Hardness testing of saw plates

[Rob Streeper]

I just posted a comment on an old link regarding some hardness testing I have done and one of the members suggested I create a new thread with the information.

I've been building saws for about two years now and have become interested in the technical aspects of saw design. In my reading I have found that Disston at some point or other decided that the optimal hardness for saw plates was Rockwell C 52. The saw building community uses blued spring steel in sheet or rolled sheet form, often 1095 in building custom saws for the boutique market. My nature is to question things, especially benchmarks that were established a century ago like Disstons Rc52, and thus I set out to learn about the hardness of saw plates.
I recently acquired an Ames 2-S hardness tester. This instrument measures superficial hardness on the Rockwell N and T scales. This device works by measuring the depth made on the test article when a measured load is applied to a point called an indenter. On the N scale it uses a specially shaped diamond point and on the T scale it uses a 1/16" carbide ball. For hard materials you use the diamond and for softer you use the ball. The N and T scales specify three different loadings, 15, 30 and 45kg for different thicknesses of test materials. This links to a table giving the appropriate loading and indenter for various test article thicknesses. http://www.continentalht.com/userfiles/file/CHT-hardness-conversion-chart.pdf See the Hardness vs Minimum Thickness Chart at the top left of the first page.

Here's the instrument.

304020


And here's the data.

304021

The conclusions I draw are:

1) 1095 steel is a consistent product. Despite the fact that the test materials were purchased from different sources over a period of several years the hardness values for the products were all very similar with a remarkably low standard deviation of 0.34 Rc points. Plus these tight results show you that both my technique and my hardness tester are working well.

2) Of the custom makers, only one is using 1095 steel. The other two makers represented in my collection are using softer material. Maker 3 is using very soft metal, 40% softer than 1095, - likely because it's cheaper and/or easier to work with. Maker 2 is using something that is about 7% softer. The highly significant t values (t < 0.05) for these measurements indicate a high degree of confidence that the measurements are not due to chance.
Maker 2's saws work just fine for me, but I'm not using them for production work. Makers 2 and 3 are names that everybody in this area knows well so I won't reveal them.

3) Disston's products varied in hardness over the years. The Disston saws in my collection were manufactured in the period spanning post WW1 to post WW2. The variability is not too surprising because the concepts of industrial quality control were in their infancy in the early 20th century. Disston also used steel that is harder by just under 4%, or 2 Rc points, than is 1095 but the difference is on the edge of statistical significance (t = 0.04).

Sharpening old Disston saws is going to be harder on your files than is sharpening modern saws. If you're really worried about the longevity of your files get some saws from Maker 3 and your files will last a long time but you may not get much wood cut.

Further, modern custom saws made with raw 1095 steel are not as hard as are older Disston saws. If the plate hardness is important and if Disston optimized his products we've actually taken a step backward in our unquestioning use of 1095.

I'm gonna go lick my wounds now.

[David Dalzell]

would there be any way to find out who is "custom maker 3"? I am about to buy several backsaws and would rather have harder than softer steel.

[Rob Streeper]

would there be any way to find out who is "custom maker 3"? I am about to buy several backsaws and would rather have harder than softer steel.
Hi David,

Sorry, I can't tell. I'm sure you understand that I would incur significant liability if I revealed that information and it's likely that the Sawmill Creek administrators wouldn't be too happy either. I debated with myself the wisdom of making this post but ultimately decided that I would do a disservice to the community if I remained silent so I settled on the compromise of anonymizing the names of the makers.

I'd suggest that you ask the saw makers you are considering about the steel that they use and whether they test each lot of their stock for fitness for use.

Given the significant investment that some of us make in these custom made saws I don't think it would be a bad idea to invest in a hardness tester for personal use. I have a lot of money in my saw collection and it really disappoints me to learn that I have been taken advantage of.

BTW, I have some chisels I'm suspicious of too...

Caveat emptor

Cheers,
Rob

[David Dalzell]

I figured that would be the case. Everyone seems sue happy these days. I did enjoy your article.

David

[Rob Streeper]

Thanks David, sorry I can't be more forthcoming. I'll be continuing my saw blade hardness work.

Cheers,
Rob

[Dale Schafer]

I'm not in the USA so not sure on the legal ways over there. But surely reporting facts i.e this saw from this maker tested at…. Can't be liable???

How often do people writing reviews that are condemning or critical get sued? If they do and that leads to all independent reviews only commenting on the positive then the reviews are worthless, even if it is just a subconscious bias on the basis of how they word it to not offend.

What happened to free speech?

I say to all manufacturers stop using scare tactics and make a quality product where you have nothing to hide and can take pride in your product. That would make for a much better world.

[Dale Schafer]

Rob, Have you had a chance to test the new breed of production saws i.e. LN & LV or just the custom makers?

[Rob Streeper]

Hi David,

People are getting sued by companies here for writing negative comments on the Internet. See: http://www.forbes.com/sites/investopedia/2013/01/25/can-you-be-sued-if-you-give-a-bad-review-on-yelp/

Speech is still free in America as long as nobody objects to what you have to say - that's why I have to be careful.

As to Maker 3 and the results I got, who knows. Maybe it's a one-off problem. Maybe one lot of steel was unevenly hardened.

[Rob Streeper]

I put anybody who charges more than $75 or so for a hand saw into the 'custom makers' category.

[Rob Streeper]

I realized over dinner that I have saws from an additional custom maker so I've updated the data table to include tests on their saw too. I also included measurements of the blade of an Atkins No. 26 pattern makers saw having a Silver Steel blade.Their saw is at the extreme upper end of hardness, Rc 61.4. Special type of saw. I'll enter some more data tomorrow as I have a few more Disstons to test.
304047

[george wilson]

I don't know what happened to my previous post. There are 3 different N and T scales. I only ever used the Rockwell C scale,in which the usual 1095 spring steel that everyone uses was a consistent 52 Rc. Much harder than that and it is impossible to file the saws,or to keep the blades from breaking in use because of being too brittle.

I don't understand your figures,therefore.



These different scales are for specific steels and materials,too. You must use a scale that is intended for the type of steel you are testing. That,to be brief,is what my post was about. But,I'm too tired now to investigate further.

I have not heard of anything but the C scale being commonly used to test spring steel. And,I have never seen an old saw that was harder than modern 1095 spring steel. That was just my own experience,of course,and I am not a collector of vintage saws,but have filed many.

I am wondering how you have arrived at a hardness in the C scale of an Atkins saw at a bit over 61 RC scale. That would be totally impossible to file. What "special saw" would that have been?

[Rob Streeper]

Hi George,
I think your response is on the original thread. I started a new one at the suggestion of another member.
For these measurements I referred to this: http://qs-hardnesstester.com/hardness-minimum-thickness.html to select the appropriate scale and loading.
For most of these measurements I used the 30N scale. For the 0.015" 1095 I used the 15N scale even though it is somewhat below the minimum thickness recommended.
The consistency of my measurements and the low standard deviation for the 1095 measurements indicate first that 1095 is very consistent material, second that I'm using the tester correctly and finally that the tester is operating correctly. The measurements on the test blocks included in the tester kit were spot-on despite the fact that the instrument was made in 1969. The low number of test marks on the blocks suggest that it spent most of the intervening years sitting on a shelf somewhere. Too bad as it's a beautiful instrument.

Cheers,
Rob

[george wilson]

What country are you in that you cannot publish correct data on saw steels? Fine Woodworking does tests upon different tools such as chisels BY BRAND NAME,and has not been sued.

I haven't investigated the odd scales you are using,but the words "superficial hardness",when the saw steel is not superficially hardened bother me,as does going below the minimum allowable thickness. Those things just MIGHT lead to inaccurate data.

At work I had the cadillac of hardness testers,the Versitron,to get my results from. Had the test blocks and a new diamond also.

[Rob Streeper]

Here is a link to the company page for my tester. The Rockwell N and T scales are also known as 'superficial' hardness scales and the testers are superficial hardness testers. They are used for thin hard materials such as thin hard steels and case hardened surfaces.
People get sued all of the time for writing critical reviews, even in Virginia. See: http://www.forbes.com/sites/investopedia/2013/01/25/can-you-be-sued-if-you-give-a-bad-review-on-yelp/. Poor lady is being sued for $750,000.
Atkins used it's so-called Silver Steel in it's saws. See: http://www.tias.com/8600/PictPage/1922983474.html. I have no idea what 'Silver Steel' is but it is sure hard stuff.

[Jim Matthews]

This is excellent.

Thank you for taking time to compile this.
It's a terrific piece of work.

This sort of essential data helps explain
lots of things I notice in my hodge-podge
collection of rehabilitated saws.

I congratulate you on this article.

[Dale Schafer]

I have heard the term silver steel used by old timers in reference to drill rod so my understanding is it would be a grade of tool steel.

[Stewie Simpson]

Interesting article on Atkins Saws.

Quote:

He went to England to have this steel made, and was told that the price would be so high the woodcutters of the United States could not buy the saws.

“That is the kind of steel razors are made of,” the Englishmen said.

“Then that is the kind of steel for Atkins saws.” the Hoosier answered. Mr. Atkins named the metal “silver steel” and during his life, as is the case with the Atkins company now, it was his boast that no other saw metal has been made to equal it.

http://blog.lostartpress.com/2013/07/05/selling-atkins-saws/

[george wilson]

Please read this whole post to make sense of it. Silver steel is a term coined by the British to mean carbon steel,even MILD steel. Wrought iron is grey. Even low carbon mild steel is silver color compared to it. ANY steel is silver steel. Even concrete reinforcing rods.

So,the term silver steel was,and sometimes is still applied to any carbon content of steel,from mild steel,to hardenable W1 drill rod. Colt revolvers had a plant in London. His 1862 Army model had slender proportions compared to earlier pistols because they were the first he made from the stronger and more reliable(against bursting open) mild steel. The British dubbed them as made of "silver steel".

I am sorry,but,with all due respect, I must say that I find the study here faulty and misleading. The OP has not used the proper testing device,or the proper hardness scales. You cannot get the proper answers using a device intended for SUPERFICIALLY hard metals. Saw steel is NOT superficially hard. It's hard all the way through,which throws the penetrator way off.

Old saws almost never exceed the 40's in the Rockwell C scale,which is the proper scale for testing steel such as spring steel used in them. One saw user has mentioned to me that he had a DEFECTIVE Simmons saw that was not tempered properly,and was left with a straw colored blade,which indicated a very low tempering color. Straw color means it was heated to about 400º. It had never been filed and the teeth were worn to a rounded state. The saw was impossible to file with the very best of USA made files. He gave up on it. It was in the 58 Rockwell C range. It had been tempered in the 400º range instead of the 750º range,which would have softened it to a fileable hardness. Such things happen. Things get out of factories that are sometimes defective,as most of us should well know.

That(58) range is how hard GOOD chisels are made. These days some are made softer to avoid liability from chips flying into someone's eye while prying open a can of paint,stupidly. Try filing a GOOD chisel. You will ruin your file.

Since the term "silver steel" was applied to ANY grade of carbon steel,it could mean carbon steel of a lower carbon content(.18%-.20% for MILD steel that would not harden),to higher percentage used to make saws( .70 to .95% carbon),or an even higher carbon content( over 1%) used to make razors and files. It was a salesman's ploy using confused semantics. Other saws were made of the SAME silver steel.

Lately I have seen on television,an ad that states that ROLLED STEEL makes GM pickup trucks superior. Well, ROLLED STEEL has been used since the late 18th. C. to make everything from washing machines to locomotives. Cars are nearly ALL made of ROLLED STEEL!(Except a few like Saturn,made with plastic body parts,or aluminum cars like old Jaguars).That is a silly ploy used on the average customer,who has no idea about steel. The same old ploy as Atkins used,basically.

[george wilson]

The lady in Virginia who is being sued falls into an entirely different situation from scientific data such as Fine Woodworking has used in testing BRAND NAME chisels and their performance. They named Sears,Barr,Pfiel,and several other chisel brands.They have not been sued.

