Tension in hand saws.

[Cheshirechappie]

We had a long and rather rambling thread on the subject recently, which some may recall with sinking heart - panel-saw-tensioning-t96004.html

After much diligent searching, Corneel discovered a German text, dealing with machine saw blades, that outlined a method to check bandsaw blades for 'tension' put into them by hammer or roller during manufacture or servicing. The blade is held in a gentle curve, and a short straightedge applied across the width. The 'tensioned' areas will be revealed by slight out-of-flatness - very slight humps - of the blade when so curved. When the blade is flat, no such humps are detectable. A hump height in the order of about 0.4mm was mentioned, detected by looking for light under the straightedge.

Corneel tried this on a couple of handsaws, and found some evidence that the same thing happening (go to the end of the thread for photographs).

I tried the same, and - rather indistinctly - also found 'something' on a couple of saws, though no evidence on others.

On another forum, Rob Streeper has documented his work to investigate 'tension' by hardness-mapping old saw blades. He found two lines of extra hardness on some vintage saws, one about an inch above the toothline, and one following the line of the blade back. (Those lines correspond to the 'humps' Corneel observed on his saws, leading to the possible conclusion that they were the same thing.) He subsequently hammered a new saw blade along those regions, and observed a slight change in behaviour of the blade in service - it seemed less 'whippy' at the toe end on the back stroke, and more inclined to return to straight quickly.

Additional hardness (slight work hardening, in effect) would be caused by extra hammering along the regions identified, so things do begin to look promising in the hunt for 'saw tensioning'.

We know also that surface work to sheet materials introduces residual stresses (long and boring explanations in previous thread), and that hammering or rolling have such an effect.

I've been doing a bit of rummaging. One of my old textbooks is "Principles of Industrial Metalworking Processes" by G.W.Rowe. In his chapter on "Metallurgical Factors in Metalworking" he discusses residual stresses.

He notes the following, "Residual sresses arise wherever there has been inhomogeneous plastic deformation, but it is important to recognise that they are elastic stresses due directly to differences in elastic strain and cannot exceed the yield stress of the material." (page 322). In plainer English, if you work a piece of metal, you'll get some, but unless the metal changes it's physical shape, they'll always be limited to the yield stress of the material.

Perhaps more interestingly, he also notes (page 323), "After a long time at room temperature, or a shorter time if heated, the residual strain and hence the residual stress will be relaxed by processes involving the diffusion of crystallographic defects."

In other words - tension might not last in a saw blade. It will gradually relax itself, in a process rather like the seasoning of a casting. I've absolutely no idea how long that might take, but it could well be that vintage saws that were 'hammer tensioned' a century ago have long ago lost whatever tension they had by a slow, gradual, natural relaxation of internal stresses.

 

[woodbrains]

Hello,

Thanks for this info. I'm not sure it answeres any of the questions about the mysteries of saw tensioning, though, but more info is welcome all the same.

Mike.

 

[Corneel]

He notes the following, "Residual sresses arise wherever there has been inhomogeneous plastic deformation, but it is important to recognise that they are elastic stresses due directly to differences in elastic strain and cannot exceed the yield stress of the material." (page 322). In plainer English, if you work a piece of metal, you'll get some, but unless the metal changes it's physical shape, they'll always be limited to the yield stress of the material.

From this I understand that the deformations themselves did exceed the yield stress (plastic deformations) but the results of these in the other areas is an eleastic stress.

I don't know if it disappears over time, if we can still see the result in saws over a century old. Rob Steeper had another suggestion: Repeated sharpenings bring the toothline up into the "hammer line", and that indeed changes things, resulting in a wavy toothline.

 

[oakmitre]

I would suspect the tension is of benefit if it keeps the blade straight and true in the kerf against side friction.

Beyond that I would doubt if there is any benefit.

I have a Knew Concepts frets saw with a rigid frame which maintains blade tension and hence straight at all times.The difference between this and a traditional fret saw which uses the saw frame as a spring type tensioner is huge.

I also have Lenox HT50 hacksaw which is advertised to achieve 50,000 psi tension on the blade. It is more than twice the weight of a cheap hacksaw I own, but I haven't really found any huge benefit in the Lenox extra tension.

I think that the simple application of some baby oil to the sides of a panel saw would outweigh any tensioning effects because of the reduction of kerf friction, unless we are talking about a very floppy blade.

 

[Trafalgar]

The only part of a saw engaged with the wood should be right at the toothline to the depth of the teeth. If you're banging the middle of the plate against the wood it's like a bad bobsled run.

There's an old Tom Law video that starts out with him sawing and it looks as effortless as Nureyev. The saw is in great tune, for sure, but nothing is touching the wood but the teeth. The saw is perfectly upright in the cut, it is not rattling along and banging side to side in the kerf. That's how it's supposed to be done. Talking about fine fettling points is practically a waste of time until you achieve this level of expertise in use. I certainly have not, but I do recognize it when I see it and understand its importance -- good first steps.