Hammering Hand Saw Blades

[rob streeper]

I've decided to make up a couple of Harvey Peace style hand saws. These are going to be 0.042" plates, hammer tensioned and cryogenically treated. I spent the day cutting both 0.035" and 0.042" steel. I also started hammering the cut plates and got a couple of them ground and shaped.

First the cut plate ~26" at the toothline.

352322


Given that I'm using rolled steel I wanted to keep track of the curvature of the steel as I work it. To do this I conceived an ad-hoc droop meter. The white line on the plate is a reference mark for positioning the plate at the edge of my work table. I also marked the positions of my clamps.

352323


I used my laser level to create a reference plane.

352324


Hold a rule out at the end of the blade and take a reading.

352326


Flip the blade over and read the droop on the other side.

352327


Big difference. Now for some corrections. First the anvil. I have two smaller than this but the large one pictured here is much easier to use for this type of work.

352330


5 lb diagonal peen saw hammer.

352329


And strike some guidelines on the plate. I intend to strike above the two white lines away from the area to be toothed.

352328


The blade blank is curved due to the rolling of the steel. The effect is quite pronounced.

352335


After hammering, back to the droop test. Side 1.

352333


And side 2.

352334


After grinding the surface of the blade is clean of hammer indentations.

352332


After a little more hammering, some more grinding, some more hammering and file work we have two blades, a 26" and a 28".

352331

As to the amount of force applied in hammering the steel think dribbling a basketball, allow the hammer to rebound and then gently drive it back to the steel. Don't take the attitude of driving a stake through the heart of Dracula or you'll screw up the plate.

[Corneel]

Thanks for the post Rob. Did you hammer both sides equally to get rid of the curvature in the 26" blade?

[rob streeper]

I frequently find that both sides need hammering. My preferred method is to hammer first the convex side. If the plate straightens up I check the droop. If the droop is uniform on both side I quit. Sometimes I find that I need to hammer the convex side until it becomes slightly concave then I flip the blade over and hammer the formerly concave side to bring the plate back straight.

[DaveTTC]

Wow. This is an intersting thread. A lot of work too

Dave TTC
Turning Wood Into Art

[rob streeper]

Wow. This is an intersting thread. A lot of work too

Dave TTC
Turning Wood Into Art

Thanks Dave. This is fairly easy to do though the 0.042" steel is a bit of a work-out. I didn't document all of the blades I did yesterday, only two of the three Harvey Peace type. I also cut blanks for and hammered 6 blades in 0.035" steel. The other blades were not profiled or ground however.

[DaveTTC]

I see there is something for me to learn here

Dave TTC
Turning Wood Into Art

[Isaac S]

Is the hammering that you show being done to "tension" the blade or to flatten it?

[rob streeper]

Is the hammering that you show being done to "tension" the blade or to flatten it?

Hi Isaac,

The hammering is doing both as best I can tell. Flattening can be accomplished by simply hammering out the curve from one side only. As I mentioned above, I often find myself needing to hammer both sides. After doing this I can see changes in the droop of the tip of the blade. In addition the blade feels 'snappier' when flexed. The tensioned blades also exhibit better resilience when flexed to extremity.
In my thread on blade hardening I also measured a stress induced increase in apparent hardness in areas adjacent to the areas that are hammered. This stress hardening is better known today in the guise of shot-peening, sometimes used to work harden the connecting rods of engines for example.

Cheers,
Rob

[Isaac S]

Assuming for the moment that the hardness is affected by the hammering, what does that translate to in use? I guess I am having a hard time seeing what this does beyond flattening the blade.

Shot or hammer peening, as I understand it, results in a compressed layer of material at the peened surface. This layer of compressed material counteracts applied tensile forces, preventing tensile cracks from forming on the surface. Every reference I have found says that peening is used for stress relief and fatigue resistance, and does a very good job of that.

So far, I understand all of this, at least at a beginner's level. Here's what I've always had a hard time understanding. There are two material properties that are of particular interest to me as a potential handsaw maker: yield strength and Young's modulus (stiffness). Yield strength dictates the degree to which a saw can be bent before being permanently deformed, while Young's modulus tells me how hard it is to bend it that far.

I can sort of see that peening will increase the yield strength through work hardening. My only reservation is that the hammering is done on such a small portion of the blade that its overall effect can't be very large.

Where I have real problem understanding is how hammering increases the stiffness. Young's modulus (for steel) is practically independent of alloy, hardness, and yield strength. I have never seen a claim that peening has any effect (in a quick online search, the one reference I found showed no effect). I'm not sure how "snappiness" could be measured, but my best guess is that it is related to stiffness.

