Observations on edge retention of some stainless alloys using a visual inspection tec

[knarfeng]

In a thread in the General Forum Vassili asked me about how I compare edge retention. Rather than further distract that thread about Buck 420HC, http://www.bladeforums.com/forums/showthread.php?t=640524
I said I would make a separate thread. Here we go, even though I’m not quite ready.

One of the comments I had made in that thread was that you need both good steel and good heat treat to get good edge performance. Actually, it’s more than that. You need
- Good alloy
- Good heat treat
- Good geometry. This is a combination of edge angle and overall blade profile.
- Good sharpening (Sharp knife cut better and longer than dull knife. Sometimes we forget that.)

The cutting performance of a blade is the combination of all those properties. If you want to look at the effect of just one property, you need to control the others so that performance differences are due to the one property you are examining.

Buck’s published data about their development of their “Edge 2000” blade geometry illustrates this. By using a 420HC blade with an optimized shape and comparing it to the performance of a non-optimized blade of BG-42, they were able to make 420HC outperform BG-42. Of course, when they used optimized blade shapes for both alloys, the BG-42 outperformed the 420HC. The profile and edge angle totally overshadowed the difference in alloy.

Knife maker Phil Wilson uses identical blade shapes and controls the hardness of each blade when he does steel comparisons. This is not an option for most of us. I don’t make my own blades and I don’t heat treat my own blades. So what is a curious fella going to do on a shoestring budget? Limit as many variables as possible.

Here is how I chose to control the different variables
- I control the impact of the blade profile by choosing manila rope as the medium to be cut. The thing about manila rope is that, as soon as it is cut, it pulls away from the blade, so, while edge angle still impacts the results, the impact of differences in blade profile is minimized. This is especially true because of the way I evaluate the edge retention (see below). I use 3/8” manila rope, because that is easy to find at my local Home Depot.

- I control the edge angle by changing the edge angle of every blade I test to the same set angle by using a Sharpmaker. As mentioned above, when cutting manila rope, the edge angle does make a difference, so I control it. I change the edge bevel of all the blades I test to 30° inclusive.

- I control the sharpness by using the Sharpmaker on the same settings. I sharpen on the coarse rods, finishing on the corners of the fine rods.

- I measure the Rockwell hardness of the blades I test so that I know what I am testing. I can’t control the hardness of the blades I test. But I can at least identify the hardness of the alloy sample I am testing. When possible, I compare blades with the same measured hardness.

So, now that I have limited my variables, how do I measure the edge retention? I have chosen to look at the amount of edge deformation I can see after a set number of cuts. Before any cuts, there is zero deformation. Using a 3x-hand lens, I cannot see the edge. I make the cuts, then examine the edge to see what has happened to it.

I set a pair of boards on my workbench with a small gap between them. The boards project over the edge of the bench top. The boards support the rope. The cuts are made over the gap so that I do not end a cutting stroke by cutting into the wood. This ensures that each blade only cuts rope.

Before I cut, I mark off a 2” length of blade. I use that 2” length to make a slicing motion through the rope. So, it is 2” of steel that is the same hardness, the same edge angle, and the same sharpness that is cutting the rope. I make 20 such cuts with each blade.

I do this with anywhere from 3 to 5 alloys at a time.

After I make 20 cuts using each blade that is in a test run, I use that 3x lens to examine the blades. I examine to determine how much deformation of the edge I can see. I then rank the blades going back and forth between them. I look at them and set them in order, most deformation to least. To me, less edge deformation equates to better edge retention. I note the order, then resharpen the knives and repeat the process. I do this several times with a batch until I am satisfied I am seeing a repeatable pattern of retention ranking. It is a time consuming process. When I do another batch of blades, I include one or two of the blades I have already ranked and compare the new ones to those.

So, have I determined anything amazing? Nope. The results pretty much mirror what I would expect to see if I were just thinking about the alloys. More carbon in an alloy gives better edge retention than an alloy with less carbon. A softer blade does not perform as well as a hard one. (This second observation is not shown in the data I am posting today.)

