Panel Saw Tensioning

UKworkshop.co.uk

Help Support UKworkshop.co.uk:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
MIGNAL":2q6aa1ft said:
http://www.crosscutsawyer.com/viewtopic.php?f=4&t=750

Interesting.

One question it doesn't answer is what difference there is between a blade so rolled and one not. If any.
 
MIGNAL":2frmzlpo said:
http://www.crosscutsawyer.com/viewtopic.php?f=4&t=750
Yes, that is interesting! :) Its still a bit vague though, do you have an account on that forum? maybe could ask for some details on the process of tensioning? Might finally answer some of the questions rolling round this thread (thank you all for the debate and input by the way, I found it very good read :D ).
 
No account. I wasn't even aware that a forum existed specifically for crosscut sawing!
It has been an interesting thread. I'd certainly heard of 'tensioning' saw plate but I've never known anything about it. I'm not sure that I've ever believed it existed. In truth I probably haven't paid any attention to it. Planes are my thing. Saws are for my next life.
Someone needs to join and PM that French chap Diabolo. No doubt he'll have more to say on the matter, especially since he seems to be a saw maker, even if not in a commercial capacity.

There's also mention of it on the following, especially the final post on the difference in steel. (post 16). Stephen Shepherd is the guy who wrote a very good book on traditional Finishes and a very good book on Hide glue.

http://www.fullchisel.com/blog/?p=137


Assuming this hammering did take place (and I've no idea either way) when did it stop? Roughly, when did saw plate stop getting hammered and when did it become rolled? Is a Disston from eg. 1935 inferior?
Presumably it wasn't only Disston that was doing this and it applies to other makers of quality saws.
 
Aren't Dissiton still about? I thought they were?

Also I thought Lie-Neilsen claimed to tension saws (have not checked that yet)... i vaguely remember someone i knew having a problem with a small LN tenon saw and it being fixed via tensioning... but I cannot remember the details so don't rely on that account.

Thanks for the link Mignal, will read it tomorrow... bed time for me should have been 40mins ago ;)
 
Yes - it's all speculation at the moment. Lots of questions, no answers (though it seems some are more certain about things than others).

I suppose in the fullness of time, somebody will work it out. After all, the skills of making moulding planes disappeared. I dare say some said the skills had gone for ever, and we'd never replicate what the old makers could do. Then along came Clark and Williams, Matt Bickford, Philly, Caleb James.....not to mention John Whelan and Ted Ingraham, and probably several others. If it's been done before, it can be done again.

I still think the clincher will be if somebody produces two identical saws, one 'tensioned' and one not, and demonstrates a significant difference. Until then, it's all hot air.
 
Cheshirechappie":2vg1sge1 said:
I still think the clincher will be if somebody produces two identical saws, one 'tensioned' and one not, and demonstrates a significant difference. Until then, it's all hot air.
I see where you coming from with this, but I think there are too many variables for a test like that to be meaningful. For example: User's: skill, psychology, preference, physical capabilities, past expierence. Saw's: Metallurgy, past treatment, overall age, sharpness, size, weight.... etc. There are just to many that you cannnot be certain about in an old saw alone (prove me wrong here) and thats not even including the huge influence the user has.

Plus as we don't actually know for certain what tensioning is yet, so how can we know one of the saws is tensioned (properly)? Particularly as all the examples seem to be 100 years old. So for me the clicher is definalty finding that out.
 
Rhyolith":31kfsuvz said:
Cheshirechappie":31kfsuvz said:
I still think the clincher will be if somebody produces two identical saws, one 'tensioned' and one not, and demonstrates a significant difference. Until then, it's all hot air.
I see where you coming from with this, but I think there are too many variables for a test like that to be meaningful. For example: User's: skill, psychology, preference, physical capabilities, past expierence. Saw's: Metallurgy, past treatment, overall age, sharpness, size, weight.... etc. There are just to many that you cannnot be certain about in an old saw alone (prove me wrong here) and thats not even including the huge influence the user has.