She claims she was charged for work not done,and her jewelry was stolen. How is she going to be able to prove these claims? How does she know WHO might have stolen her jewelry? How specific was the contract she signed to have the work done? That contract is the all important feature of her case. And,if it was not very specific,she has NO CASE,and is open to being sued.

When I had my shop building built,the contract only said they were going to build me a building. That was IT,repeat IT. I furnished drawings I made. They kept getting into my house and using my shovels,breaking the handle off one,and leaving my tools out in the rain. We were not living in the house yet,but my wife and I were using the tools to fix up the interior of the house. It was necessary to leave it open during the day when I was at work. I had a devil of a time getting them to FOLLOW THE DRAWINGS. But,I prevailed.

I just forgot about the broken shovel,and the rusty tools. They had no business using my tools,but it wasn't worth a law suit. In the end,I was just happy that the building got built. I HATE using contractors!!!!!

You just cannot compare these issues with scientific data PROPERLY gathered and made into an article. But,you need to do it PROPERLY.

[Stewie Simpson]

Hi George. I don't disagree that it was a marketing ploy by Atkins. The following article seems to validate the Disston was also using similar sales tactics.

Stewie;

http://www.disstonianinstitute.com/steel.html

[Mel Fulks]

I have seen planer knives marked "silver steel" that were just semi high speed steel. Low grade stuff, but what a beautiful name... SOUNDS better than M2 or T1.

[george wilson]

The fact is,Stewie,I misread your post at first,and then didn't have time to change my reply. I still don't. Cooking a late breakfast.I stay up very late. Sorry. Deleted it.

[Rob Streeper]

I am sorry,but,with all due respect, I must say that I find the study here faulty and misleading. The OP has not used the proper testing device,or the proper hardness scales. You cannot get the proper answers using a device intended for SUPERFICIALLY hard metals. Saw steel is NOT superficially hard. It's hard all the way through,which throws the penetrator way off.



Hi again George, In regard to your assertion of my ignorance I refer you to this: http://www.nist.gov/manuscript-publication-search.cfm?pub_id=853006. This publication is the reference I use for selecting and performing these types of measurements. For those who don't know, The National Institute of Standards and Technology is the national standards reference agency/laboratory for the United States. NIST sets standards that are used worldwide, i.e. they are THE authority in this area. Their mission is to: Promote U.S. innovation and industrial competitiveness by advancing measurement science, standards and technology in ways that enhance economic security and improve our quality of life.

I have abstracted this from the above referenced document on Rockwell testing.304082

[george wilson]

I have had my say,Rob. That's all I can do. I had a proper Versitron hardness tester at work that cost several thousand dollars. I tested every tool I own,nearly. All my saws and chisels have the little punch mark on them. 1095 saws checked out very consistently at 52 RC. RC is important. It is the CORRECT scale to use for hardened steels or DEEP case hardened. And,deep means greater than 1/16". Chisels,like my 1960's Marples checked out at 58 to 60 RC. And,you can't file them. Old saws stayed in the 40's.

Use of a scale that is for superficial hardness just isn't going to give proper readings. Period.

I will add that a cracked diamond will give erratic readings. You might not be able to see a crack. I had to replace the diamond in my tester for $120.00 when it became erratic.

I don't mean to be mean or confrontational here. I just am giving the benefits of my many,many years as a professional tool maker and heat treater. When I turn loose information for general consumption,I want it to be as accurate as I can make it,and proven by actual experiences multiple times.

Look carefully at your chart and see what materials are intended to be tested with your S and N scales. THIN case hardened. That can mean only a few thou. deep. Not through hardened. You already admitted that in one instance you tested .015" steel,which was less than the minimum thickness required. Those errors will simply not give true readings.

[Stewie Simpson]

The fact is,Stewie,I misread your post at first,and then didn't have time to change my reply. I still don't. Cooking a late breakfast.I stay up very late. Sorry. Deleted it.

Fair enough George. Take care of yourself.

Stewie;

[george wilson]

By the way,Stewie,you look like you need a haircut and a shave!!:)

[Rob Streeper]

Well George, what can I say. You seem to think you're correct and NIST is wrong.

[Stewie Simpson]

By the way,Stewie,you look like you need a haircut and a shave!!:)

Your right there George. I could also do with a lighter sun tan as well. ;)

[ian maybury]

There quite likely was as George says a much wider historical usage of the term silver steel, but it surfaces these days in the UK and environs as a very specific form of brightly finished ground rod as described here - so it probably has specific applications for production of easily hardened parts: http://en.wikipedia.org/wiki/Silver_steel

It's a very definite designation, the subject of a British Standard. It hardens to 64 HRC according to the link. It's quite a high carbon tool steel (0.95 - 1.25%), with a decent amount of chromium, manganese, silicon and vanadium in there too. e.g. http://www.westyorkssteel.com/tool-steel/silver-steel-bar/ It's possibly not a lot different in some ways to O1 steel - has a bit more carbon, a bit less manganese, similar chromium, a bit less vanadium, a little silicon and no tungsten - but i don't know enough to know what the differences are likely to amount to.

It's a very common tool room supply item. I don't know enough about the metallurgy or usage in practice to say much about its detail characteristics, but personal experience says that even in the annealed form in which it's normally supplied it's quite awkward to handle. e.g. it doesn't thread well, because bits chip out, and it's likewise not keen on being more than moderately bent - it tends to crack.

[David Weaver]

Well George, what can I say. You seem to think you're correct and NIST is wrong.

I don't post much here any more (because like this thread, the discussion doesn't advance and there isn't proper credibility given to the people who actually know what they're talking about), but I'd like to point out that George is correct and you're wrong. You need to test the saws on something like a versitron (or trade for some time on one) if you want to assert their hardness. Right now, you're just wasting your time.

I had the saw that was light straw temper, it couldn't be filed. Period. The very tips of the factory teeth cracked off because it was overhard and then what remained got rounded over (worn) by some tough handed but weak minded individual and that is how the saw stayed until I got it. It was a disston 7 (not a metal cutting saw or anything, just a defective 8 point crosscut saw - maybe someone pulled it off the line as a spoof before it got tempered to spring temper).

There is little discernible difference between old saws brand to brand and 1095 at 52 hardness is harder than any old saw that I've filed (save the one), except those old enough to have been hardened inconsistently (where most of the saw is fairly soft and there are surprise hard spots that kill files). There are not many saws like that in use, and they don't say atkins on them.

Silver steel is a marketing term. It means nothing at this point other than in old things, it was a generic term for high carbon water hardening steel, commonly used on razors (I guess it assured the buyer they weren't being slipped medium carbon steel in a low quality good). Only the original alloy coined by some brit had silver in, and then not that much (silver does nothing for the steel) - nothing you find modern enough to use will be any different than any other generic water hardening high carbon steel. Not coincidentally, every saw made by disston and every other maker at the time was high carbon water hardening steel. When specialty steel was actually used, the mill and the alloy were specified (see schoeller Bleckmann razors, or "phoenix steel").

If those saws have ever seen a file that left marks in them on all of the teeth, they are not high 50s or 60 hardness, and not even mid 50s. They would've been thrown aside/away in favor of a saw that was easier to file. Not to mention, at mid to high 50s, spring temper is gone and the teeth would snap at set.

Most of the old saws I've filed are probably mid to high 40s in hardness.

Find yourself a chisel that's 60 hardness, stone the bevel off until it's about saw width, and then file five full depth teeth in it with a saw file and let us know how it goes.

I don't think the NIST is wrong, I think you're missing the difference between theory and practice and you need to give your saws to someone who can test them competently.

[Rob Streeper]

Hi Mel,

The Atkins saw I tested that has the Silver Steel blade is a pattern makers saw. It is very hard but not brittle. It's hardness may be due to the relatively unique design of the saw.

Cheers,
Rob

[David Weaver]

Hi Mel,

It is very hard but not brittle.

if it was 61 or even 58 hardness, you wouldn't be able to flex it. It would break just like high end japanese saw plates break when they are flexed (and those are often yellow steel, which is much like 1095).

Find a thin chisel, bend it (or a knife) and let us know how it turns out. They are heated past spring temper, and most pocket knives are a touch softer than chisels.

What you're perceiving as a magic combination of hardness with ductility is due entirely to the fact that the saw is not actually 61 hardness.

[Brian Holcombe]

I don't post much here any more (because like this thread, the discussion doesn't advance and there isn't proper credibility given to the people who actually know what they're talking about), but I'd like to point out that George is correct and you're wrong. You need to test the saws on something like a versitron (or trade for some time on one) if you want to assert their hardness. Right now, you're just wasting your time.

I had the saw that was light straw temper, it couldn't be filed. Period. The very tips of the factory teeth cracked off because it was overhard and then what remained got rounded over (worn) by some tough handed but weak minded individual and that is how the saw stayed until I got it. It was a disston 7 (not a metal cutting saw or anything, just a defective 8 point crosscut saw - maybe someone pulled it off the line as a spoof before it got tempered to spring temper).

There is little discernible difference between old saws brand to brand and 1095 at 52 hardness is harder than any old saw that I've filed (save the one), except those old enough to have been hardened inconsistently (where most of the saw is fairly soft and there are surprise hard spots that kill files). There are not many saws like that in use, and they don't say atkins on them.

Silver steel is a marketing term. It means nothing at this point other than in old things, it was a generic term for high carbon water hardening steel, commonly used on razors (I guess it assured the buyer they weren't being slipped medium carbon steel in a low quality good). Only the original alloy coined by some brit had silver in, and then not that much (silver does nothing for the steel) - nothing you find modern enough to use will be any different than any other generic water hardening high carbon steel. Not coincidentally, every saw made by disston and every other maker at the time was high carbon water hardening steel. When specialty steel was actually used, the mill and the alloy were specified (see schoeller Bleckmann razors, or "phoenix steel").

If those saws have ever seen a file that left marks in them on all of the teeth, they are not high 50s or 60 hardness, and not even mid 50s. They would've been thrown aside/away in favor of a saw that was easier to file. Not to mention, at mid to high 50s, spring temper is gone and the teeth would snap at set.

Most of the old saws I've filed are probably mid to high 40s in hardness.

Find yourself a chisel that's 60 hardness, stone the bevel off until it's about saw width, and then file five full depth teeth in it with a saw file and let us know how it goes.

I don't think the NIST is wrong, I think you're missing the difference between theory and practice and you need to give your saws to someone who can test them competently.

I was starting to wonder where you were.

PS the Snow white is workin' like a charm.

[Winton Applegate]

Hi gang,
I am having a great time reading this thread and learning.
When I heard the term superfical I thought that might mean case hardened.
At least I was right about one thing today.

The term gray metal . . . that's the English term right?
As far as I know in America black is iron and white is aluminum. As in Black Smith.

For a bit there I was guessing silver steel was stainless; I recall reading in his book that Norm Abrams had a nice stainless steel panel saw that some one stole from him. Any way . . .

George,
I stay up too late as well.

Thanks guys for hammering all this out !
I'll just shut up and learn something.

PS: in this day and age, the last ten years or so, of flagrant incompetence, I wouldn't even be surprised that the Standards Institue (in America) was full of beans.

[Rob Streeper]

if it was 61 or even 58 hardness, you wouldn't be able to flex it. It would break just like high end japanese saw plates break when they are flexed (and those are often yellow steel, which is much like 1095).

Find a thin chisel, bend it (or a knife) and let us know how it turns out. They are heated past spring temper, and most pocket knives are a touch softer than chisels.

What you're perceiving as a magic combination of hardness with ductility is due entirely to the fact that the saw is not actually 61 hardness.

Hi David,

The hardness and brittleness are critically dependent on the alloy used. As I said I don't know what Silver Steel is but the saw and box look like they might have been made in the 50's or 60's. For instance, I tested a Lenox carbide toothed bandsaw blade just now. Thickness is 0.022" and I don't know what alloy is used for the blade body but 5 replicate measurements on the 15 kg and 30 kg N scales indicate that the hardness is 58-60. The blade is of course quite flexible. Another Lenox blade I have, a 14/16 Diemaster II, gives hardness readings in the range of 32-35. This blade is M42 according to Lenox literature. Both are great blades but each is obviously very carefully engineered. The CT blade is designed for heavy duty cutting of large pieces of hard metal. The Diemaster II, which has teeth sharpened into the body of the blade, is also a heavy duty metal cutting blade, especially thinner stock.

