have lousy impact strength or toughness. My favorite example is my cracked vinyl siding that can bend and stretch all day and not handle a whack at all. Blades are not supposed to be prybars but large ones are supposed to be able to whack and chop things, thus impact strength is also our most logical definition of toughness.

If you approach it from this standpoint the two properties we are looking for in this thread are impact strength and abrasion resistance. But a couple of other properties are worth mentioning, tensile strength and compressive strength. These two are often interestingly connected in an inverse way, but the most interesting is the ability of very brittle materials to have high compressive strength. Two great examples are concrete and cast iron. Both of these materials have excellent compressive strength and yet rather low tensile strength, they are what we would call brittle and have little to no impact strength. Bearings would benefit most from abrasion resistance and compressive strength to handle the wear and the loads, I have found that sudden loads such as impact to be rather bad for most bearings.

Philip, thus far I have used my impact tester mainly for specific checks of my process and materials, and most of that has been with damascus. I have isolated numbers from different tests I could look up but in order for me to giver out numbers and be confident of them I would really prefer to do direct comparative tests of these specific steels with them all being prepared the same. Charpy values are very touchy things to try to compare loosely, and this isn’t helped by all the methods of notching used. For this reason I only work with “V’ notch and un-notched values. I would be happy to run tests if supplied with steel but currently all I have in 10mm x 10mm stock is O1 and prepping stock not in these dimensions is a real pain.

 

[Mike Krall]

Kevin,

So, if related as abrasion resistance and impact strength, is the ordering Ken listed in post #22 correct?

.....AR...............IS
1. O-1.............1084
2. 52100.........52100
3. 1095...........1095
4. 1084...........O-1

Mike

 

[Phillip Patton]

Philip, thus far I have used my impact tester mainly for specific checks of my process and materials, and most of that has been with damascus. I have isolated numbers from different tests I could look up but in order for me to giver out numbers and be confident of them I would really prefer to do direct comparative tests of these specific steels with them all being prepared the same. Charpy values are very touchy things to try to compare loosely, and this isn’t helped by all the methods of notching used. For this reason I only work with “V’ notch and un-notched values. I would be happy to run tests if supplied with steel but currently all I have in 10mm x 10mm stock is O1 and prepping stock not in these dimensions is a real pain.

If you give me all the specs, I'd be happy to make the samples for you.

 

[Kevin R. Cashen]

Kevin,

So, if related as abrasion resistence and impact strength, is the ordering Ken listed in post #22 correct?

.....AR...............IS
1. O-1.............1084
2. 52100.........52100
3. 1095...........1095
4. 1084...........O-1

Mike

Well let's say that it is in the ball park, once again results will vary due to peoples differing ideas and applications, and this is more so with the impact strength. Most of these steels fall into a narrower range in the foot pounds they can handle and are relatively close enough that margins of error can overlap. Introduce something like L6 to the group and then impact strength becomes much more clear since it can be more than double these steels and even large deviations will still be well seperated.

 

[Kevin R. Cashen]

If you give me all the specs, I'd be happy to make the samples for you.

You need enough material to make 3-5 samples of each steel- 10mm X 10mm x 3 inches long. All should have a good, identical longitudinal finish of no less than 220X. Most importantly, of course, all steel of the same type must have ALL thermal treatments identical and all samples must be within 1 point HRC. I will do the notching since it must be done after HT and rather precisely.

This applies to anybody who may want such tests done. If all the prep work is done I am more than happy to break steel for people and give them numbers on it. It is actually the least labor intensive (and most fun ) test I can do if the prep is already done. I may need to wait until better climate conditions though since impact is VERY temperature sensitive and I am struggling to keep my shop habitable right now. Even L6 may behave like glass in Michigan right now! I had to direct traffic around a downed line with the fire dept. this morning and it was darned cold!!

 

[Phillip Patton]

I'm not sure I have enough O1 right now, but I have plenty of the other stuff.
I'm not sure if I can swing the 1084, though. All I have is 1/4" flat. I may be able to upset it to get a thicker sample, unless someone knows of a source for thicker stock?
Thanks for being willing to do this, even if it is fun. I've wondered for long time how the simpler steels would stack up against some of the higher alloyed stuff. I haven't seen any impact toughness data for 1095 or 1084.

