Home Articles Forums Gallery Marketplace Videos

WoodCentral Forums

Knowledge You Can Build On — Est. 1998

Pictures from Testing for Article

Posts

Pictures from Testing for Article

#1

Pictures from Testing for Article

David Weaver

I did a more controlled test (not a scientific statistical test, but reliable enough to test a theory).

I thought at the outset that the unicorn would make the cheap chisels better really good chisels with a flat bevel. Well, not quite. But close chopping maple.

All chisels were flattened on a washita that has 1 micron diamonds on the surface. The washita cuts fast enough that I think the diamonds are meaningless, but the finish on the back of the chisel from the washita could be the difference in effort needed.

13/16ths of maple, 1 3/8" tall and one inch of cut length chopping. Using a verawood mallet and only rotating my wrist so as to have a similar strike each time. The two V11 unicorn tests were first and last. I was pleased to see them so close together in effort, but I didn't run them to prove that. I ran them because my first go at unicorn on V11 didn't yield no edge damage and in my test with the verawood mallet, I'd need to experiment with it a little bit more to accommodate it. That aside, I ran a single test with a 25 degree edge and aside from some large areas of failure, some parts of the v11 edge were clear for the width of the microscope view (2 hundredths). You could chop for a while at 25 degrees if finish isn't important and you were will to do a very hefty job on the next sharpening. It's different than the others.

Here is a table of the results (note that I managed to do some very minor rounding of the back edge of the Mk2 chisel and have to admit it was beneficial for one reason - that chisel went straight into the cut without any diving, even with a heavy cut. I intended to just apply the method and record data, not do it over and over and over until I couldn't see any faults. Because nobody here will do that, either.

I'm going to post this as a gallery to imgur without any explanation and see what kind of goofy comments come from it :)

(first pane in the picture is just after honing, and the rest are after chopping maple)






Re: Pictures from Testing for Article

#3

Boy am I a DS sometimes..picture fix

David Weaver

The last two pictures are japanese. One should've been V11.

I'm sure I could find a way to accommodate the V11 with this method to be damage free, but it is truly different in is behavior from what I can see in regard to sensitivity to bevel. It cuts well at a low bevel angle and it's harder to get it to cease all damage at higher angles.

At the same time, I was able to significantly limit the sorby damage, but in the test, it wasn't eliminated. Since I'm unfamiliar with the V11 chisel and bought it for this test, I unicorned it "harder" the second time until it as visually similar to the other chisels and the picture is from the second trial and not the first.

The yellow cells in the data table are intended to show that there's a real possibility that if the average person unicorned a sorby chisel, they may get equal performance to V11 in some variations. You can see from the pictures of the edge damage that this is also in regard to durability.

Otherwise, the V11 substantially outperforms the sorby and that's not a surprise. V11 also seems to give less feedback when damaged (whatever it does to fail, or perhaps something to do with the amount of slick chromium in the alloy, it doesn't feel blunted and bounce in the cut as much as the damage adds up).

These are all just observations. The sweetest of the chisels in sharpening and use combined is the japanese chisel, and the MK 2 AI sharpens and grinds much better than does V11 (even when we're talking about using diamonds and CBN), so when unicorned, it's my non-japanese preference. That won't be in the article.

I didn't use the japanese chisel unicorned before this, either - like the V11 chisel, I just set it up and did the tests. Similar with the MK2, I unicorned it a bit before, but not much. Too much fascination with stepping up wonky chisels, I guess.






Re: Pictures from Testing for Article

#4

why not use the same number of mallet strikes?


Re: Pictures from Testing for Article

#5

I don't understand the experiment

Bill Tindall, E.Tn.

Winston did a similar experiment and preceded the edge pictures with the chisel chopping conditions captured with a picture. A good idea I think.

Is it correct that your definition of Unicorning involves a primary 20 degree bevel, a ?? secondary bevel and a buffed tertiary bevel? The others are flat 32 for example?

You report 3 measurements 13/16 (thickness of maple board?), 1 3/8" length of the chop, board on edge?, and in the results 1" cross section (width x height I presume?) .