In your pictures measuring the droop, one side measures about 1 3/4", and the other about 4 1/2", although that measurement looks like it may actually be a bit less. That gives a total droop of approximately 6 1/4", or an average of 3 1/8".I'm not sure how careful you were taking those measurements, since they were just for illustrative purposes. But if those are indeed fairly accurate, I would say that you haven't changed the stiffness to any meaningful degree.

To me, a really interesting experiment would be to take a piece of spring steel (hopefully flat, but I'm not sure if that is necessary) and subject it to a three point bending test. Then take that same piece of steel and hammer it just as you have done, then repeat the test to see if stiffness has increased or not. I have no access to the testing equipment needed to do this, but might be able to rig something. I'm thinking of two cylinders at a fixed distance apart that the test strip can rest on, with a dial indicator beneath the strip at the center. The only way I can think of to load the test strip is with fixed weights. If you're up to doing some hammering, I'd be willing to try setting that up. I can supply a few test strips.

[planemaker]

http://www.crosscutsawyer.com/viewtopic.php?f=4&t=750

[rob streeper]

Assuming for the moment that the hardness is affected by the hammering, what does that translate to in use? I guess I am having a hard time seeing what this does beyond flattening the blade.

Shot or hammer peening, as I understand it, results in a compressed layer of material at the peened surface. This layer of compressed material counteracts applied tensile forces, preventing tensile cracks from forming on the surface. Every reference I have found says that peening is used for stress relief and fatigue resistance, and does a very good job of that.

So far, I understand all of this, at least at a beginner's level. Here's what I've always had a hard time understanding. There are two material properties that are of particular interest to me as a potential handsaw maker: yield strength and Young's modulus (stiffness). Yield strength dictates the degree to which a saw can be bent before being permanently deformed, while Young's modulus tells me how hard it is to bend it that far.

I can sort of see that peening will increase the yield strength through work hardening. My only reservation is that the hammering is done on such a small portion of the blade that its overall effect can't be very large.

Where I have real problem understanding is how hammering increases the stiffness. Young's modulus (for steel) is practically independent of alloy, hardness, and yield strength. I have never seen a claim that peening has any effect (in a quick online search, the one reference I found showed no effect). I'm not sure how "snappiness" could be measured, but my best guess is that it is related to stiffness.

In your pictures measuring the droop, one side measures about 1 3/4", and the other about 4 1/2", although that measurement looks like it may actually be a bit less. That gives a total droop of approximately 6 1/4", or an average of 3 1/8".I'm not sure how careful you were taking those measurements, since they were just for illustrative purposes. But if those are indeed fairly accurate, I would say that you haven't changed the stiffness to any meaningful degree.

To me, a really interesting experiment would be to take a piece of spring steel (hopefully flat, but I'm not sure if that is necessary) and subject it to a three point bending test. Then take that same piece of steel and hammer it just as you have done, then repeat the test to see if stiffness has increased or not. I have no access to the testing equipment needed to do this, but might be able to rig something. I'm thinking of two cylinders at a fixed distance apart that the test strip can rest on, with a dial indicator beneath the strip at the center. The only way I can think of to load the test strip is with fixed weights. If you're up to doing some hammering, I'd be willing to try setting that up. I can supply a few test strips.

During my hardening experiments I happened across this reference: https://en.wikipedia.org/wiki/Bauschinger_effect

My understanding is that the irreversible compressive deformation of the steel that is hammered places adjacent regions under stress.

[rob streeper]

http://www.crosscutsawyer.com/viewtopic.php?f=4&t=750

I've seen that too. The crosscut saw enthusiasts also have discussed the use of English wheels for blade tensioning.

[rob streeper]

Making the most of long summer days. Above is Mr. Wenzloff's version, below is mine. 5 1/2 ppi, 4o rake, mahogany handle, 3X large screws, medallion and stainless bushings. Both saws have breasted toothlines, the Wenzloff is filed crosscut and is, if I remember correctly, 6 1/2 ppi.

352400

[Isaac S]

http://www.crosscutsawyer.com/viewtopic.php?f=4&t=750

Thank you very much for that, Stewie. I stumbled upon that forum last year when I was thinking about this very topic, but do not remember finding that thread. That is the first explanation I have seen or heard that makes any sense to me. Specifically, this quote from Diabolo was very enlightening:


Tensioning is expanding the central portion of a blade. Thus the belly of the saw blade is under compression, whereas the toothline and the back are under traction. On a circular saw blade, tensioning compensates the mechanical and thermal dilatation of the toothline under operation. If not, it becomes "too long", and the saw blades is unstable and wobbles in the kerf. On a handsaw, this is for compensating the thermal dilatation of the toothline (my emphasis). It doesn't warm a lot when cutting, but it is enough to make the saw cut less straight because the blade is thin.

