Panel Saw Tensioning

Help Support

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
If the behaviour of a saw blade in regards to vibration control can be changed by hammering a bandsaw blade (that one is proven beyond doubt), then I don't see a reason why you can't change the behaviour in a handsaw. How benificial it really is in a handsaw is not proven in a scientific way. Probably never will.
CStanford":26o8rvlq said:
CC, they did not hammer the saws exclusively for the purpose of making them straight.

"Tensioning" was reserved for higher end saws, at least in Spear and Jackson's case. Unless they sold all their lower-end saws with a slow bend in them, you are completely wrong or would need to explain why they would make their lower end saws straight by some other method and reserve the expensive, laborious method for their best-quality saws.

I'm reasonably sure there are still people out there tensioning saws, Bob Smalser appears to be one, your insistence on finding one on this particular forum (a not particularly heavily traveled one at that) leaves one baffled by your contorted logic and what this means in the real world.

Perhaps all that is needed is a broadening of your horizons?

Perhaps all that is needed is for those claiming that 'tensioning' does something noticeable (such as stiffening it) to a hand saw blade to demonstrate the fact. If there are people out there tensioning saws, that shouldn't be too difficult to do.
There was/is a demonstrable difference or sane, honest firms like Spear & Jackson would not have engaged in the practice. They were doing it as late as the mid-1960s and perhaps later. I would think the most likely reservoir of old craftsmen is in England, though my guess is that the ones still living are not internet savvy.

I'm not sure how one would demonstrate 'stiffness' via the internet without some elaborate laboratory procedure. So, you see, you've set yourself up to have 'won' the argument more or less by default. Your only problem is that the history of the practice speaks to its legitimacy more than your strident arguments otherwise. It's easy to walk into a room and make the announcement that you'll only believe in black holes when somebody can provide, and explain, a mathematical model for their existence. And you proceed to do this in a room full of people who never took a math course after Algebra II. Therefore, black holes do not exist. Your limited audience could not provide you with the necessary rigorous proof.

I do not currently own any top-of-the line antique saws. I have in the past. They did seem stiffer than a cheaper 'warranted superior' saw but they were not alike enough in other ways to draw a conclusion from my obviously limited experience.

Again, they had other means of straightening the saws. They were straight by the time they arrived at the bench of the craftsman who would tension the, again, top of the line saw. Something was occurring. It beggars belief that it was all a ruse or silly superstition. They paid people to do this work.
Charles - all sorts of people SAY all sorts of things about 'tension' in hand saw blades. All I'm suggesting is that the questions and doubts could be laid to rest if someone, somewhere, demonstrates beyond doubt that there's a way to make hand saw blades stiffer (or whatever) by hammering them (or whatever), and demonstrates it by doing it.

So far, lots of words, lots of ideas, lots of assertions. But no proof or demonstration (that I know of) that 'tensioning' a hand saw blade does anything noticeable to it.
Not difficult to prove. Difficult to organise and you would need people sufficiently interested in taking part. You just need someone who can tension a saw and a handful or two of experienced woodworkers. Same saw type of course, one 'tensioned' the other not. The woodworkers aren't told which is which but they have to pick out the 'better' saw.
That's pretty much how I'd do it but I sense that CC would not be satisfied with the methodology.

I'd love to hear anybody's theory about why any of the major manufacturers would pay guys to beat on already straight and otherwise presumably salable saws if there were no effect from the exercise. I admit to be a bit stuck on this point.
Oh I dunno, sounds as good as way as any. Two identical saws, one tensioned and one not. Blind testing by independent woodworkers.

Another way might be to obtain a fresh saw blade, measure its stiffness, tension it, and then measure its stiffness again by the same method. See if there's a difference. I'm sure other people could come up with other methods, too.

As to why major manufacturers etc etc. Erm, did they? We know they had to smith saws straight as many as three times - once after heat treatment, once after grinding and once after polishing (or 'rubbing'). So if they did hammer-tension, they had three opportunities to do it. Simon Barley's account of saw making suggests that it came under 'smithing', the first of three hammering stages. The 1844 list of prices he shows on page 6 of 'British Saws and Saw Makers' gives prices for 'smithing', 'hammering' and 'blocking', but does not mention 'tensioning'.
I suggest we find someone who can do this, if we can, we use some form of crowdsourcing to fund the commissioning of two saws, possibly with video/text explanation, the saws open to be tested by whoever, and would end up as the property of an appropriate museum, as would rights to any video/text, I've sent a message to the Ken Hawley trust, perhaps a member of TATHS could contact them, and anywhere else people can think of,


As mentioned in Spear and Jackson's Story of the Saw, tensioning by hammering was reserved for only top shelf models. The others had to have been made straight by another method (why hammer to make straight and not 'tension' the saw too?) which logically implies that they arrived at the bench to be tensioned already straight and in essentially salable condition except for the application of a little whipped cream on top (tensioning).

Phew, long sentence. Sorry.
Two links to other forums.

Rob Steeper has done a lot of research into old Disston plates, how they were hammered and trying to replicate it. You should read this thread on woodworkforums to see some of his results. After hammering two lines, one along the back, one through the middle of the saw plate, the plate still has the same eleasticity, but now more even from one side compatred to the other. But the plate feels "snappier" and oscillates less.

Another forum, where a guy named Diabolo explains the tensioning process and the why of it all in regards to large crosscut saws.