Surely it did not take 100 whacks to drive a chisel 1 3/8" length one time? If it did this whacking is far more dainty than I employ and many will conclude the whacking was designed to go easy on the edge not test its robustness.

What is the objective of recording # of whacks? I don't see what conclusions can be drawn with bevel angle, progressive dulling, and nature of dulling as it effects penetration convoluted in the experiment.

Re: Pictures from Testing for Article

#6

Re: Pictures from Testing for Article

Winston

It's great to see your test results. I have to say I'm surprised there's such a big difference between 30 and 32 degrees for these chisels. I'm no longer surprised that the buffed edge holds up so much better.

> 13/16ths of maple, 1 3/8" tall and one inch of cut length chopping.

Can you explain your test a bit more? It's not clear to me what the procedure was.

I wonder if some of the differences in durability between the chisels is due to different behavior when buffed. The PM-V11, for example, gets shaped more slowly when buffed, and maybe the edge didn't get as much of a convex profile as the others.

I know it can be difficult to get profile pictures (especially when the chisel has no lands, as in the Veritas chisels) but they may be enlightening.

The reason I'm thinking about this now is because I went back and looked at profile photos of the Buck Bros chisel when buffed vs. stropped. I saw that the buffed chisel had an extremely high angle right near the edge, as in 60 degrees or more, and it's possible that the high angle is the reason that the edge held up so well when buffed.

In the profile photos you posted earlier, it looked like your buffed chisels were 45 degrees right at the edge. I wonder if a higher angle would increase durability without adding noticeable cutting resistance.

Re: Pictures from Testing for Article

#7

penetration

Bill Tindall, E.Tn.

There is no reason I can think of nor literature data to support the hypothesis that a stainless steel chisel has less friction against wood than carbon steel. I think Wiley and I compared and reported these results.

There are enough variables convoluted in this experiment I don't see how a cause of the minor differences in whacks can be assigned. To list a few- bevel behind the convex bevel, profile of the convex bevel, shape of the damage as this shape affects penetration(I'll bet folds are worse than chips), damage vs whacks.

If number of whacks revealed something about damage where it could be a surrogate for photography there would be a reason to explore this possibility. I don't see this correlation likely. So, what use can be made of the number of whacks results?

I do like the comparative photo presentation. Worth 2000 words but the impact is obscured by the whacking table.

Re: Pictures from Testing for Article

#8

Re: I don't understand the experiment

David Weaver

hard maple 1 3/8 wide, 13/16" thick and then chopping a cross section 1" long.

In volume, I'd figure it's about equivalent to three half blinds, so the chopping numbers are not out of the realm of expectation.

I would normally hit chisels a little bit harder, but not too much harder.

The reasoning for the test is simple - if you are doing lots of hand work with chisels, chopping the waste out of half blinds will probably be one of the most intensive things that you'd do (mortises, of course, but we're not addressing mortise chisels because the sash mortise size chisels benefit from having steeper bevel to be able to rotate in shallow mortises).

One of the things that's often said is that people cannot efficiently waste half blinds. If you can do two things:

1) lower the effort that it takes to waste out a half blind in strikes

2) improve the behavior of the chisel so these strikes remain uniform and in this case, don't involve bouncing out of the end grain cuts (so that is no reason to pause cuts), then you may find that it's no faster to rout these out than it is to chop them out.

I'll explain this briefly in the article, but good behavior across the grain in any cut is a good thing, even if it's just things like pushing the marked line back less or making the small number of cuts that one would make trimming the remaining bits off of sawn dovetails.

I have no idea what warren will attest to, but one of the things that brings a hand tooler to not explore faster ways to remove waste is the ability to keep tool in hand working evenly and at rhythm.

Finally, if I propose a method of modifying edges that improves perceived sharpness, but increases effort to do a similar volume of work, that's a trade off. we don't want a trade off if we don't have to have one. If this method improves edge holding, shortens sharpening cycle time, makes good results easier to get and then results in less effort chiseling, then we have a unicorn instead of a different colored horse.