So, given that we don't know yet what Silver Steel is I think it best to reserve judgement on the question of whether it can be simultaneously hard and flexible.

Cheers,
Rob

[Patrick McCarthy]

[QUOTE=David Weaver;2359714]I don't post much here any more (because like this thread, the discussion doesn't advance and there isn't proper credibility given to the people who actually know what they're talking about),

DAVID, good to see you post. I for one, would encourage you to keep posting, as I am one of the silent majority out here who has benefitted from your breath of experience. I understand the frustration with getting sucked in (or under) by some of the posters, but I think most out here who appreciate the comments are reticent to say "Great post" and, on the other hand, equally reluctant to do a "Dear dunce head" post to some of the "always certain, but seldom correct" posters. Like everything out here in real life, the vast majority of us are open minded and appreciative of those who answer our questions. Also like the real world, there are also those that seize upon facts and beliefs without the depth of knowledge or experience to really understand the nuances necessary to make some of the blanket statements they utter.

BTW, saying George is correct is kinda like saying the Pope is Catholic. Of course he is correct; not even sure his wife could find fault with his knowledge, skill and experience.

[Rob Streeper]

There quite likely was as George says a much wider historical usage of the term silver steel, but it surfaces these days in the UK and environs as a very specific form of brightly finished ground rod as described here - so it probably has specific applications for production of easily hardened parts: http://en.wikipedia.org/wiki/Silver_steel

It's a very definite designation, the subject of a British Standard. It hardens to 64 HRC according to the link. It's quite a high carbon tool steel (0.95 - 1.25%), with a decent amount of chromium, manganese, silicon and vanadium in there too. e.g. http://www.westyorkssteel.com/tool-steel/silver-steel-bar/ It's possibly not a lot different in some ways to O1 steel - has a bit more carbon, a bit less manganese, similar chromium, a bit less vanadium, a little silicon and no tungsten - but i don't know enough to know what the differences are likely to amount to.

It's a very common tool room supply item. I don't know enough about the metallurgy or usage in practice to say much about its characteristics, but experience says that even in the annealed form in which it's normally supplied it's quite awkward to handle. e.g. it doesn't thread well, because bits chip out, and it's likewise not keen on being more than moderately bent - it tends to crack.

Hi Ian,

Thanks for the additional information. The additional elements, if they represent what Atkins used of course, may explain some of the apparent discrepancy between hardness and flexibility of this particular saw blade. When I say flexible I mean that I can deflect the tip of the saw relative to the centerline of the blade by 50 mm or so and the blade doesn't break.

Cheers,
Rob

[george wilson]

Rob,NIST is not wrong. You are just not using it correctly. You are apparently ignoring the stated metals and conditions of those metals to be tested with the scales your machine has access to. I suggest that you study the chart carefully,in a non emotional way. You have only a specific type of tester,and it is not suitable for the saws you are testing. I am sorry to burst your bubble,but this is factual.

About silver steel again,it means every type of steel that has carbon in it. That includes everything in use. Mild steels still have a silvery appearance compared to grey looking wrought iron.

I have tested many a tool,especially punches, that I made from W1 steel. Fully hardened,it reaches to 67 RC,much harder than HSS,which is normally about 63 RC. These were tested on my Versitron. I don't know where that figure of 64 RC maximum hardness was arrived at for silver steel,of which W1 is a member.

. I always got 67 RC as a maximum hardness.There are many types of HSS,though,so a blanket statement about them cannot be made. Some,like M 42 HSS has a lot of tungsten in it,and the weight and color it has can look a lot like carbide.

After hardening my W1 steel letter or number punches,I had to be careful about tempering them neither too soft nor too hard. I found that heating them to a purple color gave the best balance,so they could stamp a crisp impression onto soft steel,or un hardened tool steels,without shattering or quickly getting dulled. This left them at about 54 RC. Letter and number stamps have a finite life. I have forgotten how many impressions a normal letter stamp is expected to make,but it is in the thousands.

[David Weaver]

It's not 60 hardness. There is no spring steel that is 60 hardness. We will wait for you to find and confirm the error in your tooling or interpretation of how to use it. You are getting far too swept up in trying to get more precise about how your tool was used, but ignoring the fact that the results aren't possible. If they were, everyone would've had saws that hard, and files would've been another 10-15 points harder. I don't know if you don't know enough about metallurgy to understand why there are not 60 hardness or 58 hardness 0.8% carbon water hardening steels or what, but you have made an error at some point and everyone here with enough knowledge of tools, tool steels and the temper of the same knows that you've made an error. You're digging yourself a hole.

Even a razor is only about 60 hardness, possibly a touch harder, and they are thin and would benefit enormously from ductility, but they don't have it because it doesn't exist at that hardness level. Simply tapping a straight razor against a faucet can cause a section to break out of it.

If the saw is from the 1960s, it's just going to be some low dollar mid to mid-high carbon steel that is more highly alloyed. The term spring steel (which originally intended water hardening and probably then later oil hardening carbon steels - used a generic term, often of british origin) was completely defiled later on by adding a lot of chromium or manganese and still using the term.

[Steve Voigt]

There quite likely was as George says a much wider historical usage of the term silver steel, but it surfaces these days in the UK and environs as a very specific form of brightly finished ground rod as described here - so it probably has specific applications for production of easily hardened parts: http://en.wikipedia.org/wiki/Silver_steel

It's a very definite designation, the subject of a British Standard. It hardens to 64 HRC according to the link. It's quite a high carbon tool steel (0.95 - 1.25%), with a decent amount of chromium, manganese, silicon and vanadium in there too. e.g. http://www.westyorkssteel.com/tool-steel/silver-steel-bar/ It's possibly not a lot different in some ways to O1 steel - has a bit more carbon, a bit less manganese, similar chromium, a bit less vanadium, a little silicon and no tungsten - but i don't know enough to know what the differences are likely to amount to.

It's a very common tool room supply item. I don't know enough about the metallurgy or usage in practice to say much about its characteristics, but experience says that even in the annealed form in which it's normally supplied it's quite awkward to handle. e.g. it doesn't thread well, because bits chip out, and it's likewise not keen on being more than moderately bent - it tends to crack.

Ian, thanks for posting this. I was sure I had heard the term, in the specific sense you're using it, bandied about on knife (or maybe razor?)-making forums, but I wasn't aware of the link to west york. .95-1.25% carbon is very high, on a par with Japanese white.

[george wilson]

Steel that is actually sold to be used as spring steel,is generally .80% carbon. Old saws can run between .70 and .80% carbon. Blacksmiths like to use .70% carbon if they are forge welding bits into softer bodies as in making bitted plane irons,or large timber framing chisels. Higher carbon steel catches fire and burns up at welding heat. Since steels were hand processed in the old days,makers were frugal about making tools from solid high carbon steel.

The old saws were made with .70 to .80 carbon steels. They really cannot compare to the boutique saws being made these days from .95% carbon steel. 1095 carbon steels are pushing the boundaries of spring steel. It is not real difficult to break a tooth off while setting teeth on 1095 saws. I had to recut a number of the saws we made for the museum. When they were done,though,they were said to be better than any other saws the craftsmen had ever used. Now,nearly ever small maker is using 1095 spring steel shim stock.

As I have said several times here,steel hardness vs toughness is a balancing act. Gain hardness and you lose toughness and flexibility. You can't have both,except in the case of the Ancient Wootz steel. Their secret was to only reach a red heat in forging a hockey puck size ingot of very high carbon steel into a sword. By not exceeding a red heat,the carbides in the steel were not melted(to form,when they cooled,into large crystals). The smiths crushed the still solid carbides into a fine matrix inside a body of soft iron. That is how they made such superior weapons. We don't do it that way,but ought to. The reason we do not is,high carbon steels at red heat,are VERY hard and a big chore to forge. I have done that myself,making some nearly indestructible engraving tools.

[Rob Streeper]

If I'm wrong George please tell me how, specifically and instruct me on the appropriate Rockwell scale and indenter to use. This instrument and it's test blocks are traceable to NIST standards.
I think everyone who has been critical of my testing is missing the fact that I am the only one so far to present data. If you have data that conflicts with mine please share it.
Superficial testing is specified for thin materials because the plastic deformation induced by the heavier loadings used on the C scale causes the readings to be anomalously low.
Tellingly, the instrument I have came with two hardness test blocks. The testing of the instrument on those blocks returned values that were within specification. Further, my testing of the 1095 specimens shows very tight readings. The average is about a point higher than the steel manufacturer specification but the specified range is within the tolerance of the instrument.
I am open to well intentioned discussion, even if the discussion is contentious but I ask everyone to please leave the personal attacks out.

[David Weaver]

There quite likely was as George says a much wider historical usage of the term silver steel, but it surfaces these days in the UK and environs as a very specific form of brightly finished ground rod as described here - so it probably has specific applications for production of easily hardened parts: http://en.wikipedia.org/wiki/Silver_steel

It's a very definite designation, the subject of a British Standard. It hardens to 64 HRC according to the link. It's quite a high carbon tool steel (0.95 - 1.25%), with a decent amount of chromium, manganese, silicon and vanadium in there too. e.g. http://www.westyorkssteel.com/tool-steel/silver-steel-bar/ It's possibly not a lot different in some ways to O1 steel - has a bit more carbon, a bit less manganese, similar chromium, a bit less vanadium, a little silicon and no tungsten - but i don't know enough to know what the differences are likely to amount to.

It's a very common tool room supply item. I don't know enough about the metallurgy or usage in practice to say much about its characteristics, but experience says that even in the annealed form in which it's normally supplied it's quite awkward to handle. e.g. it doesn't thread well, because bits chip out, and it's likewise not keen on being more than moderately bent - it tends to crack.

i would expect that atkins silver steel probably wouldn't fall in that fairly generous range (which appears to be an industrial standard now, despite being a very large range of carbon that literally occupies three different categories of japanese steels).

I checked around, and there appears to be more than one label that it applies to. Michael Faraday used the term on steel that he put in razors (that had actual silver in it, and then not). Sawmakers used it for steel that had chromium in it because the saw stayed bright longer. They claimed they could make harder saws, too. Perhaps that's true a point or two, i haven't noticed it in saws that I filed, but there may have been more in putting the chromium in for manufacturing - as in good behavior of the steel in hardening and tempering - than necessarily saw characteristics in use or storage. Plus anything you can market as a differentiator in a commodity market is a good thing, it's branding a commodity.

But the fundamental problem still exists with any claim that the saws are substantially harder than any other saw (or that a disston saw is hardness 58), and that's that the saw couldn't be filed and would thus be useless.

It looks like the terminology in the UK applies to a range of things the same way that we would use the term "oil hardening drill rod" (without specifying the carbon level), except where applying it to razors (where it suggests a carbon steel razor that isn't one of the highly alloyed or very high carbon specialty steels)

[Winton Applegate]

Well so much for staying silent.
I think the word flexed and hard enough to break when applied to Japanese high end saws needs a little more massaging for clarification.

What I am saying is you are both right but some where the terms are getting messed up.

There is flexed, a little, and there is bent.

I know that Japanese saws have teeth hard enough to snap off if the filing is too aggressive for the type of wood being cut. Yet every video I see by in the know guys, like Toshio Odate, while handling their saws they can't seem to keep from flexing the saw (some times quite a bit; way past five inches) to feel the spring of the blade. Call is showmanship or compulsion.

No doubt if they were to choke up on the blade and bend it over a short span it would snap or at least crack.

[Rob Streeper]

Well so much for staying silent.
I think the word flexed and hard enough to break when applied to Japanese high end saws needs a little more massaging for clarification.

What I am saying is you are both right but some where the terms are getting messed up.

There is flexed, a little, and there is bent.