Plus as we don't actually know for certain what tensioning is yet, so how can we know one of the saws is tensioned (properly)? Particularly as all the examples seem to be 100 years old. So for me the clicher is definalty finding that out.

Just to clarify, I didn't mean that somebody compares two vintage saws. What I meant is that somebody takes some new spring steel and makes two new saws from it, one 'tensioned' and one not. If there is a noticeable difference between the two in performance, or stiffness, or waviness of the toothline when it gets hot, or whatever, then it's clear that the 'tensioning' has done something. The maker would also know exactly what they had done differently with the two saws, and could describe the process, if not the exact metallurgical effect.

I suspect there may be a bit of 'suck it and see' involved in this. The links Mignal posted suggested some sort of rolling process rather than hammering, but the point of controlled testing is that you try something on one workpiece and compare with an identical workpiece not so treated to see what the effect is. It may be that quite a few controlled tests along these lines are needed to work out which is the most effective - or it may be that it demonstrates that none of them are particularly effective.

A fun project for somebody!
 
I'm away from books again, but perhaps someone could check what our best saw historian, Simon Barley has to say on this. I think it's that final tensioning was only done on the top grade of saws, not on the second or third grade. That would fit with the collective idea that it's something not everyone would notice or appreciate.
 
Simon describes 13 steps in the manufacture of saws.

1) Preparing the flat sawplate.
2) Paring, toothing and marking.
3) Hardening and tempering.
4) Grinding.
5) Smithing.
6) Further grinding, glazing and smithing.
7) Rubbing.
8) Blocking.
9) Stiffening.
10) Setting and sharpening.
11) Fitting the back.
12) Handling.
13) Packing and despatching.

The most relevant section is probably section 5, Smithing. Simon writes, "The saw plate might at this stage to the uneducated eye look flat, but it contained multiple and almost invisible irregularities which the trade called 'fast' and 'loose'. These were corrected by a saw maker's hammer that had two faces at right angles to one another and which produced the marks shown in figure 1.13 [The marks look rather like chequer-plate, for those familiar with it.] The sheet also had to be hammered to produce stretching of the edge and hence "tension", in order to increase its ability to return to straightness when bent. Flatness and tensioning were achieved by the smith's hammering in one place after another, and on both sides, constantly checking for the right result using a sawmaker's straightedge (figure 1.14)."

Under section 9, Stiffening, he writes, "Although there is no mention of this stage in the statement of prices, it appears always to have been part of the manufacturing sequence in the Sheffield saw industry. It involved heating the sawplate up to a fairly cool red heat, thereby restoring the flexibility that had been diminished by rubbing and blocking. There are indications in one manufacturer's advertising material from the 1970s that stiffening was reserved for the best of their four qualities of handsaws. A stiffening trough was also referred to in the Wilson saw stock inventory of 1774, but there is no other information to enable us to know how or at what stage it might have been used at that time. In the account of saw manufacturing at Spear and Jackson's works in 1861, stiffening was described as coming after setting."

Which suggests that there was a process of hammer tensioning of some sort, carried out during the plate flattening stage (or one of the plate flattening stages, more accurately). How much difference it made, and whether it was done to all saws, isn't stated, and figure 1.14 shows smiths working on circular saws, not hand saws.

However, the description of 'stiffening' is very confusing. Heating a near-finished saw blade to a red heat, even a fairly cool one, is most likely to anneal (soften) it, and certainly won't stiffen it. It may have been a way of recovering plates that wouldn't flatten easily by annealing them to take out all the hammer stresses put in, before re-hardening and tempering, and then going through the processes again - but that's just me speculating.

Again, more questions....
 
AndyT":j2spp7di said:
I'm away from books again, but perhaps someone could check what our best saw historian, Simon Barley has to say on this. I think it's that final tensioning was only done on the top grade of saws, not on the second or third grade. That would fit with the collective idea that it's something not everyone would notice or appreciate.

That goes toward my earlier comment, too, about similar hardness saws not being similar in stiffness. I've not noticed the floppiness I described other than on unmarked saws, but I have to admit I don't have any mid-line saws. Just saws that are old enough to be pre-roller tensioning.