You need enough material to make 3-5 samples of each steel- 10mm X 10mm x 3 inches long. All should have a good, identical longitudinal finish of no less than 220X. Most importantly, of course, all steel of the same type must have ALL thermal treatments identical and all samples must be within 1 point HRC. I will do the notching since it must be done after HT and rather precisely.

This applies to anybody who may want such tests done. If all the prep work is done I am more than happy to break steel for people and give them numbers on it. It is actually the least labor intensive (and most fun ) test I can do if the prep is already done. I may need to wait until better climate conditions though since impact is VERY temperature sensitive and I am struggling to keep my shop habitable right now. Even L6 may behave like glass in Michigan right now! I had to direct traffic around a downed line with the fire dept. this morning and it was darned cold!!

 

[Mike Krall]

You need enough material to make 3-5 samples of each steel- 10mm X 10mm x 3 inches long. All should have a good, identical longitudinal finish of no less than 220X. Most importantly, of course, all steel of the same type must have ALL thermal treatments identical and all samples must be within 1 point HRC. I will do the notching since it must be done after HT and rather precisely.

This applies to anybody who may want such tests done. If all the prep work is done I am more than happy to break steel for people and give them numbers on it. It is actually the least labor intensive (and most fun ) test I can do if the prep is already done. I may need to wait until better climate conditions though since impact is VERY temperature sensitive and I am struggling to keep my shop habitable right now. Even L6 may behave like glass in Michigan right now! I had to direct traffic around a downed line with the fire dept. this morning and it was darned cold!!

Does the exact chemistry of the samples need to be known?

Mike

 

[Kevin R. Cashen]

Obviously those elements heavily affecting impact toughness would be important to know. The problem comes in with methods of heat treatment. If you end up with grain boundaries laden with carbides the impact strength is going to be in the basement for even a fairly tough steels, until you fix that issue, also the morphology of plate instead of lathe martensite is going to substantially lower impact values.

Very many of the operations that bladesmiths are fond of will result in this in 52100 or 1095:

This will make these steels have the lowest impact values regardless of the hardness.

While it is beyond the scope of most guys not using very tight temperature controls, it is also possible to dissolve only enough of the carbon in steels above .8%C to achieve hardness and leave the rest in the form if very fine carbides. If this is done in solution amounts from .6% to .8% you can effect Ms enough to eliminate more plate type martensite in favor of lathe type which will greatly enhance impact strength. Thus if done correctly one could get much of the abrasion resistance from 1095 and still have it handle impact like it was 1084, all depending on where you put the carbon with the heat treatments.

Plate martensite forms at lower temperatures under greater strain and at irrational habit plane angles so that the plates impinge upon each other like chunks of ice in a jammed up river which can cause microfracturing, like this:

Lathe martensite forms at higher temps under less strain and in more orderly packets like this:

Thus it is much "tougher". Ms is affected by chemistry and carbon in solution is a major chemical factor. So one can adjust it up or down by mastering the ability to determine carbon in solution.

I like to bring this up to help illustrate that I am not just over thinking things when I stress what exacting controls in heat treating can accomplish. Guys who feel I may be disrespecting the idea that heating to non magnetic and dunking in oil is "good enough", need to understand that what I am really saying is that there is a whole world of limitless possibilities they may never even be aware of if they simply settle with basics of what appears to work. This stuff is so cool and fascinating that the exploration and experimentation and be endless even after you remove many of the unknown factors.

 

[Kevin R. Cashen]

Another point about the above and how it illustrates that it is important to have all the pieces of the puzzle and not just one. Imagine two pieces of 52100 both reading exactly HRC 56, yet one takes 35 foot pounds to break because it has little carbide in the grain boundaries but instead in many nice little fine carbides within the matrix, while the other fails miserably at 15 foot pounds due to eggshell like boudaries or huge carbides in between. Both samples would look exactly the same to the eye, and to the Rockwell tester, both would skate a file and hold an edge, yet one would be ready to fail on the first substantial whack on a solid object. The only differnce between the two could be somthing as big as heat source and soak time, or something as small as a few minutes more of cooling time from non-magnetic in normalizing.