Re: Pictures from Testing for Article

#9

Re: penetration

David Weaver

two things:

1) I can feel a difference in behavior between V11 and the others, no matter how many times I do the test. It slips through wood more easily. I can notice much less friction when planing with V11. I'm not sure why there would be less friction with the wood sliding across the metal planing, but somehow not less when wedging through wood. I have no idea why anyone would study this on a stainless steel made for knives, kitchen equipment and scissors when it's not a high speed steel, and is relatively new.

It can be felt both on planes and chisels, though.

2) The chisel strikes are important and are explained in the test summary. If I mention there is a decrease in effort needed chiseling due to this method, it's important. Perhaps not to all, but it heads off the accusation I'd immediately suppose if I weren't the one doing the test ("OK, you've made the edge more blunt, and it holds up well, so will any blunt edge. It'll just take more work to use"). It doesn't, and that's important. The implications are important both for malleting and paring. if you aren't sure why they are for paring, set up two mock tenons about 3" long and saw leaving waste above a knife line. Unicorn a chisel, then pare the waste off. Then use a chisel that you would otherwise use with a flat bevel and pare the waste off to the marked line. Note the effort needed and look at the results in combination.

All of these are important, not just edge damage. Edge damage is the biggest issue most people face, because you can hardly worry about the effort question if you can't keep a chisel sharp in the first place.

Here's an instance of why, though - my real life experience. I found years ago that the difference for flat bevel edge holding between a mediocre chisel and a really good one is usually about 2 degrees. So the tests that are out there in droves comparing chisels at 30 degrees are all fine and good, but most of the tests would've been well served by adding 2 degrees to the underperformers and noting that little of the damage incurred continued with two more degrees.

This was one of my ongoing irks with destructive standardized chisel tests. We got blanket accusations with some chisels appearing to hold up perfectly and others lasting a tenth as long in use. Realistically, anyone with nerve endings would've explored accommodating the chisels that were a little less good to see if the accommodations left them usable.

This is just my opinion, but if a chisel fails at 30 and fails a 20th as much at 32, then that's tolerable. For some reason, when you start getting toward 34 or above, using the chisel becomes intolerable if you have another better option at hand. Even for really heavy use like mortising a plane body - instead of shearing chips out, they're broken off, the chisel bounces in the cut a lot more instead of penetrating and the waste flies across the shop, and since I was standing toward this test, the broken off shards of maple actually hit me in the face twice. I'm not sure how they managed to get up to my face, but they sting like gravel off of a string trimmer.

Re: Pictures from Testing for Article

#10

I'd have to measure the maple...

David Weaver

..which would then be difficult because at some number of strikes, I'd be partially through a cut.

Re: Pictures from Testing for Article

#11

the test bed

David Weaver

stick of maple thickness and ripped/planed to width.

One inch for each test iteration - backed by very soft fir plywood that challenges no edge and even planes without damaging edges (I can't even see the glue in it - it's ideal in a test like this).

To do the test, I strike the chisel in the middle of the inch section to make it easier to see how much I'm chopping at a time and then tackle each half in five or 6 leaves.


Re: Pictures from Testing for Article

#12

practical chiseling

Bill Tindall, E.Tn.


Replicate of dovetail from ash desk drawer. 14 whacks to make it. First 4 were dainty to establish shoulder at base line. Remaining 10 were progressively harder as the baseline provided more support. 1/2" total depth, one half that of your experiment if I understand it correctly.

If I saw up front that it took 100 whacks to remove a total of 1" length of dovetail waste I would dismiss anything that followed as too far removed from practice to be relevant.

As an aside to the on-lookers, over sawing half blinds is common in "production" Period Furniture. I never did it before this trial. Wow, is it fast. This dovetail took a fraction of a minute and done. I am astonished, but I suppose I should not be.

Re: Pictures from Testing for Article

#13

Now I am more confused

Bill Tindall, E.Tn.

The 1 inch is the amount of maple you turned into chips, not the total length the chisel chopped? If correct, that is not the metric I would find most useful. Rather how many inches of maple did the chisel drive through. This variable is what dulls.

Now we have another convoluting variable on number of whacks, chip thickness.

Re: Pictures from Testing for Article

#14

Bill - Friction

David Weaver

Will likely bore others.

https://www.engineersedge.com/coeffients_of_friction.htm

I sent an email to bill and supposed that the friction of high hardness steel feels like it's less than lower hardness, and that I suspect also that V11 at the same edge and wedging angles above the edge moves through wood more easily.