The japanese describe tensioning as a separate step from straightening, also, and at least in one place, I saw it described as ensuring that the plate stays straight.
 
Cheshirechappie":1foiz021 said:
Rhyolith":1foiz021 said:
Cheshirechappie":1foiz021 said:
I still think the clincher will be if somebody produces two identical saws, one 'tensioned' and one not, and demonstrates a significant difference. Until then, it's all hot air.
I see where you coming from with this, but I think there are too many variables for a test like that to be meaningful. For example: User's: skill, psychology, preference, physical capabilities, past expierence. Saw's: Metallurgy, past treatment, overall age, sharpness, size, weight.... etc. There are just to many that you cannnot be certain about in an old saw alone (prove me wrong here) and thats not even including the huge influence the user has.

Plus as we don't actually know for certain what tensioning is yet, so how can we know one of the saws is tensioned (properly)? Particularly as all the examples seem to be 100 years old. So for me the clicher is definalty finding that out.

Just to clarify, I didn't mean that somebody compares two vintage saws. What I meant is that somebody takes some new spring steel and makes two new saws from it, one 'tensioned' and one not. If there is a noticeable difference between the two in performance, or stiffness, or waviness of the toothline when it gets hot, or whatever, then it's clear that the 'tensioning' has done something. The maker would also know exactly what they had done differently with the two saws, and could describe the process, if not the exact metallurgical effect.

I suspect there may be a bit of 'suck it and see' involved in this. The links Mignal posted suggested some sort of rolling process rather than hammering, but the point of controlled testing is that you try something on one workpiece and compare with an identical workpiece not so treated to see what the effect is. It may be that quite a few controlled tests along these lines are needed to work out which is the most effective - or it may be that it demonstrates that none of them are particularly effective.

A fun project for somebody!

The trouble with your test is that it can only prove that tensioning is successful if someone who is not skilled in it can find a result. Your initial hypothesis suggests or at least implies that you think there's a great chance that it doesn't do anything. your test can't prove with any reliability that it doesn't stiffen a saw plate or provide some other benefit, because failure of someone who has no clue what their doing isn't proof that something doesn't work.

The best tensioned saws (with more language coming in later posts that it was reserved for higher lines of saws just supporting what I've already noticed, but you discount, anyway) were done with rollers. that's my experience. Saws like a Disston #12 or even the older D8s (which are wonderful saws, despite their ugly appearance and gigantic handles - unequaled by anything anyone is making now).
 
Very late to this thread but just wanted to go back to something way back on the first page which might be critical.

Cheshirechappie":3etfnz8i said:
Edit to add - the only way I can think of to make a sawblade stiffer (other than to make it a backsaw) is to make it from thicker material to start with. Stiffness of flat pieces of material is proportional to the cube of their thickness, so not much thicker makes it significantly stiffer. Conversely, not much thinner makes it significantly more flexible.
How about increased harness? As might be created by work-hardening the steel.

This is a well-understood process among traditional panel beaters and the steel used is quite similar. Edit: also with the making of steel drums. I can't imagine some hardening not occurring when hammering a saw, although the magnitude of the effect is likely a lot less since the changes in shape are so much more limited.

I agree incidentally that the use of the word tension might be a red herring, older written sources are a minefield when it comes to this sort of thing both with regard to modern technical usage of the same term and how it might be used in day-to-day speech. However, I'm not convinced there's no internal tension in a well-made traditional saw plate, sorry :oops:
 
Carl P":3ksnwalt said:
as in 'the plural of anecdote is not data'.
That's a good one, love it!

But actually, it can be. Multiple anecdotal accounts are a form of data, they can't simply be discounted particularly when there is some consistency. I'm not saying we accept it blindly, it must be confirmed by further testing of course. But that's not to say it is being discounted as data, after all scientific test data must also be confirmed.
 
ED65":1dzsr95v said:
But actually, it can be.

