Panel Saw Tensioning

[D_W]

I can spot an engineer on a forum from a mile away, and not only this one, because this kind of thing applies (attempting to set up a straw test and draw a universal conclusion) and happens all the time.

I don't think your saws are representative of a quality lot of saws from the 1900s or so. I don't know anything about groves saws other than that I have a groves rip saw, and it's not floppy at all. Anything made after 1935, I'd question the quality - saws were in a downward spiral of cost cutting running from circular saws.

Maybe the saw market is less uniform in the UK. In the US, 80% of the saws directed toward professionals were disstons, probably the bulk begin no 7s and D8s. There are plenty of second line saws to compare them to. I'm less familiar with what is first and second or third or worse in sheffield.

I have no clue what the context of your use is, but the only way you get a real sense for saws is to use them for everything in several projects (rough ripping, fine ripping, resawing, etc). If you're using them as an ancillary tools to power tools, you can tolerate a lot of substandard characteristics, just as people can easily forget about how excellent a double iron plane is if they only use their bench planes to smooth already machine planed wood.

At any rate, if you shoot for drawing a universal conclusion before you have more exposure than the saws you list, nobody will have any regard for the conclusion, anyway. The forums are full of people who assert something that is the opposite of what people have known for a hundred or more years. Most people won't use saws enough to know what's correct or not, anyway, you may become a guru and have your results parroted on the other guru's blogs. If I were you, and I saw all of the old texts discussing tension, I would spend more time trying to figure it out and less trying to say it doesn't exist.

 

[D_W]

I have been on the wrong side of these things, by the way, stating that old text was vague and what was behind it was hot air. On oilstones, on double iron planes, etc.

That's partly what makes me less apt to just write off older common knowledge even though I haven't tensioned a saw myself yet (but do have similar hardness saws with similar cross sections that are way different in stiffness - which would be due to ____ if it's not tensioning?).

 

[Trafalgar]

Why would Disston hire, train, and pay the skilled labor required to do this step in the manufacturing process if it weren't necessary? For $hits and giggles?

If it weren't necessary I'm thinking somebody would have noticed.

 

[D_W]

Why would Disston hire, train, and pay the skilled labor required to do this step in the manufacturing process if it weren't necessary? For $hits and giggles?

If it weren't necessary I'm thinking somebody would have noticed.

in 2015, they could target white collar retirees with a scheme like that, but not in 1900.

 

[Trafalgar]

If there was (is) nothing to tensioning then it was one of the biggest marketing ruses of all time -- complete with employees doing faux work but being paid real salaries, etc.

Or maybe they really were that stupid back then. Nah, doubtful.

 

[Carl P]

I've found this discussion very interesting and while I know nothing about it, it seems to me that the term 'tension' may be something of a misnomer, possibly stiffening may be more descriptive. Also, as to evidence, unless you can tension a saw yourself, no matter how many saws you may try, because of the variables, all you can really prove is one saw is better than another as in 'the plural of anecdote is not data'. Some kind of test would be essential, grinding down the saw plate will, naturally, make it more floppy, it being thinner and all, and may change the hardness, so is somewhat pointless. The only meaningful test I can see would be to make two identical saws, but with only one tensioned. Alas this is very much beyond my meagre capabilities, so I'll leave that to someone with actual talent, until the results of such a test are known, most of what can be said on the matter seems so much unsupported hot air.

Cheerio,

Carl

 

[AndyT]

I've enjoyed it too. I'm no physicist or engineer, but I now have a new mental image of what might be going on inside an old saw blade, which I think fits in with what used to be done in manufacturing.

First, think of a concrete beam. It has some resistance to bending, but not much. When it is supported at the ends and loaded in the middle, it sags down, with the upper parts in compression and the lower parts in tension.