I know that Japanese saws have teeth hard enough to snap off if the filing is too aggressive for the type of wood being cut. Yet every video I see my in the know guys, like Toshio Odate, while handling their saws can't seem to keep from flexing the saw (some times quite a bit; way past five inches) to feel the spring of the blade. Call is showmanship or compulsion.

No doubt if they ware to choke up on the blade and bend it over a short span it would snap or at least crack.

So as Clark Griswald says (https://www.youtube.com/watch?v=mk74WprmZxY)

Hi Winston,

One of the saws I tested was a Japanese style saw with a yellow/gold hued blade. Very hard but still flexible enough to be used.

Cheers,
Rob

[george wilson]

Your tester is simply not the correct instrument for testing solid,hard all the way through steel. The Rockwell C scale is the scale to use. Your tester does not do C scale.

I have mentioned this in previous posts.

You are not the only one who has ever posted this type of data here. I have done so over a period of some years.

And yes,Japanese saws can flex. The question is HOW MUCH. Not as much as Western type saws. You can bend the tip of a decent Western saw till the tip touches the handle. It will spring back straight.

Now,I have to go make a punch and die set for my wife's jewelry making business.
'

[David Weaver]

Well so much for staying silent.
I think the word flexed and hard enough to break when applied to Japanese high end saws needs a little more massaging for clarification.

What I am saying is you are both right but some where the terms are getting messed up.

There is flexed, a little, and there is bent.

I know that Japanese saws have teeth hard enough to snap off if the filing is too aggressive for the type of wood being cut. Yet every video I see my in the know guys, like Toshio Odate, while handling their saws can't seem to keep from flexing the saw (some times quite a bit; way past five inches) to feel the spring of the blade. Call is showmanship or compulsion.



Japanese saws are all over the board in terms of hardness. I'd imagine odate knows which saws he's doing that to (the very expensive custom made saw that stan covington sent to me was very tough and could have handled flexing, but it wasn't 60 hardness - it was fileable).

The ones that are above what we normally use for spring temper (they are probably mid to high 50s) will break with a catch in wood.

They're not really comprable to our atkins type saws for two reasons:
* you can't file them with a western file of average quality (not that they're the right shape to begin with)
* you could never begin to put the same amount of flex on them that you get on a jam with a western saw, they'd just snap

[Rob Streeper]

When someone tells you that you're wrong, it's not a personal attack. When someone tells you that it appears that you don't know enough about the application of steels to tools to know that you're wrong, it's not a personal attack.

It just appears to be the case here, and I think everyone with a lot of experience with various tool steels at various tempers is scratching their heads wondering how you think you could create a saw that you can literally take the tip and bend it to the handle and have it be 58 or 61 hardness. It doesn't exist.

David, taking measurements under controlled conditions is what I do professionally. When I take measurements I do it right or I don't do it. Sawblades are not a life or death type of product like jet engine parts so what's the beef? Why so many indictments of my competence, knowledge, methods and so on?
The uniting logical failure of all of the criticisms of my work is that none of them has offered anything but opinion. No contrary references, no data sets, nothing. Another failure is that nobody has been able to tell me with any specificity why I'm wrong. If I'm using the wrong hardness scale please educate me. I have a Rockwell B/C tester in addition to the N/T tester that I've used to generate these data.
If I'm doing this the wrong way prove it to me.

[Winton Applegate]

I respect a lot of David's work and knowledge.
Though
Now that you mention it some of my Japanese pull saws have the yellow (to brown) color left on the blades to "prove" their temper process. Though not right at the teeth so much.
The Western saws are all thicker though that can be backless; they have to be because they are pushed.

I am not saying anything obviously just echoing stuff I read here. And thinking/mulling more like.

Hey . . . give that B/C tester a go and let us know or better yet take the saws to a behemoth hardness tester like we had across the alley from the community collage Adult ed welding classes I was taking while I was still going to High School. Long story.

PS: When I looked at the hand held job my thought was "that is one light duty hardness tester . . . not like the ones I am flirtingly familiar with".

[David Weaver]

Rob, I don't know anything about your tester or its application. I suspect it's not being performed correctly, but because I don't know the testing device or what the pitfalls of its translation could be (or where it could have error).

I know that the saws you tested are not 58 or 61 if they have been filed before. That much is clear from actual experience. When I harden a chisel, I check it with a file. When it is light straw, it can't be filed - the teeth of files get burnished or chipped off by something that hard, and that's without it being thin like a saw plate (which puts fewer file teeth in a cut

I have filed a dozen and a half different backsaws from the early 1800s to now, and carpenters saws from the mid 1800s to probably the 1970s, and I guess it would irritate me to see someone going elsewhere and telling everyone that disston saws are harder than 52 hardness boutique saws, same with backsaws. That is very very seldom true.

You have a technical problem with your method somewhere, but until you can even suggest as to why someone would want a saw that couldn't be filed, you shouldn't be leaning on a tool that you don't have much experience with and asserting it's correct because you have a process that you adhere to.

[David Weaver]

better yet take the saws to a behemoth hardness tester like we had across the alley from the community collage

Stan's saw was dark brown at the tang. The rest had been polished off, so I don't know where it was in hardness, but I think stan told me something like 54? I can't recall for sure. I would never have bent it back around to itself like you can do with a western saw.

[Winton Applegate]

David,
Yup . . 54 too hard to file with western files, especially most all of the curent shlock, and I never bend my saws like the showman so I think I am safe here.
That was sure a nice saw you shared with us. Gives us a bench mark for everything else.

[Steve Voigt]

I don't have enough expertise with saws to take a firm stand here; all my saws are old Disston's and I've never owned a modern premium saw. The numbers do seem high to me, and don't necessarily pass the "common-sense" test, but that's not proof.
More to the point, I think it's a little discouraging, and off-putting, when a guy does a crapload of research, posts it, and then immediately comes under withering criticism.
It would be great if someone else with a decent quality hardness tester could do some similar tests and let us know the results. Otherwise it's all just talk.

[Winton Applegate]

It would be great if someone else with a decent quality hardness tester could do some similar tests and let us know the results. Otherwise it's all just talk.

Well . . . we are not looking at a chart of numbers from him it is true but . . . George did exactly that from what he is telling us and I am sure he has no reason to bend the facts.


I think it's a little discouraging, and off-putting, when a guy does a crapload of research, posts it, and then immediately comes under withering criticism

Yes.
Though that is the risk one takes (my chip breaker rants and ultimately taking a head shot for my mistake is a good example) for doing the best one can and posting about it.

I would like every one, especially those who read but very rarely post to take more risks and post the findings. Some one is bound to learn something some where and those who have typed about that beat subject too often can just say out of it.

[David Weaver]

David,
Yup . . 54 too hard to file with western files, especially most all of the curent shlock, and I never bend my saws like the showman so I think I am safe here.
That was sure a nice saw you shared with us. Gives us a bench mark for everything else.

All you have to do is cough up a grand and get on the waiting list and you, too can have one :) I wish I was a more competent user of it, I can't convey the difference between it cutting 7/4 material and every run of the mill long kataba or ryoba that I've had. It was like riding 120 psi on cobblestones and having someone point you to a turn off on a freshly paved road with no expansion gaps on it.

[David Weaver]

my chip breaker rants and ultimately taking a head shot for my mistake is a good example..

Well, there's a lot more opinion there, of course. Some of us are partisans in that debate!

(I didn't, by the way, file stanley's saw at all. I figured it would be inappropriate to do something like that - I think I could've filed it with my vintage files. Stanley said to me when I complained that I might not like to test an expensive saw that I could break, that I would be quite sure as soon as I got a hold of it that it's not a saw that would be easily broken, and he was right).

[Winton Applegate]

partisans
But patriots when it comes to Neandering.

not a saw that would be easily broken, and he was right
Poetry in steel.

Problem is if I coughed up a thou and got it then I would want to use only that saw for every thing and would start posting about how one can cut dovetails with it and the stuff would hit the fan all over again.

. . .. . .
. . . .
. . . sayyyyy . . . that would be fun . . . .:)

[David Weaver]

I thought when I got it "this would be a good saw for stuff thicker than 4/4, but I might like to have one in 6 or 7 tooth for that..." (as in buy two).

(there was only one left when I tried Stans - it was the matching plate to stans from a pair. I think he said the maker makes a few dozen saws a year and that's it, so it's not like you can say "can you put another one with X# of teeth in the mail monday". I ALMOST bought that second saw, but I wavered a little bit and someone else did - which ultimately is good for my marriage)

[Rob Streeper]

David,

The Ames testers are very simple instruments to use. If you're interested go here http://www.amesportablehardnesstesters.com/rockwell-tester-instructions/ and watch the video.
True, it is not a bench style or other high precision tester but, like I said above, these are saw blades - hardly worth getting your blood pressure up. Plus or minus a half point isn't critical here.
I find it very disturbing that so many responses have included personal jabs directed at me. In my world, data is data and opinions are not data. Experience is not data either. If anybody has that data let's compare notes and be objective in the interpretations.
As I keep saying, nobody else has presented any data of any kind.

Thanks,
Rob

[george wilson]

Steve,I find it absolutely stunning that you seem to want an anything goes approach to posting "facts" here. Do you think anyone should post any old "facts" here and not be questioned about its validity? I hope our reporting press does not do things that way,or we'll never know what is really going on.

I just think getting facts straight is important. And I have given them several times in this thread,yet am still accused of just voicing my opinion. I know the hardness of every saw and chisel I own because I tested all of them with a reliable and correct RC range tester. Now,THAT was a "Crap load" of research that I did for my own education.

Rob,no personal jabs have been directed at you. And,the range of hardness for good old saws-40 to 45 RC,and new boutique saws @ 52RC compared to a figure of yours- 61,is not 1/2 f a point.

[David Weaver]

David,

The Ames testers are very simple instruments to use. If you're interested go here http://www.amesportablehardnesstesters.com/rockwell-tester-instructions/ and watch the video.
True, it is not a bench style or other high precision tester but, like I said above, these are saw blades - hardly worth getting your blood pressure up. Plus or minus a half point isn't critical here.
I find it very disturbing that so many responses have included personal jabs directed at me. In my world, data is data and opinions are not data. Experience is not data either. If anybody has that data let's compare notes and be objective in the interpretations.
As I keep saying, nobody else has presented any data of any kind.

Thanks,
Rob

Rob, purchase a couple of files at home depot, put the saws upside down in a vise and see if you can file them.

If you can file them at all, that will prove that you have a problem with the method somewhere.

I also work with (a lot of) data and I know when you get an oddball result from something that you haven't got a lot of experience with, you'd better be very wary. The entire group of people that I have ever talked to with saw experience (including people who file a lot of saws) disagrees with what you have told us, that old saws are generally harder than 52 RC 1095, and some substantially so (complete with "caveat emptor warning").

The only possible exception would be if you have somehow gotten only a few disston saws and several turned out to be bad, and that you got one atkins and it's also very bad. That's entirely possible, and that's why I'm asking you to file it. If the file skates across them and is burnished over in just a few strokes, that solves the problem.

Otherwise, you cannot suggest that files hardened in the low 60s will file saws in the low 60s and leave that just out there as that, especially when it's very easy test the file theory on tools of known hardness (and find that they will skate when they are chisel/plane/razor hardness).

[Bruce Page]

Everyone, take a deep breath and keep the discussion friendly. :)

[Rob Streeper]

Hi David,

I did you one better on the files. Here are some pictures of a test on a small square Nicholson file. Note that this file has one safe side - perhaps a special item?

304139

304140

The hardness reading came out to be 64, just as you would expect. I tested the same file on my B/C tester and the hardness read at 64 as well. The B/C tester is one of the Chinese built cheapies but despite having an offset it is quite linear, much better than I expected it to be.
I have sharpened the Disstons I tested and they are really hard and they tend to wear files out faster than does the 1095 that I use for my saws. The Disstons are harder to set too.

Cheers,
Rob

P.S. Oops, made an error here. The Rc value should be about 58 for this test on this instrument. Being harried makes for errors!