The more complex and carbon rich the steel the more critical these fine details become, the more simple and moderate the carbon levels the more forgiving the steel will be for these minor errors. That is why 1084 is great for beginners ad 52100 or O1 is not! It is also why so much mythology can arise around a steel like 52100; any little thing can produce a markedy different outcome that can be totally misinterpreted as something bigger than it is. Many of the "eureka" moments of bladesmiths making a steel peform like a super metal are quite easily explained by the eventual elimination of some small detrimental effect of using tools or methods not up to the challenge of such a touchy alloy. And adding one extra step may get a steel like 52100 to snap to attention quicker than say O1 due to the absence of vanadium or tungsten, thus causing many working with simple heating tools to prematurely put it above other alloys in performance.

 

[Greg Obach]

Hi Kevin

i really like what you have said here.. .. a good example of the learning curve... i used to dislike O1 alot.. ofcourse it has more alloy and i thought it would do nicely ... but it alway turned out a little soft and not such a good edge holder with my more primitive set up... hardly worth twice the price of the plain carbon steels, yet... when i got pyrometer control.. and really pay attention to the soak.. I was actually surprised.... the steel had some definite changes in properties.. ... typically, i hand sand alot and i knew that O1 was acting different now that i've got control over the process.... it was a devil to sand and need a higher tempering temp... also was much stiffer on a very thin blade...
- now i love the steel... its lovely as a knife.... but what i don't understand is why the numbers for toughness aren't there... i've never experienced that with this steel after the new heat treatments... and i've done some terrible things to that steel to test it.. ( cept that foolish bend test ) .. and it does very good... i'd have no problem making a nice bowie from it and taking it out and the woods and depending on it.. ( unlike anything stainless :thumbdn:

i think all the steels if done right will hold up very nicely in all catagories.... i wonder where these steels ( if properly heat treated ) would let you down if you are using the knife properly.... ..

just something to think about

Another point about the above and how it illustrates that it is important to have all the pieces of the puzzle and not just one. Imagine two pieces of 52100 both reading exactly HRC 56, yet one takes 35 foot pounds to break because it has little carbide in the grain boundaries but instead in many nice little fine carbides within the matrix, while the other fails miserably at 15 foot pounds due to eggshell like boudaries or huge carbides in between. Both samples would look exactly the same to the eye, and to the Rockwell tester, both would skate a file and hold an edge, yet one would be ready to fail on the first substantial whack on a solid object. The only differnce between the two could be somthing as big as heat source and soak time, or something as small as a few minutes more of cooling time from non-magnetic in normalizing.

The more complex and carbon rich the steel the more critical these fine details become, the more simple and moderate the carbon levels the more forgiving the steel will be for these minor errors. That is why 1084 is great for beginners ad 52100 or O1 is not! It is also why so much mythology can arise around a steel like 52100; any little thing can produce a markedy different outcome that can be totally misinterpreted as something bigger than it is. Many of the "eureka" moments of bladesmiths making a steel peform like a super metal are quite easily explained by the eventual elimination of some small detrimental effect of using tools or methods not up to the challenge of such a touchy alloy. And adding one extra step may get a steel like 52100 to snap to attention quicker than say O1 due to the absence of vanadium or tungsten, thus causing many working with simple heating tools to prematurely put it above other alloys in performance.

 

[B.V.]

Ok....that brings me to another question that in my mind comes next. A while ago I did read the recommended HT for O1 and it said something to the effect of a "soak time" of maybe (and I'm quessing as it was quite a while ago) 10 min. soak per inch of thickness of the material. I translated that to around a 2 min. soak for a 1/4" thick blade. That seems a bit short according to some of the things I've seen here on the forums. Is that a correct interpretation?

Brad

 

[Barkes Knife Shop]

The three are the best for a using knife. That's what I use on my knives. The heat treating is about the same. ------

Terry

 

[Mike Krall]

Ok....that brings me to another question that in my mind comes next. A while ago I did read the recommended HT for O1 and it said something to the effect of a "soak time" of maybe (and I'm quessing as it was quite a while ago) 10 min. soak per inch of thickness of the material. I translated that to around a 2 min. soak for a 1/4" thick blade. That seems a bit short according to some of the things I've seen here on the forums. Is that a correct interpretation?