I am posting a plot that I made of some of my results. These are some of the more common blade alloys found in current blade offerings. To make the chart, I arbitrarily gave 440C a 10, then scored the others the way I ranked them when I compared them to one to another. Don’t get hung up on the absolute values. Look at the comparisons alloy-to-alloy. For instance I rated 440C as a 10 and 420HC as a 6.5. You should not infer from this that 420HC had 65% of the edge retention of 440C. But a valid observation would be that I found that 420HC did not hold an edge as well as AUS8 and that AUS8 did not hold an edge as well as 440C.

 

[knarfeng]

I am also posting a table of the hardness values I measured for the blades that I tested.

So, if I only see what I expected to see, does this help me? Yes. It confirms theory. That is always nice to see. Also, it allows me to compare an unknown alloy such as the Chinese 9Cr13CoMoV alloy that Benchmade uses. I found that it was better than 8Cr14MoV, but not quite as good as my Buck 440C. Spyderco said that 8CR14MoV performed about as well as AUS8. I found that it did. (OK, Sal told us the absolute truth. That is hardly a surprise but it is still good to know.) Some folks post the oddest conclusions about alloy performance when they have no actual data but just know it’s true. I gots data.

Anyway, that is what I do.

 

[neeman]

Great post
Thanks
A move away from myth!

A very interesting curve!

Some questions:

Why did you use an arbitary # for "Relative Retention"?
The difference in Relative between 7 and 10.5 could be 3 1/2 cuts in 20 or much much higher.
This is important information, if the spread is only 3.5 cuts per 20 or a different #.

If the differences were small, the marketing guys would use the relative #s.
Aus10 is 25% better than Aus8!!! (2 is 25% of 8)

You have the # of cuts before deformation?
Show us the absolute #s
Then make that in to a rating.
This way you can use your stats to say X alloy cuts Z% more than Y alloy, and know the # of cuts involved
This would be a more useful stat to me.

I work as a Tech Writer, and if I was writing an artical; I would ask the Engineer for these % of cuts stats.


Interesting that the high end of the curve is so flat!!
Also I would love for you to test the D2 of a Queen Cattle King ....Hint hint ... LOL

Again
Thanks for your time and effort in doing this repeatable method.

 

[Phil Wilson]

Knarfing, Thanks for this information. I like the idea of inspecting the edge, I have been doing the same thing. It is also refreshing to see hardness values since of all the variables this is near the top of the list for improving cutting performance. A couple of observations: It sounds like you are seeing mostly edge rolling (deformation). With these steels-- at these hardness-- that is the main failure mode in my experience. The edge will roll as the cutting force goes up and wear resistance does not have a chance to come into play. Hardness equals strength so a harder blade will hold the edge up there long enough for the carbides in a higher alloy blade to resist wear. If the heat treat is right-- as the cutting force goes up--the edge will continue to wear and the fine edge will not fail by chipping. The other variable I have noted is how the rope is cut. It is important to try to cut close to vertical so as to not put side forces on the edge. The edge is very strong with the force down since the the whole blade supports the edge. On a side force there is almost no steel area supporting the edge. You can test this by whittling wood and twisting the blade out of the cut. This is a very severe test of a thin edge and can give a feel for the final sharpening angle required on a thin grind. As a final note if 154 Cm is hardened to about 61 then it will show much better against the other steels you tested. Phil

 

[dextersp1]

Thanks for all that work. I think the way you went about it is the best an individual can do such a test.

It is still difficult for me to wrap my mind around what the difference between a 7 and a 10.5 means in usage. Should the test have gone on to failure? Failure meaning it can no longer cut the rope.
So for example, a 7 fails at 35 cuts and a 10.5 fails at 41.
With that type of info I could then decide the marginal value of the knife steel.

 

[wnease]

cool test technique! more! more!

 

[shecky]

Nice post!

 

[knarfeng]

Great post
Thanks
A move away from myth!

A very interesting curve!