[george wilson]

Where was the Nicholson file made? The Mexican ones are terribly soft. I can easily file a Mexican made Nicholson with a NOS USA Nicholson file.

I cannot explain how you seemed to get a correct reading on the file,IF it is a USA made one. But,I can assure you that your readings on the saws are simply not correct. In all my years of being a woodworker and tool maker,I have never seen an old Disston that was as hard to file as a new 1095 saw. And,for many years,I filed all of the saws,mine and the old ones also,used in Williamsburg.

As I mentioned,if your diamond has a tiny crack in it,it will give erratic readings. That COULD enter in here,for all I know. I have no way of judging your tester.


Steve,it is hard to judge a person's words on a forum. I know David pretty well by now,and I know he is not a mean person. He has good judgement about using sound scientific procedures in both his work,which demands it,and in the study of tools that he has made. I am sure that he is only trying to get at the truth. David and I both very much care about the members getting correct information. Does it really matter if they have any real idea how hard a saw is? It was necessary for me to know as a professional tool maker. Maybe it doesn't really matter to a hobby user,as long as the saw can be sharpened. That's my nature,though,and David's too. We want accurate information to be given. Humans are nuts anyway! :)

[Ron Bontz]

Hmm. Well the only data I see is the data I receive in the form of MSDS sheets and test reports of the steel, when I buy a small roll of it. There is no secrets as to where my saw plates come from. Me or the water jet guy. :):) It is just 1095 with an RC48-51. I am able to buy 1075 with an RC of 52 and I know of 1085 as well. But I don't use it. I do not have the time or money to buy the steel and try it out to see how it holds up. Hardness is not a good indicator in and of itself of steel quality. Not being a metallurgist, I always think of the debates between O1 steel and A2 or V11. One takes a keener edge and the other holds up longer. Lathe tools are no different. M42 vs PMV10 vs PMV15. Sears used to make a saw ( 30 years ago ) That was clearly a softer steel made obvious when sharpening. It quickly took a keen edge but it did not rust as quickly and dulled faster. No doubt a calculated trade off. The old carpenters I used to work around would talk of the difference in saws at lunch break. Stanley is harder to sharpen but holds an edge longer than the Craftsman etc.. All I really know is that once a tool reaches a certain quality, it is a poor craftsman that blames the tool. AH HA! That's why my dovetails are so bad! It's the saws I use with crappy steel in them. OH WAIT! They are my saws. :( Opps. :) Have a good night. :)

[Rob Streeper]

As to the Nicholson file, I think that it's an American one. The stamp is so small that it's hard to read although it does have the standard Nicholson look to it. I tested several others, including some that appear to be quite old and they gave similar readings.
As to the condition of the diamond in the tester, it seems to be fine and I think that on the basis of the replicate measurements I have done. No wild variations, only a couple of points one way or the other.

[Larry Frank]

I have been reading this thread and one thing that I missed is the calibration tests for the instruments being used. I do assume someone ran calibration tests???? If so what were the results. As a metallurgist, I would never do a study like this without frequent calibration tests and posting the results of the calibration. With thinner materials it is easy for error to creep in and calibration tests validate the results.

[Rob Streeper]

I have been reading this thread and one thing that I missed is the calibration tests for the instruments being used. I do assume someone ran calibration tests???? If so what were the results. As a metallurgist, I would never do a study like this without frequent calibration tests and posting the results of the calibration. With thinner materials it is easy for error to creep in and calibration tests validate the results.

Yes, the surface hardness tester comes with two test blocks, one at 60.2 and one at 49. I checked the tester and the indicated values are within the specifications of the blocks, repeatedly.

[Patrick McCarthy]

This is about the most gentlemanly forum I have experienced. That being said, you cant expect to make statements and not be able to support them in a scientific way (ie, a generally accepted scientific way). I am a veritable virgin to metallurgy but even I have to tell you that thus far I sense you are using incorrect assumptions as to the application of the NIST tables, and the capabilities (ie design intent) of your gauge.
My comments and 3.50 will get you a cup of coffee, and not much more.

[george wilson]

Certainly anyone can be wrong. No one is perfect.

What I know about THIS subject has nothing to do with commonly held or cherished ideas. I have cited several times results that I got with a very choice,high grade hardness tester,calibrated with proper test pieces. Yet,you still call my information my OPINION. I don't think we are getting through to each other. I don't think you want the benefit of my lifetime of experience as a professional tool maker. What about that? Am I being mistreated too?

Patrick,you have exactly hit the nail on the head,and worded it very well. If you want correct scientific data,you HAVE to use the correct instrument to get it. And,you can't post tables,seemingly ignoring what the table says about what the particular scales are used for. If that makes Rob feel like he has been unfairly put upon,it is very unfortunate,and I am sorry he feels that way.

[Rob Streeper]

George,
My testers, fallible though everyone thinks they are are giving consistent results. Work with me here for a minute. Look at my data for the 1095 steel only. If we leave out the 0.015" material you can see that the results are quite consistent. I just discovered a math error in my standard deviation calculations but here is the table for the 1095.

304149
The standard deviations, now calculated for each thickness of test material, are quite reasonable given that the tester I'm using has a tolerance of plus or minus 1 Rc or HRC point or a range of 2 Rc points, exactly equaling the maximum standard deviation in the measurements I made. The actual values are also, within the error limits of the instrument, within the manufacturer designated specification range for this material. In fact I'm surprised that the results are as consistent as they are given that each of these stock materials came from a different source.
Thus, these results suggest to me that the measurements are reasonable. I've done some testing using my B/C tester, and where the material is thick enough the results are very close to these. I'll now revise the full table to correct the math error and post it.

Cheers,
Rob

[Rob Streeper]

Here's the revised table. I made a math error in calculating the standard deviations of the measurements, corrected now.304150

[Rob Streeper]

I just had a thought about the high values for the Disston saws. These measurements were taken in the area under the handles. On the Woodwork Forums site I have a thread running about making a replica of D-8 taper ground saw plates. I just took a few measurements of the areas that I had hammered on and then ground smooth. The values were from 62 to 64 suggesting that the hammering has drastically increased the hardness of the steel. I wonder if the steel under the handles of the Disston saws had been work hardened as well. If so my results would be consistent with the Disstonian statements about the hardness under the handles.

[David Weaver]

I just had a thought about the high values for the Disston saws. These measurements were taken in the area under the handles. On the Woodwork Forums site I have a thread running about making a replica of D-8 taper ground saw plates. I just took a few measurements of the areas that I had hammered on and then ground smooth. The values were from 62 to 64 suggesting that the hammering has drastically increased the hardness of the steel. I wonder if the steel under the handles of the Disston saws had been work hardened as well. If so my results would be consistent with the Disstonian statements about the hardness under the handles.

The simple fact exists that saws need to be double digits in hardness less than files for a file to have any longevity in a saw.

You have now suggested that you have a file that's 64, but it's not a file you'd file a saw with. You can grind off and stone an area on a triangular saw file and test its hardness. If you find it's 64 and file the atkins and the saw file does so without getting burnished immediately, you will know that your testing has a very large error in it.

This differential is something that every vintage sawmaker knew very well, because their saws were used heavily. They would not have made saws 55+ hardness when files are only low 60s in hardness.

[Pat Barry]

I just had a thought about the high values for the Disston saws. These measurements were taken in the area under the handles. On the Woodwork Forums site I have a thread running about making a replica of D-8 taper ground saw plates. I just took a few measurements of the areas that I had hammered on and then ground smooth. The values were from 62 to 64 suggesting that the hammering has drastically increased the hardness of the steel. I wonder if the steel under the handles of the Disston saws had been work hardened as well. If so my results would be consistent with the Disstonian statements about the hardness under the handles.
I did some checking on George's assertion and in fact, he is correct. This article (http://www.labtesting.com/wp-content/uploads/2012/08/chart-hardness-c.pdf)talks about the difference in readings between the C scale and the N scales. Your measurements on the N scale, particulalry at the lower forces, are significantly higher than the C measurements would be. Since you have access to the C method, I agree with George's previous point that you should take measurements with both and then it will be very interesting to see how your numbers compare

[george wilson]

I am happy to see more reasonable points of hardness for the 1095. I always got a pretty consistent 52 from the stock I tested,but I won't quibble about 1 extra point. Besides,different batches are possibly a little different.

But I still cannot agree with the very high points you are showing for old Disstons that I have never gotten more than 40-45 RC from.

However you have now found errors in your original calculations,and have recalled that you hammered the areas you tested. Yes,hammering WILL increase the hardness of the steel. In fact,old time blacksmiths who had to repair broken buggy springs from under the seats of buggies,HAMMERED the "temper" back into the springs. When the break was welded,the areas for several inches near the weld would have been rendered fully annealed. If the blacksmith had tried quenching them,they would have been glass hard, with areas a little farther away left too soft. Springs are very tricky to get right. (I have made quite a few of them,and have had a few blow up violently in my face!!) So,they let them air cool slowly,and hammered the spring temper back into them. They were clever enough to get away without large furnaces that the manufacturers would have had to harden and temper the springs initially.

So,it really looks like these errors have caused a great deal of discussion. I am glad to see them resolved.

I would like to point out that what a close and trusted friend can say is received totally differently from what a stranger might say. And,for the greatest part BY FAR,we are strangers to each other. So,how can we all expect to trust what one another says without taking it the wrong way. I think that is worth remembering and thinking about.


Pat,I'm glad to see you agree with me. After all,I was doing this work for many years,and continue to do so even this very day. How could I have made the things I have without knowing what to do with tool steels.

[Rob Streeper]

I am happy to see more reasonable points of hardness for the 1095. I always got a pretty consistent 52 from the stock I tested,but I won't quibble about 1 extra point. Besides,different batches are possibly a little different.

But I still cannot agree with the very high points you are showing for old Disstons that I have never gotten more than 40-45 RC from.

However you have now found errors in your original calculations,and have recalled that you hammered the areas you tested. Yes,hammering WILL increase the hardness of the steel. In fact,old time blacksmiths who had to repair broken buggy springs from under the seats of buggies,HAMMERED the "temper" back into the springs. When the break was welded,the areas for several inches near the weld would have been rendered fully annealed. If the blacksmith had tried quenching them,they would have been glass hard, with areas a little farther away left too soft. Springs are very tricky to get right. (I have made quite a few of them,and have had a few blow up violently in my face!!) So,they let them air cool slowly,and hammered the spring temper back into them. They were clever enough to get away without large furnaces that the manufacturers would have had to harden and temper the springs initially.

So,it really looks like these errors have caused a great deal of discussion. I am glad to see them resolved.

I would like to point out that what a close and trusted friend can say is received totally differently from what a stranger might say. And,for the greatest part BY FAR,we are strangers to each other. So,how can we all expect to trust what one another says without taking it the wrong way. I think that is worth remembering and thinking about.


Pat,I'm glad to see you agree with me. After all,I was doing this work for many years,and continue to do so even this very day. How could I have made the things I have without knowing what to do with tool steels.

George,

I just had a realization. As I wrote earlier I tested the hardness on the Disston saws in the area under the handle. If these saws were work hardened by hammering as is described in accounts of Disstons production methods then the saws may be harder there. I have been working on making a replica D-8 saw. As part of the process I tried hammering the blade along the back and down the center of the blade. I did not hammer in the area of the teeth. I ground the blades and this evening I took a few measurements. The values were 62, 64 and 64. This is the same 0.035" 1095 steel tested in the table above, just hammered and ground.

As to the proper scale I have just corresponded with Ray Gardiner on the subject. Here is what I wrote.

Hi Ray,

I used the conversion charts to produce the Rc values reported. NIST recommends the N scale for thin hard materials and they recommend the highest loading permissible for a given test material thickness. Look here: http://qs-hardnesstester.com/hardnes...thickness.html (http://qs-hardnesstester.com/hardness-minimum-thickness.html)

Referring to the linked chart and given that we're talking about materials that should have Rc hardnesses of 45 or greater you can see that the minimum thickness recommended for a C scale measurement is 0.034" for Rc45. If the material is harder, say Rc 52 then the minimum thickness decreases to 0.032" and so on.
Look to the right side of the chart now at the N scales. You will see that for a 0.015" thick test article that the minimum Rc value is somewhere between 68 and 76. Thus N15 testing of 0.015" materials is not really that good. However, on the N30 and N45 scales you will see that the minimum thicknesses recommended are 0.022" and 0.024" respectively. All of the saws I tested have thicknesses equal to or greater than than 0.022".
To accurately measure 0.015" material I need to go to the 15T or 30T scales, something I need to do in the future obviously.