I don't know the V11's exact hardness, but we have a pretty good idea of the chromium content from Patrick Chase XRF-ing the steel and posting it for all to see.

Safe to say, it probably won't rust.

I also did a very brief search for studies of this, but I didn't find anything based on hardness. What I found was someones' thesis that steel's friction on wood is lower based on *potential* hardness. As in, unhardened tool steel has less friction across wood than does mild steel.

I have actually had the benefit of discerning this having made an O1 plane and a mild steel infill at the same time. The O1 plane has less friction when you forget to wax it, and the abruptness of the wax running out isn't as stark.

Cast iron planes have less friction than mild steel planes ( in same-weight infills). Subjectively, the O1 infills and cast infills have about the same friction on wood. Bronze LN planes are also much more sticky than cast, and i'm holding off looking at bronze on that engineer's page until after I type this so that I don't get biased by it. But, I'll bet the coefficient of friction of bronze is close to mild steel.

So, I can't say that it's so much the hardness as what the thesis described as potential hardness, but the japanese chisel and the V11 chisel are easier through the cut, and it's noticeable and even affects the quality of the leaf of maple coming off for some reason. japanese and V11 chisels are less likely to fracture the shaving than the others even when identically set up.

I can't imagine anyone studying this, but when you use one right after another, then you find things you didn't expect to find, but they can be felt very easily. In isolation with large gaps between uses, they may not be quite so easy to notice.

Re: Pictures from Testing for Article

#15

you're in the weeds. here's why...

David Weaver

maple doesn't mallet as easily as ash. Ash and beech are not far from maple in hardness, but maple is more resistant to chiseling and planing and the quality of wood is far different.

If your dovetails are 1/2 of an inch deep or square...let's assume the chopping might be a fifth of a cubic inch of wood. You would be doing about 1/5th of the amount of wood volume that I chopped in a wood that's much easier to chop. (1/2 inch cubed would be 1/8th of a cubic inch, i'll bump that up a little bit, but it looks like less than a fifth of a cubic inch in your dovetails).

If that's still puzzling, find yourself a piece of maple around 7/8ths thick, rip it to 1 3/8" wide and then chop an inch of length off of it in leaves about 1/10th to 1/12th thick.

If I give you a fifth of a cubic inch in ash, and no other variables mattered (penetration in wider wood will be more difficult than narrower), you'd have around 80 strikes to remove the same volume of wood as I malleted off. But your volume is probably a little bit less than that and I'm limited to a motion with the mallet that's repeatable (rotating the wrist instead of taking a full whack). Subjectively, I can do this swiftly enough with the verawood mallet that it's reasonable. In fact, it appears to be in line with how hard you're hitting your chisel on average. I would normally mallet the wood a little bit harder, but not too much harder as i'm not looking to knock the edge off of the chisel.

In short, this test is directly applicable to what you're doing. I chose maple because it doesn't have much variance in it and it's not a contaminated (sands mineral bits, which I didn't find) or absurd (malleting verawood would've been less than representative for real work).

Re: Pictures from Testing for Article

#16

Re: Bill - Friction

Bill Tindall, E.Tn.

You don't need to find data on an obscure steel to confirm your theory. If your theory on the chromium is correct you may be able to find something on 304 stainless steel, the most common stainless out there. The LV steel is about 17% chromium, SS 304 18-20%.

Re: Pictures from Testing for Article

#17

I can't get reasonable profile pictures....

David Weaver

and the test is over, and all edges were sharpened off at some point.

I'm writing in the document that I'm sure that the V11 can be accommodated ,but I didn't explore that for the scope of this test.

I did intentionally buff it twice as long, and in the guide, I gave i twice as many strokes on the flat bevels on diamonds to get to the same point. If you just buff it for five seconds, not much is done to it -it even still retains a bit of wire edge. Doubled, and the visual result is the same.

But as you've proved, it could be accommodated.