I read a German article about pretensioning on large frame saws like they use in a mill. They run three lines along the length of the sawblade, one in the middle and two on each side of this, with a narrow roller. This tries to expand the steel in the length. And this causes the toothline and the back to be in tension. In this diagram you see various tensions in the blade. In the top you see the tensions seperately, at the bottom you see the resultant of all of them combined. As you can see the tension from the rolling process exceeds the tension from the frame. This to give you an idea of the forces involved and what actually happens. The steel itself isn't changed, it's stiffness remains the same, but the tension of the steel changes. Next thing I am trying to grasp is what that actually means.

That's a bit more like it - reference to somebody actually trying instead of just making unsupported assertions. As far as we can tell so far, Rob's experiments are in an early stage, but whilst most of the hammer work to his saw blades is about getting them flat and straight - getting the sheetstock coil curve out of them - there does seem to be a hint of something going on amongst it all. Only a hint, though. Interesting - one to watch. It's also interesting that others have gone through the same sort of discussion that we've had on this thread (albeit rather more concisely) with about the same general results.

The Crosscutsawyer forum link is the same one that Mignal posted on page 5 of the thread.

I'm not sure about the frame saw reference yet. I've got places to go now, but I'll have a think about it. Would help if I could translate the German on the graph - a short session with Google Translate later, I think!
Excerpt from post and links on page 10 of the thread:

"The UltraSTEELTM process, developed by Hadley Industries Plc (Hadleys), is a novel surface dimpling process used on steel strip prior to cold roll forming. This dimpling process increases the strength of the final rolled products and enhances other product properties such as fire test performance and screw retention."

C. J. Wang, D. J. Mynors, M. English, "Simulating the UltraSTEELTM Surface Dimpling Process", Key Engineering Materials, Vols. 410-411, pp. 449-456, 2009

This is a UK company and the professional journal being referenced a UK one as well if I'm not mistaken.

Something happens when you dimple the surface of sheet steel besides making a cute little mark.
I think I have an explanation for the words fast and loose. Loose are the areas in the plate under compression, fast are the areas under tension. During the flattening and straightening procedure, the sawdoctor will use hammers on an anvil to create or remove fast and loose points, just as neccessary to create a flat plate. During the tensioning process this is done in a more structural way. A narrow path along the length of the plate, usually in the middle is hammered and thus made loose. The toothside and the back then become fast.

Straightening, flattening and tensioning all use the same tools and work methods, so it is no wonder they get mingled. Especially when public relations write about it. It is very well possible that one man in one run flattens the plate and tensions it at the same time, without a layman seeing the difference. When they use a roller tension system, then it is a separate activity using a special machine. The effect of both procedures can be the same but the former asks for a lot more skill.
In those German texts I found descriptions how the pretension in a sawplate is checked. They bend the saw and look for a lightgap between the plate and a straight edge. When the plate is properly tensioned it will be flat when you keep the saw straight like you would in normal use. It is flat lengthwise and crosswise. But when you set the plate in a slow bend along the length, the pre tensioned parts get the chance to release part of their tension by bending differently. The sawplate gets a curved cross section. The sawdoctors have measured this with the light gap, an old technique. One of those documents mentioned a lightgap of 0.4 mm for a properly tensioned framesaw blade as used in a sawmill.

So I decided to have a look at the several old saws I have. I don't claim to be a sawdoctor, so take this with the usual grain of salt.

I have two Disston D8's from the 1930's. I have an older Spear and Jackson (no idea how old). And I have two Tyzacs, which also look rather old to me with the name punched into the plate, instead of etched.

First I looked at the straightness crosswise with the plate held straight. They were all flat or almost flat.

Then I bend the plate, pushing the toe against a stop on my bench and bending at the handle end. So I had a free hand to hold a straight edge against the plate, crosswise again, in the middle.

On both the D8's I couldn't detect much change, the plates still looked pretty flat crosswise.
The Tyzacs both showed a more convex surface on the outside of the bend in the plate. This corresponds with the information in the German texts for a tensioned blade. The convexity is not much, but I really have no idea how much it should be for a handsaw.
The Spear and Jackson has the most obvious convexity. I think I can even see how the plate is tensioned through the middle and along the back edge. I tried to shoot some pictures from the S&J. Hopefully you can see what I mean.

A picture with the sawplate straight, you can see that it is flat crosswise.

And a picture with the bend saw plate. Here you can see the effect of tensioning.

In use the two D8's both have the annoying flapping habbit on the return stroke. The S&J and one of the Tyzacs are nice saws. The last Tyzac isn't in working order at the moment.

So, overall, I would carefully conclude that this handsaw tensioning business is more then just a myth.
Not on both sides. When you bend the plate lengthwise it takes on an arched shape crosswise. So it is convex on the outside and concave on the inside of the bend. The pretension of the hammering process seeks to relieve itself a bit un this way.
Corneel - I think you're on to something. What makes it interesting is that there is some accord with Rob Streeper's experiments. His hardness measurements of old saw blades suggested lines of (something) along the back and middle of the saw, and your bend-and-straightedge checks on your saws suggest something very similar. Also interesting that the saws that exhibit this tend to have less toe-end waggle in use.

It seems (from what we know so far) that the internal stresses of the blades are being manipulated. Earlier in the thread, we discussed this at some length, but always assumed that the whole blade was being treated. It seems that instead, internal stresses are modified only in bands along the saw.

All VERY interesting.
Before we get too excited, I forgot to meassure the thickness of the blades. Behaviour of the saws depends a lot on the plate thickness too, and Disston did make them pretty thin. I'll check tonight, shouldn't take too much time.

Those bands of tensioning are similar to what I found for bandsaws and those saw mill frame saws, in the literature. You do need to make contrasting areas of compressed and tensioned steel, otherwise nothing happens.

Latest posts