To make a stiffer beam, you can insert steel rods, and pull on opposite ends while the concrete sets around them. The result is a pre-stressed beam. The pull (stress) in the steel does not move anything but it is real, and helps counteract the stretching that would happen to the underside when a load is applied. So we have a concept of an internal stress, which helps keep the beam straight (makes it stiffer).

I think this is what the skilled saw smiths were doing. By hammering, some of the steel was put in a state when it would have stretched out longer, but could not, because it was constrained by other areas of steel around it. They made these small areas of different stress across the whole saw, in a balanced way.

It's extra complicated because forces within the steel on one face of the plate need to be balanced by forces on the other face, or else the plate would bend - and the plate is very thin.
It's extra complicated because the steel was made in small batches and was not as consistent in its internal chemistry and structure as modern steel can be.
It's extra complicated because the saw went through repeated grindings and heatings, which would have affected its stiffness and elasticity.
It's extra complicated because the areas wanting to expand are small and not visible, as are the unstretched areas constraining them.

So, in my imagination, I now see the saw plate as something like a lattice structure (think of a railway bridge) where short elements are variously in tension or in compression, but the whole structure ends up being strong, with the right balance between stiffness and elasticity.

And the reason that this tensioning is not done any more, is that modern strip steel is so much more homogenous and reliable that it is not needed.

Does that make sense to anyone else?

 

[Trafalgar]

Carl P --- "the plural of anecdote is not data" Priceless!

 

[Trafalgar]

Shot peening:

https://en.wikipedia.org/wiki/Shot_peening

"Shot peening is often called for in aircraft repairs to relieve tensile stresses built up in the grinding process and replace them with beneficial compressive stresses. Depending on the part geometry, part material, shot material, shot quality, shot intensity, and shot coverage, shot peening can increase fatigue life up to 1000%.[2]

Plastic deformation induces a residual compressive stress in a peened surface, along with tensile stress in the interior. Surface compressive stresses confer resistance to metal fatigue and to some forms of stress corrosion.[1] The tensile stresses deep in the part are not as problematic as tensile stresses on the surface because cracks are less likely to start in the interior."

With saws perhaps tensile stress is a desired characteristic whereas for aircraft bodies, wings, etc. it is not. By striking the saw plate on both sides, equally, rather than relieving tension it increases it.

Things do apparently happen when metal is struck and these things are not necessarily obvious to the naked eye. That is, unless a big crack opens up in an airliner at 35,000 feet.

I think the old guys knew what they were doing. Maybe it isn't necessary these days because of the raw material and processes available but those old saws, when in good fettle, do seem to have a little something 'extra' that modern saws do not -- not a huge difference, but there is something there.

 

[D_W]

I've found this discussion very interesting and while I know nothing about it, it seems to me that the term 'tension' may be something of a misnomer, possibly stiffening may be more descriptive. Also, as to evidence, unless you can tension a saw yourself, no matter how many saws you may try, because of the variables, all you can really prove is one saw is better than another as in 'the plural of anecdote is not data'. Some kind of test would be essential, grinding down the saw plate will, naturally, make it more floppy, it being thinner and all, and may change the hardness, so is somewhat pointless. The only meaningful test I can see would be to make two identical saws, but with only one tensioned. Alas this is very much beyond my meagre capabilities, so I'll leave that to someone with actual talent, until the results of such a test are known, most of what can be said on the matter seems so much unsupported hot air.

Cheerio,

Carl

I think you're off base still on one thing - if nobody knows how saws were tensioned, then your test wouldn't prove anything.

I suggested finding two stiff saw plates, one a hundredth thicker than another or so so that you can take the thicker plate and grind it down to the size of the thin plate. Tensioning is in the surface and that would remove it. That should leave someone with two saws that were formerly tensioned of different sizes, but the ground saw would match the first saw and presumably be floppy.

What I suggested would be more relevant - find two saws that are the same cross section in metal (i.e., the plate sizes are about the same) and that are the same hardness, search until one is floppy and one is not. If those things are all true (same plate size/thickness and same hardness), what else would someone think is occurring. These are water or oil hardening steels we're talking about, there's not mystery making them stiffer other than physical working of the steel.