Some questions:

Why did you use an arbitary # for "Relative Retention"?
The difference in Relative between 7 and 10.5 could be 3 1/2 cuts in 20 or much much higher.
This is important information, if the spread is only 3.5 cuts per 20 or a different #.

If the differences were small, the marketing guys would use the relative #s.
Aus10 is 25% better than Aus8!!! (2 is 25% of 8)

You have the # of cuts before deformation?
Show us the absolute #s
Then make that in to a rating.
This way you can use your stats to say X alloy cuts Z% more than Y alloy, and know the # of cuts involved
This would be a more useful stat to me.

I work as a Tech Writer, and if I was writing an artical; I would ask the Engineer for these % of cuts stats.


Interesting that the high end of the curve is so flat!!
Also I would love for you to test the D2 of a Queen Cattle King ....Hint hint ... LOL

Again
Thanks for your time and effort in doing this repeatable method.

Hi Neeman,
I chose an arbitrary number so I could get Excel to plot the data. Excel doesn't handle "fair" or "better". It needs numbers. So, starting with 440C as a 10, I worked Down and Up in relative retention assigning numbers so I could show it would show on a graph. I like visual representations and this is the method I have developed to show "fuzzy" data at work.

I did not record the number of cuts before deformation. I did 20 cuts because one of the first steels I tried was AUS8, and after 20 cuts, it showed some deformation, but the edge was not completely gone. So that appeared to be about the right number of cuts since steels with better edge retention would show less deformation and steels with less edge retention would show more deformation. The damage to AUS8 after 20 cuts appeared to be about midrange.

The chart flattens out at the high end because, 20 cuts is not enough to do much damage to any of those alloys. I believe that if I tested VG10, 154CM etc using 30 or 40 cuts I might be able to use this technique to differentiate between those higher performance alloys.

I haven't tested the Cattle King because, being a stockman, the blades are annealed at the base and I can't get a hardness reading. I wouldn't know what I was testing. That and I have been shying away from making multiple runs and multiple sharpenings with the D2. I did rebevel the edge to 30° though, so, maybe I'll give it a try.

 

[knarfeng]

Knarfing, Thanks for this information. I like the idea of inspecting the edge, I have been doing the same thing. It is also refreshing to see hardness values since of all the variables this is near the top of the list for improving cutting performance. A couple of observations: It sounds like you are seeing mostly edge rolling (deformation). With these steels-- at these hardness-- that is the main failure mode in my experience. The edge will roll as the cutting force goes up and wear resistance does not have a chance to come into play. Hardness equals strength so a harder blade will hold the edge up there long enough for the carbides in a higher alloy blade to resist wear. If the heat treat is right-- as the cutting force goes up--the edge will continue to wear and the fine edge will not fail by chipping. The other variable I have noted is how the rope is cut. It is important to try to cut close to vertical so as to not put side forces on the edge. The edge is very strong with the force down since the the whole blade supports the edge. On a side force there is almost no steel area supporting the edge. You can test this by whittling wood and twisting the blade out of the cut. This is a very severe test of a thin edge and can give a feel for the final sharpening angle required on a thin grind. As a final note if 154 Cm is hardened to about 61 then it will show much better against the other steels you tested. Phil

Thank you, sir. You do me high honor.

Those are great comments and excellent insights on the importance of making straight cuts with no side forces. I will pay more attention and perhaps I won't have to make so many repeated runs.

hmmm... gonna have to find me a 154CM blade hardened to 61. Cool!:thumbup:

 

[PatriotDan]

Thanks for your work and good presentation! Excellent read on something new and unfamiliar to me.

Whats the reason why S30V is not included??


Thanks again :thumbup:

 

[knarfeng]

Thanks for your work and good presentation! Excellent read on something new and unfamiliar to me.

Whats the reason why S30V is not included??


Thanks again :thumbup:

Thanks Dan,

I don't own any blades made of S30V.

 

[Broos]

Thanks for the info! I think the visual edge inspection method (or better yet pics) is a good method for comparison. Your eyes can see in a second what would take many pages to describe, define, and quantify.