Re: Pictures from Testing for Article

#18

The trouble with 304 stainless

David Weaver

is that it needs to be compared to mild steel. The sliding friction of hard steel vs. soft steel is well established. Static friction is equal, but sliding friction of hard steel is less.

Top page 2-44.

But steel on wood, I don't know.

What we need to actually know is less about a low carbon stainless vs. mild stainless, and exactly what I am perceiving -V11 vs. O1 at similar hardness.

These are things I'm supposing, though, not things I'm looking to prove. The difference between the way V11 and the others behave is stark. You could almost linearly compare the feel of sorby/mk2/japanese

This is not that important of a thing except to explain why the japanese 32 degree test resulted in less effort than sorby at unicorn.

The article will have a fairly compact statement about what the strikes represent and how to interpret them, though. I just posted the table and let you guys fish.

Re: Pictures from Testing for Article

#19

Friction and surface finish

David Weaver

this isn't really a question I need to answer - but the difference in slickness through the cut for the japanese chisel and the V11 chisel is noticeable. I was surprised by that.

I saw a study from japan (where stamping is concerned) that discussed sliding friction and said they found no difference between similar stainless and non-stainless steels (similar hardness, etc), but that surface finish could generate a large difference. That's not a surprise, but i don't know the context there.

If it is surface finish instead of alloy that creates the difference, it's true that both the japanese chisel and the V11 chisel sharpen more slowly (even on diamonds) than the other two chisels and the perception of edge fineness from them may be due to the fact that the surface finish is finer with the same abrasive.

It's less important why what "is, is", and from the perspective of mine as a tester, more important how stark the difference is after very careful setup of each chisel to count strokes on the diamond hone and check bevel size, etc for the microbevel. I went so far as to hollow grind the japanese chisel instead of leaving the bevel flat to make sure that the part of the chisel in the cut was the same.

This answers a question that I don't think is that important and if you're just focused on edge holding, you wouldn't care about. That is, if cheap chisels perform nearly the same as expensive ones on a test like this, is it worth it for the cost-no-object individual to buy a $113 with tax 1 inch chisel, or do something strange, like troll japan's version of ebay looking for just the right japanese chisel (this can be done on a much lower budget, but it takes time and there's some risk of getting junk).

The answer using one chisel right after another is that even with this neutralizing method, the sweetness in the cut of the V11 and japanese chisels is better. If you're just knocking out dovetails and listening to MLB or something, then this kind of sweetness may be of little interest.

Just for the purpose of this test, too, the LV and japanese chisels create those large leaves of maple more easily, and then the tendency to produce this:


is less. this is the remnants left on the board when a chisel either starts too steep or gets damaged. They fly, and they will literally hit every part of you, including the face. With the leaves, the chisel just stays in.

With ash, beech and good cherry, I think the leaves are always nicer, but with maple, they're not.

When you work only with hand tools, one of the things that you chase is getting the chisels to work just right, the same way that a double iron plane neatly ejects its payload while a single iron plane will clog and hopefully unclog on the next shaving - it's an issue of dominance for someone who may plane a few ten thousand feet in a given day.

I'll explain a conclusion of all of this in the article in far less detail, but it's simply this. The proportion of strikes probably tells about half the story. If one method, netural of edge damage, takes 20% more strikes, the trouble to do a set amount of work is probably double that more and not 20% because of all of the little nuances that occur when a tool isn't working quite as well.

Re: Pictures from Testing for Article

#20

More on practical chiseling

Bill Tindall, E.Tn.

Summary- harder wood doesn't take more strikes. Higher resistance chisels don't take more strikes. Why, chops are not made at equal force. Force is proportioned to achieve a desired penetration. Time to conclusion winds up the same. Total work, not significant for these tasks. Total time, significant and equivalent.

Chopped an equivalent dovetail in hard maple. 6 dainty chops to establish initial baseline. 10 more to complete the 1/2" depth. (recall it was 4 and 10 in ash)

It may be that I don't chop out waste like anyone else, but don't put a lot of money on that bet. I have chopped many feet of dovetails on cases and drawers in the last decade. I'm lazy and look for efficiency where ever I can find it. So do some others.

I did not do pine or poplar but considering what I was using for feedback I believe the number of chops would have been about the same.