Please understand that I am doing these studies because I am interested. I have searched the web and have not found any systematic study of the hardness values of woodworking saw blades. I wonder why?

Cheers,
Rob

For material thicker than 0.032" I agree that HRC will work. For thinner material apparently it will not.

[Isaac Smith]

In fact,old time blacksmiths who had to repair broken buggy springs from under the seats of buggies,HAMMERED the "temper" back into the springs. When the break was welded,the areas for several inches near the weld would have been rendered fully annealed. If the blacksmith had tried quenching them,they would have been glass hard, with areas a little farther away left too soft...So,they let them air cool slowly,and hammered the spring temper back into them. They were clever enough to get away without large furnaces that the manufacturers would have had to harden and temper the springs initially.

How hard does one have to hit the steel to make this happen? Many light taps, or were full strength blows better? Were these repairs considered "almost as good as new", or more of a "that'll do for now" thing? I realize that any answer will be a broad generalization, but I am curious about this.

[Rob Streeper]

George,
On the issue of files I just looked to my left and saw that I have a Nicholson auger file sitting on my desk. I measured it and it comes back with a C scale hardness of 60. On examination with a magnifier I can see that it is a US made file. It's also in new condition and the end made for filing with the edges has two nice flat surfaces. Probably not ideally smooth but it's certainly better than the file I tested before as far as smoothness and parallelism of the sides facing the anvil and indenter.

Cheers,
Rob

[Rob Streeper]

I did some checking on George's assertion and in fact, he is correct. This article (http://www.labtesting.com/wp-content/uploads/2012/08/chart-hardness-c.pdf)talks about the difference in readings between the C scale and the N scales. Your measurements on the N scale, particulalry at the lower forces, are significantly higher than the C measurements would be. Since you have access to the C method, I agree with George's previous point that you should take measurements with both and then it will be very interesting to see how your numbers compare

Hi Pat,

Let's consider a hypothetical. In the first panel of the table of results I posted I measured samples of 1095 steel. If I was reporting N scale measurements of 52 ish those would convert to C scale measurements of approximately 32 since I used the 30N scale for most of the tests. Rc32 is well below the accepted hardness value for 1095 straight from the roll but that's not what I reported. I was measuring N30 readings on the instrument in the mid 70's and using the conversion tables I was writing down the corresponding C scale values.

Look closely at the data. The values I reported for the 1095 are what you would expect for that material. The testing on the various saw plates was done exactly the same way with exactly the same instrument.

The Disston saw values are likely high because I measured the plates under the handles. If Disston worked the saw plates in this area by hammering then they would be harder, consistent with my results on my D-8 build page. Thus these results can be explained by regional differences in the hardness - something that would be good to know, yes?

As to the custom makers, one of the saws is of different design and thus the plate is much harder. Given what others have written here that is to be expected for this type of saw.

Cheers,
Rob

[Rob Streeper]

Why are you avoiding the file test that I suggested? The simple fact exists that saws need to be double digits in hardness less than files for a file to have any longevity in a saw.

You have now suggested that you have a file that's 64, but it's not a file you'd file a saw with. You can grind off and stone an area on a triangular saw file and test its hardness. If you find it's 64 and file the atkins and the saw file does so without getting burnished immediately, you will know that your testing has a very large error in it.

This differential is something that every vintage sawmaker knew very well, because their saws were used heavily. They would not have made saws 55+ hardness when files are only low 60s in hardness.

I don't know if you're afraid of finding out that your results are errant, which is something we already know to be true if the saw is still in spring temper, but dancing around the problem doesn't do anyone any favors.

David, I'm not avoiding doing what you suggest, I'm not going to do it because it won't produce a reportable and verifiable numerical result.
As I've noted elsewhere in this thread, I've come to suspect that the region of the Disston saws I measured had been work hardened. An observation consistent with what has been published regarding Disston's manufacturing procedures and is reflected in the Disstonian discussion of their steel. See: http://www.disstonianinstitute.com/steel.html "The saws with multiple values for hardness were hardest in the area under the handle, and softest in the area close to the teeth."
One problem with measuring near the teeth will of course be work hardening as Disston punched teeth and another will be surface finish and flatness. Rockwell testing doesn't work well on rough or bent surfaces. As I wrote I made no effort to polish the test points, I just chose areas that had no pitting or obvious roughness. I wiped the saws with solvent to remove dirt and goop.

Cheers,
Rob

[Rob Streeper]

<Correct on a nominal (not relative) basis is important because he is painting people who use industrial supply 1095 coil as making saws that are substandard.>

I am not saying that 1095 is substandard nor am I painting anyone, at least anyone who is identifiable. I use 1095 to make saws, why else would I have so many thicknesses of 1095 stock available to test? I'm saying that 1095 is softer than the steel under the handles of Disston saws but only a by little bit. Within the limits of my measurements 1095 is 3.5% softer than the Disston steel under the handles. I doubt very much if there are any modern woodworkers, or for that matter woodworkers of any era, who could differentiate between the cutting capacity of two saws differing by 3.5%.
Cheers,
Rob

[Kees Heiden]

Hello Rob,

Your values for several saws indeed seem to be off. An idea to validate your measuring tool for this purpose might be to make some test strips yourself. Get a few smallish bit of 1095 steel, anneal them, and then harden them. When the pieces are small enough you can do this with a simple propane torch. Then you temper the several pieces to several different colors. Leave on at maximum harness.

Now you can check with your hardness tester to see if the value you measure, agrees with the expected hardness for the color. There are plenty color charts available for 1095 steel. This won't give you absolute numbers, but it will give you an idea if your testing technique is way of the charts or not.

[Rob Streeper]

Hello Rob,

Your values for several saws indeed seem to be off. An idea to validate your measuring tool for this purpose might be to make some test strips yourself. Get a few smallish bit of 1095 steel, anneal them, and then harden them. When the pieces are small enough you can do this with a simple propane torch. Then you temper the several pieces to several different colors. Leave on at maximum harness.

Now you can check with your hardness tester to see if the value you measure, agrees with the expected hardness for the color. There are plenty color charts available for 1095 steel. This won't give you absolute numbers, but it will give you an idea if your testing technique is way of the charts or not.

Hi Kees,

I agree that the numbers don't make sense when compared to Disston's assertion that Rc52 was the best for wood cutting hand powered saw blades. Unfortunately Disston's 52 and the limited data presented on the Disstonian Institute seem to be all of the data that is publicly available.
However, it doesn't appear that anybody has thought to acquire, compile, analyze and report any data on modern saws and compare that data to the older saws. You'll notice in the responses to my original posting that nobody has posted a single measurement on a saw that they have done nor have links to any such studies been posted. The vast majority of the posts have been some combination of rejection of my data and/or attacks on my credibility because I am not a recognized member of the club that routinely posts here.
As to the 1095 issue, I have tried various experiments in hardening using it. I posted my results on Woodwork Forums at the beginning of this year. In summary I found that it is possible to harden 1095 to the point that it shatters like glass - I couldn't even get a hardness reading because the samples exploded into shards under the indenter.
Cheers,
Rob

[ian maybury]

Seems to me that the fact that Rob has a tester to hand may amount to an opportunity for those with an interest to in a spirit of co-operation see some testing done, and to resolve some technical questions regarding the steels used in hand tools. It seems like the pooling of the knowledge around here ought to be capable of resolving some questions, and/or giving a once over to some sacred cows...

In that spirit here's a link to an ASTM paper which sets out to establish a standard procedure for the use of portable hardness testers - not sure if it might add anything, but it may be worth a look: http://www.astm.org/Standards/E110.htm

[Jim Wheeler]

Hi, Rob,

Something I'm wondering about the Disston saws: did you find any difference in hardness of the saws between the tooth lines and the bodies? I have read that Disston used a proprietary process to harden the tooth line of their saws so that it was harder than the rest of the plate.

I don't own or have access to a hardness tester, but in my experience Disston saws are a bit harder to file than some others. It also seems more difficult to remove a kink or a bow from them.

Regards,

Jim Wheeler

He who welds steel with flaming pine cones may accomplish anything!​

[David Weaver]

I have no idea what virtue a bunch of numbers have:
1) if they don't have any credibility and haven't been verified by something that's well known (it is well known that a file that's not 10 points harder or more will not last filing a saw plate)
2) if the saw numbers aren't struck close to the teeth, the part of the saw that matters the most
3) George Wilson, a professional tool maker (as in makes a living doing it) tested several vintage saw plates on a versitron and got numbers nowhere close to what you're getting

There's no real sacred cows in saw making, except that many of the people selling new parts and new saws don't do the work to make handles what they should be (mass produced and boutique) because they don't either use the saws or they don't care and they just want to sell something without having to make it good in the sense of what was good to a professional user.

The old tools (pre 1900 and around there) were made in a constant feedback loop with professional users. Boutique tools now are made either substandard (saws with blocky handles, overweight backs, unattractive metallic fixtures or goofy designs like saw bolts through the spine), or they are super finely made stuff (like what Ron B makes) that are expensive because you are paying for someone to make something as well as they can. They are not representative of what was used to actually do work. A professional user would've been too cheap to buy the high dollar (though I quite like Ron's efforts) stuff and they wouldn't have tolerated the boutique or production stuff that is blocky or has wonky proportions on the saw plate (too thick) or an obscenely heavy back.

These displays of numerical items are useless because saws are used in context of work and in context of filing them. You judge a saw on:
* how are the proportions (length, plate height, handle size and orientation, plate thickness)
* how does it file
* how are the taper and tension
* how does it feel in the cut
* how do the teeth hold up until you get to your next touch up filing after a project

The rest of the stuff doesn't amount to anything unless people who don't use saws are talking about the numbers with each other.

If there is a real problem with the saw, it's not to be found with a hardness tester, it will be found in use. Or when it's time to file the saw. Anything that doesn't show up either of those two times doesn't matter.

You can strike your saws near the tooth lines and communicate hardness numbers but in the end you're unlikely to find different from George unless you have a very odd sample of saws. Some of us already know from using the saws what is harder at the tooth line and what isn't, and how it works in the context of actual use.

You could tell any of us that a pax/flinn saw is not as hard as a 1095 saw, but we already know that. All you have to do is file one. You could tell us the deer brand saws that are sold by woodcraft are extremely soft, but we already know that from filing. If you imply that some measure of important hardness in a disston saw shows them at 58 and an atkins at 61.5, then that's not only pointless, but it's misleading.

[Pat Barry]

Hi Pat,

Let's consider a hypothetical. In the first panel of the table of results I posted I measured samples of 1095 steel. If I was reporting N scale measurements of 52 ish those would convert to C scale measurements of approximately 32 since I used the 30N scale for most of the tests. Rc32 is well below the accepted hardness value for 1095 straight from the roll but that's not what I reported. I was measuring N30 readings on the instrument in the mid 70's and using the conversion tables I was writing down the corresponding C scale values.

Look closely at the data. The values I reported for the 1095 are what you would expect for that material. The testing on the various saw plates was done exactly the same way with exactly the same instrument.

The Disston saw values are likely high because I measured the plates under the handles. If Disston worked the saw plates in this area by hammering then they would be harder, consistent with my results on my D-8 build page. Thus these results can be explained by regional differences in the hardness - something that would be good to know, yes?

As to the custom makers, one of the saws is of different design and thus the plate is much harder. Given what others have written here that is to be expected for this type of saw.