 

[nozh2002]

Major questions:
What do you count as a deformation?
I observe edge under microscope and rolled edge looks as good as hair whittling edge under x200 microscope really. And so for example AUS8 which shows prety poor results on my tests may have damage which affect sharpness but does not seen under x3 lens.

Do you inspect blade before testing?
This is is important baseline to see that you counting only deformation which are result of cutting.

Thanks, Vassili.

 

[knarfeng]

Hello Vassili,

When I examine the edge, I use a high intensity lamp and look for light reflection along the length of the blade. Before I run the tests, I cannot see any reflected light. For that matter, under 3x mag, I can't see the edge at all. After making the cuts, I base my evaluation on how much edge I can see and on how wide it then appears. On alloys such as AUS8, I can see the entire length of the edge. On VG10 and 154CM I only see small patches of it.

Since the AUS8 blade that I use and the VG10 blade that I use are both 59HRC, and the edge angles are the same, this is telling me that VG10 has better edge retention than AUS8. This is hardly news to anyone, but it illustrates my methods.

I would guess that Phil Wilson is right and that what I am seeing on the AUS8 is edge rolling. But the VG10 doesn't roll at the same hardness. So something is going on there that maybe a metallurgist could talk about. I can't tell you what is happening, only that it is.

 

[nozh2002]

Hello Vassili,

When I examine the edge, I use a high intensity lamp and look for light reflection along the length of the blade. Before I run the tests, I cannot see any reflected light. For that matter, under 3x mag, I can't see the edge at all. After making the cuts, I base my evaluation on how much edge I can see and on how wide it then appears. On alloys such as AUS8, I can see the entire length of the edge. On VG10 and 154CM I only see small patches of it.

Since the AUS8 blade that I use and the VG10 blade that I use are both 59HRC, and the edge angles are the same, this is telling me that VG10 has better edge retention than AUS8. This is hardly news to anyone, but it illustrates my methods.

I would guess that Phil Wilson is right and that what I am seeing on the AUS8 is edge rolling. But the VG10 doesn't roll at the same hardness. So something is going on there that maybe a metallurgist could talk about. I can't tell you what is happening, only that it is.

Let me just spell it for myself - so you are looking at like bright line at the very edge which indicates different flat then edge sides made by sharpening.

Do you check both sides?
Do you account chips as well?
What differ 420C from AUS-8 - if AUS 8 already deformed on entire edge?

Thanks, Vassili.

 

[knarfeng]

The 420HC edge is wider if I remember correctly. Been a while since I collected the data. I rank the ones I test and record the ranking. Lacking your photographic skills and equipment, I have no way to record the actual amount of damage.

I angle the blade so that the point of the knife is away from me, I angle the knife so that any flat spots on the edge will reflect light toward me. So the edge of the knife is facing me.

I suppose it is hard to visualize. I'll try to figure out a way to make a diagram. That will make it more clear, I think.

 

[knarfeng]

I did not show the hand lens, but this should give you the idea.

 

[Big Mike]

Very nice work, your approach seems quite sound to me.

As you state, no real surprises here; taking the high end steels a bit farther might be interesting.

I thank you for your time and effort put into this, and encourage you to continue your work.





"If you're not living on the edge, …you're taking up too much space."

Big Mike

Forest & Stream

 

[Condition 1]

Now that's a review!!! :thumbup::thumbup:

 

[dextersp1]

Thanks for all that work. I think the way you went about it is the best an individual can do such a test.

It is still difficult for me to wrap my mind around what the difference between a 7 and a 10.5 means in usage. Should the test have gone on to failure? Failure meaning it can no longer cut the rope.
So for example, a 7 fails at 35 cuts and a 10.5 fails at 41.
With that type of info I could then decide the marginal value of the knife steel.

Any insights available?

Basically, I'm trying to understand how much better is a 10.5 vs a 7?

Can I get 50% more use (before requiring sharpening) with a 10.5 vs a 7 or 5% (that type of info)?