Brad

Brad,

You will find differences in O1 HT practices as listed by different manufacturers... some heat slowly to a 1200F preheat, some go directly into an austenitizing temp. oven or bath. The soak is the time it takes to move carbon, etc. to where it needs to be and though thickness has an effect, 10 minute soaks for blades is usually the recommended minimum... except when it isn't, because Carpenter Technology states a 5 min. soak per inch of thicknesses once the piece has reached austenitizing temperature.

At any rate, soaking longer, even hours, at a proper austenitizing temp. doesn't have adverse effect.

Mike

 

[Kevin R. Cashen]

What Mike said. And this as well...

Part of the preheat and soak is also a matter coming up to temperature for larger pieces. One thing that is controversial but is none the less a myth is the idea that the center of a piece of steel can be cold while the outside is hot. Smaller cross sections attached to larger can heat differentially but if you are say working with a cylinder, due to the way conductivity works, the whole section will come up in the same increments and not have the outside reach temp while the core still needs heating.

All the same the larger pieces will take longer to reach temp so they require longer soaks. Once temperature is reached throughout, it will then take time for the chemistry to undergo diffusion, and there are minimums for this to be done thoroughly. Fortunately, as we have discussed, studied and proven here on this forum (even though it is just accepted fact in the rest of the world), there is no disadvantage to soaking longer (except for decarb). So if you have the ability to hold at temp long enough to achieve proper solution, hold it for whatever minimum the manufacturer recommends and rejoice! Holding at heat without an accurate temp is not called soaking, that would be called overheating the steel.

 

[Kevin R. Cashen]

Hi Kevin

i really like what you have said here.. .. a good example of the learning curve... i used to dislike O1 alot.. ofcourse it has more alloy and i thought it would do nicely ... but it alway turned out a little soft and not such a good edge holder with my more primitive set up... hardly worth twice the price of the plain carbon steels, yet... when i got pyrometer control.. and really pay attention to the soak.. I was actually surprised.... the steel had some definite changes in properties.. ... typically, i hand sand alot and i knew that O1 was acting different now that i've got control over the process.... it was a devil to sand and need a higher tempering temp... also was much stiffer on a very thin blade...
- now i love the steel... its lovely as a knife.... but what i don't understand is why the numbers for toughness aren't there... i've never experienced that with this steel after the new heat treatments... and i've done some terrible things to that steel to test it.. ( cept that foolish bend test ) .. and it does very good... i'd have no problem making a nice bowie from it and taking it out and the woods and depending on it.. ( unlike anything stainless :thumbdn:

i think all the steels if done right will hold up very nicely in all catagories.... i wonder where these steels ( if properly heat treated ) would let you down if you are using the knife properly.... ..

just something to think about

Greg you pointed it out quite well why many of the numbers for these steels don't seem to make sense in knives, it is because real knife use should never push the steel to the limits involved. Perhaps in edge holding and abrasion resistance very hard knife use could tax the steel, but most human arms cannot generate the forces needed to show the differences in impact strength. For years I have seen heat treat methods that I know are resulting in unacceptable levels of fine pearlite that are making blades which are held up as "the best", and they maintain this facade simply because most knife use just does not push steel to the levels needed to notice the difference.

 

[Mike Krall]

Kevin,

So, if related as abrasion resistance and impact strength, is the ordering Ken listed in post #22 correct?

.....AR...............IS
1. O-1.............1084
2. 52100.........52100
3. 1095...........1095
4. 1084...........O-1

Mike

I'd like to go back here. Kevin's response to this was, "Roughly, yes and noted qualifications".

So a knife is abrasion resistance and impact strength. There are steel chemistry's that predispose a steel to excellence for abrasion resistance or for impact strength.

Heat-treating a knife is torch/forge or kiln/salt bath... a difference in ability to control temperatures.

For steels considered to be forging steels that require exact temperatures (kiln/salt bath), it seems O1 and L6 fill the max. abrasion resistance/max. impact strength categories. What two steels fill those categories for torch/forge temperature control (predisposed for excellence)?

Is there a middle-ground forging steel in each temperature control category? Or is it a matter of using the excellent-impact-strength steel, making the overall and edge geometry more "slicer" like, then tempering a little harder? Or does that fall into the "diddling" category? Or ???

Mike

 

[B.V.]

Thank you very much for the clarification. That make sense......even to me!!!!

Brad