Presuming my comments about the stiffness of actual saws are being called an anecdote (as well as the writing - I don't follow that much of historical writings, etc, you have to know if the source was a writer turned toolmaker or a toolmaker turned writer, etc).

My comments are a data point (as are any others where someone has actual experience), and not an anecdote.
 
ED65":1om23m6c said:
Very late to this thread but just wanted to go back to something way back on the first page which might be critical.

Cheshirechappie":1om23m6c said:
Edit to add - the only way I can think of to make a sawblade stiffer (other than to make it a backsaw) is to make it from thicker material to start with. Stiffness of flat pieces of material is proportional to the cube of their thickness, so not much thicker makes it significantly stiffer. Conversely, not much thinner makes it significantly more flexible.
How about increased harness? As might be created by work-hardening the steel.

This is a well-understood process among traditional panel beaters and the steel used is quite similar. Edit: also with the making of steel drums. I can't imagine some hardening not occurring when hammering a saw, although the magnitude of the effect is likely a lot less since the changes in shape are so much more limited.

I agree incidentally that the use of the word tension might be a red herring, older written sources are a minefield when it comes to this sort of thing both with regard to modern technical usage of the same term and how it might be used in day-to-day speech. However, I'm not convinced there's no internal tension in a well-made traditional saw plate, sorry :oops:

You are quite right that the phenomenon of work-hardening exists. Indeed, it was used in the 18th century to make cheap saws from sheet iron - the soft iron was hammered until it became springy. One of the contemporary writers (Moxon, if memory serves) warned against the purchase of such saws, stating that they were much inferior to steel ones.

Panel beaters normally work with low-carbon steels rather than the high-carbon steel used to make springs and saws, but they do encounter work-hardening, having to anneal work (heat it to red heat and allow to cool slowly) before they can carry on. Saw steels are a bit different, in that they are deliberately heat-treated to a fairly high degree of hardness before any hammering takes place, so there's much less scope for increasing hardness by hammer work. It will happen to some extent, though, and if saw blades are over-hammered (or worse, bent sharply, then bent back by hammering), the increase in hardness can be sufficient to push the steel to it's upper limit of tensile strength, and a crack - a catastrophic failure in engineer-speak - will occur. The same can happen if a set tooth is bent the 'wrong' way.

You're perfectly entitled to your opinion on internal tension in a traditional well-made saw blade - and at least you expressed it politely! My position remains, as it has been throughout the discussion, open-minded but sceptical. The links Mignal found certainly provide some interesting material, though as the first suggests that 'tensioning' was done to reduce the likelihood of a saw developing a wavy toothline if it became hot, and Simon Barley's writings suggest that it was to give a saw the capacity to return quickly to straight if bent, there still seems to be some - well, lack of clarity, shall we say.

No doubt in the fullness of time, the problems will be solved and the questions answered!
 
I really really want to know what saw tensioning is now... as I am pretty much convinced it exists.

Can someone post a photo of a saw that is defiantly tensioned? I am just curious to see one (I do have a few old Dissitons, but I know little about them).
 
Rhyolith":o2cysnln said:
I really really want to know what saw tensioning is now... as I am pretty much convinced it exists.

Can someone post a photo of a saw that is defiantly tensioned? I am just curious to see one (I do have a few old Dissitons, but I know little about them).

Any disston #12 or D8 (at least the older D8s...I have one newer D8 that has a shorter plate and it's less stiff, even accounting for the plate height).

Even the #12 small panel saw that I have for crosscut has excellent stiffness (a 24 inch saw that's 11 point), despite having a smaller plate cross section than some larger crosscut saws that I have. It files like any other saw (I have had saws that are hard enough such that they're really hard on files, but it's not one of them).

Of course, those were tensioned with rollers, so there's no evidence of the tensioning.

Bob Smalser's write-up on saw tensioning describes what's occurring in the metal pretty well. Whether or not it does a 100% job of explaining what's going on, I don't know.

re: earlier comments about a layer of hardened surface - that may be true, but it would be very shallow if it was there.
 

Latest posts

Back
Top