Cheers,
Rob
From the research I have done spurred by this thread it is evident that the precision and accuracy of the test is strongly influenced by the test method, the accuracy of the tool itself, and the material being tested. Each of the scales are designed for their purpose and not really intended to be used to do anything other than provide a relative measure of hardness within their own dataset. Persoanlly, I would repeat the measurement process on the manufacturer C which you recorded in the 30 range. I would do this with both the Rockwell C and N measurement methods. I would also do this with one of the Disston saw plates which you recorded in the 50 range. Keep in mind that the material thickness does have a bearing on your test results. This is evident from the tabulated data (http://qs-hardnesstester.com/hardness-minimum-thickness.html) you referenced early in this thread. Again, keep in mind, the point is really not whether your test results are accurate, its whether or not conclusions can be drawn from them. Therefore the relative results should be the focus, not the absoluteness of the results. Is the saw plate that you previously recorded as in the 30 range significantly softer than the Disston plate you measured in the 50 range, regardless of the scale used?

[Rob Streeper]

David,
I've filed the saws in question and they can be sharpened, I've also set them and not broken teeth. These saws feel harder than modern saws and saws made of 1095. I suspect now that the teeth are softer than the steel under the handle is. Unfortunately the fact that the teeth have been work hardened both by Disston's use of punching in making them and by setting over the years will distort the readings. In addition the bend in the metal will also throw off the readings. In other words, measuring the hardness of the teeth can't be done using the Rockwell testers I have. Another limitation of doing measurements on the plate in the area of the teeth is the surface finish. Rough surfaces will return anomalously low hardness readings than will smooth.

[Rob Streeper]

From the research I have done spurred by this thread it is evident that the precision and accuracy of the test is strongly influenced by the test method, the accuracy of the tool itself, and the material being tested. Each of the scales are designed for their purpose and not really intended to be used to do anything other than provide a relative measure of hardness within their own dataset. Persoanlly, I would repeat the measurement process on the manufacturer C which you recorded in the 30 range. I would do this with both the Rockwell C and N measurement methods. I would also do this with one of the Disston saw plates which you recorded in the 50 range. Keep in mind that the material thickness does have a bearing on your test results. This is evident from the tabulated data (http://qs-hardnesstester.com/hardness-minimum-thickness.html) you referenced early in this thread. Again, keep in mind, the point is really not whether your test results are accurate, its whether or not conclusions can be drawn from them. Therefore the relative results should be the focus, not the absoluteness of the results. Is the saw plate that you previously recorded as in the 30 range significantly softer than the Disston plate you measured in the 50 range, regardless of the scale used?

Pat, I have tested the Maker 3 saw on both instruments/scales. I first tested it on the C scale 150kgf loading and it recorded ridiculously low and widely scattered values - both bad signs and of course the measurements were not accurate. The values were all below 25. Those results were obtained on a saw plate that is 0.035" thick. That thickness is within the thickness range considered acceptable for C scale testing if the material is fairly hard. But obviously the plate was softer than is acceptable. Those readings are one of the reasons that I bought the N/T scale tester. Testing with the N/T tester and the diamond indenter also showed that the Maker 3 plate is soft and tellingly the scatter in the readings is lower but not as low as the scatter for the other test specimens.

[Rob Streeper]

Hi, Rob,

Something I'm wondering about the Disston saws: did you find any difference in hardness of the saws between the tooth lines and the bodies? I have read that Disston used a proprietary process to harden the tooth line of their saws so that it was harder than the rest of the plate.

I don't own or have access to a hardness tester, but in my experience Disston saws are a bit harder to file than some others. It also seems more difficult to remove a kink or a bow from them.

Regards,

Jim Wheeler

He who welds steel with flaming pine cones may accomplish anything!​

Hi Jim,

I'm planning on more testing but I'm going to use fresh 1095 straight from the roll. I've just run a power analysis on the basis of the results posted above and I see that I need an estimated 38 data points, I'm going to do 50. So I'm going to test raw 1095 and compare it to 1095 that has been hammered and ground and compare those readings to 1095 that has been ground with no hammering.

[Steve Rozmiarek]

Does anyone have access to other testers, like the one George mentioned? I'd personally enjoy seeing the data of a test replicating Rob's test subjects on a different tester.

Of course David is right, all this probably matters very little in practice, but it does have good entertainment value.

[Rob Streeper]

Of course David is right, all this probably matters very little in practice, but it does have good entertainment value.

I think that the posters here are very interested in this subject. They've spent a lot of time criticizing my results but again, nobody has produced any data but me.

[Kees Heiden]

I think everyone reacted so strongly becaue your numbers don't agree with the "seat of the pants" feeling. We have all sharpened our saws, and I guees many of us have tried to file chisel or plane blade steel once or twice. I have tried to make a float from an old Stanley blade but quickly gave up. I have also filed many teeth in many saws and that never was awfully hard to do. Chisel and plane blade steel is usually advertised to be somewhere between 58 and 62 HRc. So when you publish data with these kinds of values, then it is no wonder people are protesting. Your values just didn't stroke with normal workshop experience. So somewhere something should be wrong. As you allready posted, testing under the handle might be one of the explanations.

[Prashun Patel]

As usual, I've learned so much from a thread like this. The debate about the validity of the results is as illuminating as the data.

I've taken a heavy hand and removed a lot of the posts that were not germane to the discussion.

If you feel you've been unfairly edited, please pm me.

[David Weaver]

I think that the posters here are very interested in this subject. They've spent a lot of time criticizing my results but again, nobody has produced any data but me.

You implied the practical hardness of a disston and atkins saws (at least some) were 58 to 61. It's data, but it's either out of context or not correct. George provided results of what he's seen from a versitron and that the 1095 is accurate and the vintage saws were somewhere in the low to mid 40s (as in actual data, george provided you with data from a much more accurate machine).

It's fine to be enthusiastic, but if you're going to assert something, you need to be correct and at this point, everything you have told us is very misleading, and you've refused to validate your numbers by a simple filing of the teeth of the saws (certainly if your saws are 58 and 61 hardness and cannot be filed practically, they need to be thrown out).

[Rob Streeper]

Kees,

I was surprised at the results I got too. Nonetheless the tester checks out fine when used on the test blocks that came with it. For testing of saw blade materials that are both thick and hard enough the results on the C scale tester I have are the same within the error window of the measurement.
If it turns out that the Disston saws have different regions of hardness across the blades that was intentionally worked in during manufacture I think that information would be of great value to those currently making saws.

[David Weaver]

David,
I've filed the saws in question and they can be sharpened, I've also set them and not broken teeth. These saws feel harder than modern saws and saws made of 1095. I suspect now that the teeth are softer than the steel under the handle is.

If they can be reasonably filed, then the 58 and 61 results are meaningless. In order to test what feels harder (1095 vs vintage) plates of the same thickness and teeth of the same size need to be filed with the same file.

Without exception, every vintage backsaw that I've filed (10, 15?) is softer than the 1095 backsaws that I've filed (4 separate 1095 plates ordered at different times). Every 1095 backsaw seems easier to file than every large rip saw because it's harder to file large rip teeth than it is to file smaller teeth and because the plate on the rip saws is twice as wide.

I would suggest using a C tester and striking the saw plate 1/2 to 1 inch above the tooth lines front, center and back.

[Rob Streeper]

You implied the practical hardness of a disston and atkins saws (at least some) were 58 to 61. It's data, but it's either out of context or not correct. George provided results of what he's seen from a versitron and that the 1095 is accurate and the vintage saws were somewhere in the low to mid 40s (as in actual data, george provided you with data from a much more accurate machine).

It's fine to be enthusiastic, but if you're going to assert something, you need to be correct and at this point, everything you have told us is very misleading, and you've refused to validate your numbers by a simple filing of the teeth of the saws (certainly if your saws are 58 and 61 hardness and cannot be filed practically, they need to be thrown out).

David,
George told us about his experience and in a later post he acknowledged that his measurements on 1095 were comparable to mine. Given that the measurements of the various saws were done in exactly the same way using the same tester it seems that there is something going on here, right?
Don't you want to know why these values look weird? I do and that's why I'm doing the testing and that's why I reported the results. I did this despite the fact that I knew the community would not necessarily be comfortable with the data.

[David Weaver]

Like most others, I want to see relevant results. Personally, if I saw 58 and 61 hardness results from a vintage saw, knowing what I know, I would've figured out what the problem was before publishing any information.

[Rob Streeper]

Like most others, I want to see relevant results. Personally, if I saw 58 and 61 hardness results from a vintage saw, knowing what I know, I would've figured out what the problem was before publishing any information.

Okay, lend me one of your vintage saws and I'll be happy to poke a bunch of little divots in the blade. Seriously, I tested under the handles to keep from messing them up. Don't forget that nobody, even the 'experts' commenting here, said anything about my testing under the handles, I brought it up, everybody else missed it.
This information is not intended for a rigorously peer reviewed scientific publication, it's intended to inform and provoke discussion among hobbyists.

I once had somebody angrily shout at me "You don't know everything...". My reply was "Yeah, I know but at least I'm working on it."

[george wilson]

I can't understand why this debate is STILL going on. I thought it had been resolved last night when Rob admitted that he made errors in his calculations,and checked the hardness of the Disston saws where he had hammered the metal,further hardening it.

I tested all kinds of tool steel for well over THIRTY YEARS with my Versitron hardness tester on the ROCKWELL C SCALE. NO COMPUTATIONS NEEDED to muddy the waters.

These were my findings: 1095 blue spring steel from Precision Steel Warehouse in Greensboro,N. Carolina,consistently tested at 52 RC. Old Disston saws checked out at 40 to 45 RC. There were 80 craftsmen working in Williamsburg,and I took care of all their saw filing. There were a large number of saws involved. I am a curious person,and enjoyed testing them. I tested all of my own saws and chisels too.

Regarding the reading Rob got from "custom maker #4,at 30 RC,the guy must be using tin cans for his blades because that is about the hardness of common MILD STEEL. I still don't know how that reading was possible. Rob,are there any MORE mistakes you have made in your calculations?

[george wilson]

I think Rob is to be commended for admitting that he did make mistakes. Not everyone is willing to do that. It shows humility in a person who does that. I wanted to make a separate post to make this thought clear.

I have made mistakes on this forum (not about technical matters),and have apologized. It is the thing to do when you foul up.

[Ray Gardiner]

Hi Rob,

Stick with it, and keep an open mind, listen carefully to the advice being offered, eventually the data will make sense.
I have a few old Disstons and an Avery 6402 hardness tester, I'll see if I can get some sensible results. in the next few days.. I can't actually test any 1095, since all the 1095 I've got is too thin 20thou and 15thou for the 150kg Rc test. And I don't have a RN Superficial tester.

Ray

[Rob Streeper]

George,

The mistake in the calculations had no impact on the hardness readings. It was in the data range that I used in the first table posted for calculating the standard deviation of the measurements. As such that mistake had no influence on the actual hardness readings.

The mistake on taking the measurements reported on the Disston saws related to the region of the saw plate tested, which I recognized and pointed out. Nobody else noticed the possibility for error here. I'll select one of my saws and test it nearer the tooth line and report the results but I'm not going to pepper the whole plate with indentations.

Something else I revealed here that nobody else commented on had to do with sample size. To have confidence in a set of measurements you need to do repeated testing. It helps you get a feeling for what reality is. Statisticians have developed mathematical tools for calculating how many tests you should perform to get a given level of statistical confidence. Unfortunately you can't do these calculations unless you have some numerical data to start with. I used the data in the tables to calculate that number of samples. If you want I'll post the output from the program I used but I think that it's a little heavy for this kind of thread. That being said the number of replicates needed to give a power of 80% with an alpha value of 0.05 is 38 replicate measurements. Thus I plan to take a new piece of 1095, test it 50 times and compare those readings to a piece of 1095 that has been hammered and then ground flat and compare all of those values to measurements of a piece of 1095 that has been hammered but not ground.

I don't dispute your experience and please extend to me the courtesy of not disputing mine. I have 30 years of professional experience in designing and executing high precision experiments across a large range of testing methodologies. In fact I am a certified quality auditor. All of that does not mean of course that I am immune from making mistakes in Excel calculations but I think you should consider it. I make saws too and if you want me to prove that I'll post some pictures though I hesitate because I don't want to violate the forum rules.