I don't chop with the same force each time. At first the force grows proportionally and depends on the support the base line shoulder provides. The deeper the cut, the taller the base line, the better the support to drive the chisel deeper. But, at some depth the deflection becomes the limiting factor. I found myself bashing to get depth that would not result in unacceptable deflection. The bash force was adjusted by this feed back, not wood resistance, which was easy to overcome with modest increase in bash force. (I bash with a small rubber mallet, lighter in weight than what many people use. )

While David's experiment was convincing for comparing edge durability, it would not predict how I would rank the chisels in how many strikes (and in turn how much time) it would take to chop out a dovetail waste. I would find them all the same with no concern for the small difference in force exerted.

That said, a more acute bevel angle would result in less deflection. I think I would respond to that fact with deeper blows resulting in faster production. For my application, more durability and better penetration are selling points. The number of bashes to get across 1" is not relevant to the way I work. These data would confuse the attributes I find important.

Now you can decide how strange I am and adjust emphasis accordingly. I don't have a horse in the race from here out. I know what Unicorning will do for me in putting product out the door and it is significant.

Re: Pictures from Testing for Article

#21

one of the things that...

David Weaver

i had to quickly remind myself of when doing this test was not to adjust the mallet strikes to be anything other than equal.

I would say on something like a 5/8" wide half blind in cherry , it would be not that common to strike a chisel more than 3 or 4 times per leaf. This test in maple is far different. It's much wider, it's a much more resistant wood, plus the width, and on top of that, I need to be consistent. That means if the stick is one small blow away from going through, I have to hit it anyway and let it go into the substrate.

We make all kinds of adjustments while we're working - i tend to make adjustments that make wasting half blinds the fastest, which is working as fast as the chisel will tolerate without taking unnecessary damage. As long as the drawer front or case top being half blound (how's that for a word?) will tolerate it.

This test is intended to be something that provides a material method. It's not intended to have randomness so that materiality can't be discerned, and it's also not created so as to be so precise that you can't tell if you need such a level of precision to get the material differences. I'm comfortable with it. The only thing I wish it did was test the strength of corners on chisels, as low bevel limit pushers will lose corners on chisels. They crumble off on soft chisels, and can be lost spectacularly on overhard chisels, especially when the angle is too shallow.

Removing material with side wood is different than released wood or just an end less than full width here, and I didn't want the extra variables. This is literally just to tell if we're gaining cutting ease, losing it, or staying the same. It does that.

(i recall hearing that folks who worked in factories basically cut drawer joints with the same number of blows for every step - that becomes important if you're going to start a rhythm and never stop to think, only adjust mid work. Most people don't do that. Still, these strikes are pretty brisk and you'd be fighting other things if you tried to go much harder - such as the chisel diving inward during the cut).

Re: Pictures from Testing for Article

#22

Re: one of the things that...

Warren in Lancaster, PA

Absolutely, David. There is no gain in hitting ("bashing","mashing") the chisel so hard that there is damage, so we have to get a feel for what the chisel can stand and stay within that safe limit.

Re: Pictures from Testing for Article

#23

The  efficacy of Davis's discovery

Bill Tindall, E.Tn.

I can bash according to what the wood will endure and not be concerned with the chisel damage. It removes one more concern in getting a drawer made.

To put it more delicately, I can feel what the chisel is doing that is bad- crowding the baseline or deflecting from the lateral force on the bevel- and adjust the mallet force to approach these limits without exceeding them, and not be limited by steel properties. The wood controls my work not the cost of the chisel.

(I prefer the bashing version.)

The primary factors governing these forces are probably wood properties and chisel bevel. I remain skeptical frictional differences are significant.

Re: Pictures from Testing for Article

#24

Bash away, bash away all!

Jim Matthews

Bash is the more broadly acceptable technical term.

Most of the others have been adopted by Rappers to construe other meanings.

Re: Pictures from Testing for Article

#25

Re: Bash away, bash away all!

Bill Tindall, E.Tn.

My goal is to make woodworking sound accessible. Fiddly could convince someone to take up horseshoes where close enough counts. Everyone can bash. I believe anyone can make stuff that tries, including dovetails.