OR, to my knowledge, there are a very few remaining sawsmiths in japan making saws entirely by hand. You will read in some places that there are non, but stan covington on another forum knows a guy who makes about 20 saws a year entirely by hand. Ask the guy about tensioning, see what he says. As charles says, you can believe he's tensioning a saw for no reason, or you can take the much more probable view that he knows what he's doing and you can learn something from him.

A "talented" amateur reproducing what took the saw industry a century to perfect in manufacturing is not a good test. First, the amateur has little chance of getting things as right as a history of progressive improvement, and second, they have zero ability to do tensioning the way it was done in manufactured saws (which, of my use, were better and more consistent than the more handmade saws from an earlier era).

 

[D_W]

It's suggested that it's "not needed" due to the homogeneous 1095 available now, but to assume that the current batch of steel is a lot better or more homogenous than the house steel disston would've used is suspect. It was probably made with a similar process.

I've got a roll of 1095 that is 4 inches tall and 0.042" thick. It's stiff compared to unhardened steel, but not stiff compared to a saw made in 1900 or so.

I wouldn't trade a 1900 D8 for a modern boutique carpenter's saw, even without the price difference. Modern backsaws are a better bet because there's no need to tension a plate held in a saw back.

When someone doesn't notice the difference between various saws, I wonder how much they're sawing - dovetails? sticking? crosscutting boards here and there? It's sort of like someone who only smooths weighing in on the important aspects of a fore plane.

 

[AndyT]

It's suggested that it's "not needed" due to the homogeneous 1095 available now, but to assume that the current batch of steel is a lot better or more homogenous than the house steel disston would've used is suspect. It was probably made with a similar process.

Ok, how about, "now that steel strip is sufficiently consistent in its characteristics to make a decent saw, further treatment by hand hammering would cost more than the market would bear, given the very small number of new saw purchasers who would appreciate the marginal improvement and be willing to pay the large amount it would cost, given that the skill and knowledge needed to tension saws by hand have virtually disappeared."

 

[Trafalgar]

That pretty well sums it up Andy. Plus, people already pay a lot of money for boutique saws that have not been hand hammered. Assuming one has all the orders one can handle, has apparently distinguished his firm from the crowd, there would be nothing much to gain.

 

[Rhyolith]

To me, what Andy described (about the lattice structure) makes sense. There are other examples where there are forces present within a single piece of steel, such as plane soles where the annealing process is used to remove/manipulate them for better overall performance of the tool. I do not see why it is not possible to manipulate them to the overall benefit of sawing in a hand saw too. We know that hacksaws benefit from a stretched/tensioned blade so its not much of a jump to assume panel saws would benefit in a similar way (why wouldn't they?); particularly if the steel was not very uniform to start with. So as far as I am concerned it probably could have been done and defiantly would have been beneficial.

The only question in my mind is can stretching/tensioning the blade in the manner that a hacksaw frame does be have been done in a Panel saw? to answer that I/we need to ask How? And that can only be answered by a actual saw maker. So has anyone got any contacts on this front?

Further: I really think the argument of whether the type of stretching/tensioning mentioned above (like a hacksaw) of benefit is a red herring... of course it would be just try sawing with an un-tensioned hacksaw to see that :? The debate is whether that is what "tensioning" means in a saw makers context (what it means in physics is irrelevant, science is a language on to itself) and how it was done.

 

[Trafalgar]

I guess the forum's next move is to try to identify a maker of new saws who is tensioning by hand-hammering. Might be a tall order.

I suppose an old 'saw doctor' would suffice as well. We have all Bob Smalser has written on the subject unless he could be encouraged to post on this forum and in this thread.