Cheers,
Rob

[george wilson]

Rob and all saw makers: You can save yourself trouble by buying STRAIGHTENED 1095 sheet,instead of buying it rolled up. It is difficult to get the curve out of 1095 that has been rolled. We did it by pouring boiling water over the blade and bending it backwards. It will not straighten by just bending it backwards cold. This is particularly aggravating when making crosscut and rip saws that have no back to hold them straight.

Rob,I will not dispute your readings if you have done them correctly. I already have applauded your admission of mistakes. I did disagree with your previously wrong numbers. Post reasonable numbers and I will not disagree. You seem to have done that by now,and I'd let the matter lay. With all due respect,it would have avoided this long discussion had you checked your work more carefully before posting.

It is not necessary to lecture me about the scientific method. I am well educated,and in high school scored in the 98th. percentile on a nation wide physics test about 1958. My experience in serious woodwork dates from 1954.

I wish now that this whole discussion just be left alone. I'm not going to keep this up for 10 more pages.

I have seen your saws,and,again,with all due respect, would rather not get into a discussion about them.

Below are pictures of ONE batch of planes,ONE batch of saws,and a couple of back saws I made for my own use. This might give you some perspective of what I did as toolmaker.

[John Stankus]

Does anyone have access to other testers, like the one George mentioned? I'd personally enjoy seeing the data of a test replicating Rob's test subjects on a different tester.

Of course David is right, all this probably matters very little in practice, but it does have good entertainment value.

I've been debating about chiming in, or keeping my mouth shut. I went to talk with my buddies up in our engineering department, and should be able to get access to their hardness tester (good ol' Rockwell C tester). Due to space it was moved off to another building on campus, but I should be able to access it if it is the appropriate test. I know in my industry days making the assumption that the properties were uniform over the whole article was not usually good. I wish this had come up before my brother had moved and I could grab some of the saws out of his foot locker full of flea market finds and due a spatial study of the saw plates as well as an era study. The other thought I had was doing a chemical analysis of the steels used in the older saws.(hey I'm a chemist, though I played a materials scientist for 15 years in industry, before going back to teach)

John

[David Weaver]

This information is not intended for a rigorously peer reviewed scientific publication, it's intended to inform and provoke discussion among hobbyists.


It should at least be remotely correct then, and not laced with commentary casting aspersions at other saw makers. Right now the results are not relevant. The trouble is you feel that it is someone else's responsibility to tell you how to correct your results, but that is not the case. We have gone from a few of us knowing your results were not correct, to now you've stated that you won't test the part of the saw where the hardness actually matters, and you have stated that the saws can be filed (and for certainty, if that's the case, they are no harder than 10 points or more below the hardness of the file, or you'd have certainly remarked that the saw ruined the files).

Pat suggested that it was OK as long as the relative comparison was relevant, but it doesn't even meet that standard. The nominal numbers are not correct, they should be corrected before you continue on with any assertions, and then the relative comparisons would be better. Posting information that is incorrect (or of little value since the value at the handle doesn't mean much) and then sticking hard to it doesn't do anything but mislead people. Those who know better but who are too polite to engage in a factual disagreement will pass it by and the rest of the group who could believe anything you post will be misled. At the same time, the simple fact is that you are selling items directly in competition with the folks you are casting aspersions about (even if you don't name them), and based on data that is either irrelevant (if the saws are in fact that hard under the handle) or incorrect. You have to hold yourself to a higher standard when you have $kin in the game.

If you are going to assert the hardness of saws, you need to do it near the teeth. If you can't spare the saws to test them relevantly, then you shouldn't post the results. .

When I have pushed you to correct your information and figure out why it's not correct, it's been perceived as a personal slight, and it is in fact not. I have never seen nor heard of you, I care only about whether the results are correct and the conclusions your drawing can be drawn from them. It is a shame when discussion on this forum or any other forum about factual items and their usefulness has to be truncated because people get offended when someone says they are wrong and supposes reasons why they may be. I would be perfectly comfortable if you had suggested my assertions were wrong and maybe I didn't have enough experience in making them if you could prove otherwise. I don't get offended in that case, I learn, and then I give better information - we learn by sharing information, but we only learn efficiently if we share valid information.

[David Weaver]

I've been debating about chiming in, or keeping my mouth shut. I went to talk with my buddies up in our engineering department, and should be able to get access to their hardness tester (good ol' Rockwell C tester). Due to space it was moved off to another building on campus, but I should be able to access it if it is the appropriate test. I know in my industry days making the assumption that the properties were uniform over the whole article was not usually good. I wish this had come up before my brother had moved and I could grab some of the saws out of his foot locker full of flea market finds and due a spatial study of the saw plates as well as an era study. The other thought I had was doing a chemical analysis of the steels used in the older saws.(hey I'm a chemist, though I played a materials scientist for 15 years in industry, before going back to teach)

John

It would be interesting to see what you come up with, John, if you're able to do that. We (a friend and I) offered up only chisels and my friend may have offered some plane irons to the hardness tester. I didn't consider saws at the time, but having hammered them, it seems to me as that you can feel the steel giving if it's not bedded perfectly on the hammer stroke, that such a thing could be a problem on a C tester. Any give in the metal or influence from a backer will be troublesome.

I'd be interested in the chemical analysis across brands, too. I suspect that for the top quality saws, we'd see a lot of the same things both in hardness and in composition of the metal.

[Rob Streeper]

I've been debating about chiming in, or keeping my mouth shut. I went to talk with my buddies up in our engineering department, and should be able to get access to their hardness tester (good ol' Rockwell C tester). Due to space it was moved off to another building on campus, but I should be able to access it if it is the appropriate test. I know in my industry days making the assumption that the properties were uniform over the whole article was not usually good. I wish this had come up before my brother had moved and I could grab some of the saws out of his foot locker full of flea market finds and due a spatial study of the saw plates as well as an era study. The other thought I had was doing a chemical analysis of the steels used in the older saws.(hey I'm a chemist, though I played a materials scientist for 15 years in industry, before going back to teach)

John

That sounds good John, please let us know what they say. In particular, ask them about the appropriate use for the Rockwell C and N scales for materials of the hardness that we are discussing here.

[Rob Streeper]

Rob and all saw makers: You can save yourself trouble by buying STRAIGHTENED 1095 sheet,instead of buying it rolled up. It is difficult to get the curve out of 1095 that has been rolled. We did it by pouring boiling water over the blade and bending it backwards. It will not straighten by just bending it backwards cold. This is particularly aggravating when making crosscut and rip saws that have no back to hold them straight.

Rob,I will not dispute your readings if you have done them correctly. I already have applauded your admission of mistakes. I did disagree with your previously wrong numbers. Post reasonable numbers and I will not disagree. You seem to have done that by now,and I'd let the matter lay. With all due respect,it would have avoided this long discussion had you checked your work more carefully before posting.

It is not necessary to lecture me about the scientific method. I am well educated,and in high school scored in the 98th. percentile on a nation wide physics test about 1958. My experience in serious woodwork dates from 1954.

I wish now that this whole discussion just be left alone. I'm not going to keep this up for 10 more pages.

I have seen your saws,and would rather not get into a discussion about them.

George,

This is off the topic of saw plate hardness but it is still important generally here. I have received several PM's that are supportive of my efforts and none that are critical. As we all know the majority of comments posted openly on this thread have been critical with the least insulting of them asserting my incompetence.
I think it a matter of concern to the community that uses this forum that there is at least a portion of the population that does not feel free to share their thoughts openly. I propose that everybody think a little while about why some here don't want to comment openly and what factors have impelled them to feel that way.

[george wilson]

Your NIST chart already gives you the information about what is appropriate to test with both S and N scales. Why is it necessary for you to ask for more answers?

You got the same sort of replies in the Woodworking Forums in Australia,where you previously posted your incorrect material.

You got PM's of support? None that were critical,even of your incorrect findings? Many people may not know what the truth might be when a bunch of numbers and charts are presented to them. We all could fall into that category when presented with important LOOKING numbers about subjects we are not familiar with.

I posted a few pictures of my work back in post 106.

[David Weaver]

George,

This is off the topic of saw plate hardness but it is still important generally here. I have received several PM's that are supportive of my efforts and none that are critical. As we all know the majority of comments posted openly on this thread have been critical with the least insulting of them asserting my incompetence.
I think it a matter of concern to the community that uses this forum that there is at least a portion of the population that does not feel free to share their thoughts openly. I propose that everybody think a little while about why some here don't want to comment openly and what factors have impelled them to feel that way.

When you're done, perhaps you could do a review of what caused you to come up with numbers that are not credible, at least not in the context they are being used, and then describe what it is that caused it.

[george wilson]

I have been retired since 2009. I left a lot of things at the shop. The new toolmaker has entirely different aims(seems to be trying to turn the shop into a backup silver smith shop). He has thrown out a lot of things that he should have kept.


WILL SOMEONE PLEASE LOCK THIS SILLY THREAD?

[Rob Streeper]

When you're done, perhaps you could do a review of what caused you to come up with numbers that are not credible, at least not in the context they are being used, and then describe what it is that caused it. When you assert a saw is hardness 61 and is still sprung and as you have finally admitted, can be filed, you have to expect people will suspect that you are a bit green on what you're trying to assert.

If I am new at something and you torpedo my conclusions and assert that I have a problem that I should figure out, I'm not offended by that. I'd be much more bothered if I gave out bad information due to inexperience or lack of analysis to control quality of the results or how they are implied. For several years, I derided warren mickley because I thought he was wrong about plane double irons. I turned out to be 100% wrong. I'm embarrassed by that, and I should be, and the result of being critical of what I discuss is that the accuracy has improved. Warren told me and others that we didn't know enough about what we were talking about to draw the conclusions, and he could tell that by the conclusions we drew. As a beginner, I was probably offended, I don't recall, but in the end for warren to insist that I and others need more experience was correct.

There's really no other way for people who know the answer to the question of whether or not your data was bad to go about it other than try to figure out why you were so quick to dismiss what is already known in favor of your data. We have to assume you either have a motive for it, you're avoiding analysis (why didn't you admit right away that the saws could be filed as any others?) too prevent their conflict with your numbers, or that there is something that you don't know about the testing methods you don't use. There's no way around it other than to sit back and allow false assertions to be drawn on errant data. Is that a good conclusion?

David,

We're over 100 posts now on this thread and nobody has posted any data that contradicts mine. George speaks from his experience but again there is no data - just words.
Ask yourself why is there no data? Handsaws have been produced for centuries. You might think that somebody somewhere would have done a study, even a limited one like mine, but I have never seen one.
I have identified all of the technical errors in my results, the commenters here have simply said 'you're wrong' without giving any data to support their positions.
George maintains that the Rockwell C scale is the only scale to use. However all of the cited references, most of them mine, point out that for materials of this hardness range C scale only gives valid results with materials of 0.032" or thicker. For thinner materials the N and T scales are specified as appropriate.
Let me lead you through my reasoning.

First open this chart: http://qs-hardnesstester.com/hardness-minimum-thickness.html

Now look under the heading of the C scale on the left side of the table and you will see that, for materials of the hardness range were talking about here, the minimum thickness is 0.034" for Rc45 and 0.032" for Rc 52. Thus the practical lower limit of C scale testing of saw plate materials is 0.032" to 0.034". For thinner materials the C scale is not appropriate.

Now look at the 30N scale of the chart. You will see that for materials of at least Rc hardness of 47 the minimum thickness is 0.22". The minimum thickness for materials of Rc57 is 0.020". Materials of these thicknesses are below the application range of the C scale.

All of the saws I tested are 0.020" thick or thicker, meaning that the N30 scale is the scale to use and that is what I did. For thicker plates I cross checked on the N45 materials and in the past I have used the C scale for saws that are thicker than 0.032", including the soft one made by Maker 3.

So, if you're asserting I'm wrong because I've used the wrong hardness scale you're indicting not only my competence but the competence of the entire industry of professionals in this area.