 

[MIGNAL]

I've tried tapping on my saws with the knuckle. All the new saws ring like a church bell (even the back saws), my old saws don't!
Unfortunately I don't have enough examples of old saws to make it a meaningful test. It's probably not a meaningful test anyway but if someone catches you listening to the sound of an old saw at least you'll be termed eccentric. They just might use a slightly stronger term of phrase though.

 

[D_W]

It's suggested that it's "not needed" due to the homogeneous 1095 available now, but to assume that the current batch of steel is a lot better or more homogenous than the house steel disston would've used is suspect. It was probably made with a similar process.

Ok, how about, "now that steel strip is sufficiently consistent in its characteristics to make a decent saw, further treatment by hand hammering would cost more than the market would bear, given the very small number of new saw purchasers who would appreciate the marginal improvement and be willing to pay the large amount it would cost, given that the skill and knowledge needed to tension saws by hand have virtually disappeared."

That's pretty fair. I would sum it up as if the customer can't tell the difference, then why bother with the expense?

 

[Cheshirechappie]

It's suggested that it's "not needed" due to the homogeneous 1095 available now, but to assume that the current batch of steel is a lot better or more homogenous than the house steel disston would've used is suspect. It was probably made with a similar process.

Ok, how about, "now that steel strip is sufficiently consistent in its characteristics to make a decent saw, further treatment by hand hammering would cost more than the market would bear, given the very small number of new saw purchasers who would appreciate the marginal improvement and be willing to pay the large amount it would cost, given that the skill and knowledge needed to tension saws by hand have virtually disappeared."

That's pretty fair. I would sum it up as if the customer can't tell the difference, then why bother with the expense?

Or perhaps, if the customer can't tell the difference, there isn't one.

One thing that does occur to me is that if somebody had discovered a process that made thin metals stiffer, it would have had huge commercial potential in many other fields of engineering. Aircraft work, for example, in which it might have reduced or eliminated a great deal of the expense and trouble of fitting stiffening ribs, and eliminated their additional weight. Automotive body panels are another application that springs to mind. Engineers (and especially business accountants) are always looking for ways to do things using less material. That no such process exists (well, to my knowledge, anyway) suggests that the difference it made to a sawblade wasn't especially dramatic.

Also bear in mind that sawblades had to be hammered to straighten them after heat treatment, the processes at the time being incapable of producing flat, hardened and tempered stock, which is something the steel mills can do now, as Andy says. The understanding of material science, stresses and metallurgy was even more imperfect in the late 19th century (indeed, it was pretty much in it's infancy then), so the use of words in a technical context tended to be very loose by modern standards. They very probably looked at springs and thought of 'tension' - hence phrases like 'proper tension, spring and character'. Just meant 'springy' to them - which is, after all, what you're after; flat, and springing back to flat if deflected.

 

[D_W]

I'll pose a scenario:

* A 1900s customer who worked professionally with saws (without the help of power tools, often on site work) could tell the difference and would pay for it (despite having very little disposable income)
* A modern hobbyist can't tell the difference

So the odds on favorite is .....to start by assuming the hobbyist is correct?

I can tell the difference, and i'm only a hobbyist. But I use my saws for everything - ripping, crosscutting, resawing. And it is in ripping and resawing that the differences become so stark, and for the same reason, conversion of a saw designed for crosscutting isn't always a great choice for creating a fine-toothed ripping saw.

We have no chance of recovering what disson knew about tensioning handsaws at a commercial level. We may have some hobbyist makers or boutique makers who would like to recreate a different process using hammers, something that could've been done prior to perfecting rollers (and was probably done).

We're talking about the same audience who pretty much was universally sold, for a while, on a plane having an extremely thick single iron, bevel down, as being tops for functional planing (but for some reason, planes weren't made that way 125 years ago).

The buying audience these days isn't very demanding, especially the people actually spending money. Those are the folks who are just coming into the hobby after making a living elsewhere and having discretionary money to spend.

 

[Cheshirechappie]

Ah - so you can tell the difference, but the rest of us modern hobbyists can't....