I am aware of this phenomenon on backsaws, but never made the leap to this happening in a handsaw.


Nice looking saw, Rob.

[planemaker]

Your welcome Isaac.

Stewie;

[IanW]

The way I've always understood it is that the reason for hammering a saw is to put the plate into tension on each side, which has the effect of stiffening the blade. This is analogous to sapwood in trees being in tension to resist bending in wind. The old mast makers for the tall sailing ships apparently understood the principle and liked to find trees that only needed the bark removed - any trimming or straightening removed the sapwood & reduced the stiffness & strength of the mast...

Cheers

[RayG]

For large circular saws tensioning the blade is a common requirement, I worked on sawmills for a bit, ( not on doctoring saw blades however :) ) they had a guy who would periodically come along and "tension" the blades that needed it. The saw doctor, would hammer the center in a pattern to provide some tension so that when the blade heated up while in a cut the blade would still track rather than become "floppy" in the center and run off track. The saw mill operators described it as a properly tensioned blade would "stand up" through the cut, whereas a blade not tensioned would bend off to one side and not track.

As to how this translates to hand saws, I guess it's the same principle, but more to do with stretching the steel to provide tension along the tooth line.

Full marks to Rob for going to the extra effort. :2tsup:

[rob streeper]

I am aware of this phenomenon on backsaws, but never made the leap to this happening in a handsaw.




Isaac,

I think that I've learned that the word 'tensioning' has a different meaning when applied to a backsaw in that the tension is applied to a backsaw blade by the back and in a non-backed blade by the lines of hammering.

Another factor that likely contributes to the snappy feel of a tensioned handsaw blade is that the blade is no longer flat top to bottom. The distortion is not easily or casually visible but if you look at an older saw blade in the right light you'll see what I mean. Analogous to the situation in corrugated tin sheets. Without the corrugations the flat metal sheet is very flexible and it gains a lot of longitudinal stiffness with them.

In my hardness mapping of the older blades that DW sent I was able to measure the pattern of hammer blows on the No.7 plate. The pattern is reflected in my paint strikes made on these plates to guide my hammering.

Cheers,
Rob

[Isaac S]

For large circular saws tensioning the blade is a common requirement, I worked on sawmills for a bit, ( not on doctoring saw blades however :) ) they had a guy who would periodically come along and "tension" the blades that needed it. The saw doctor, would hammer the center in a pattern to provide some tension so that when the blade heated up while in a cut the blade would still track rather than become "floppy" in the center and run off track. The saw mill operators described it as a properly tensioned blade would "stand up" through the cut, whereas a blade not tensioned would bend off to one side and not track.

And not just heat, but centrifugal forces expands the steel at the rim more than at the center.




I think that I've learned that the word 'tensioning' has a different meaning when applied to a backsaw in that the tension is applied to a backsaw blade by the back and in a non-backed blade by the lines of hammering.

For all of the words in the English language, it can really be used imprecisely.




Another factor that likely contributes to the snappy feel of a tensioned handsaw blade is that the blade is no longer flat top to bottom. The distortion is not easily or casually visible but if you look at an older saw blade in the right light you'll see what I mean. Analogous to the situation in corrugated tin sheets. Without the corrugations the flat metal sheet is very flexible and it gains a lot of longitudinal stiffness with them.

I don't think this is a good analogy. The bulk of the stiffness of corrugated sheet metal is due to its cross-sectional shape. The corrugations are orders of magnitude larger than the thickness of the metal sheet. In a saw, any variations are be on the order of 1%, give or take a bit. If there is any appreciable effect on stiffness, is has to come from either internal stresses in the saw plate or changes to the material properties itself.

[rob streeper]

I managed to get Harvey Peace #1 sharpened and put on the scratch coat today.

352489

352488

Some clean-up of the lines, finishing and off to the laser guy.

[acowboy]

What a great job...
And thank you for the info on how to tension the blade.
Have an old Disston TH saw, has a slight bend near the tip, same concept should work to straighten it to some degree.

That saw is just beautiful.

[rob streeper]

I think that this is what happens when a saw plate is hammered and then ground. I've been able to produce this effect repeatedly.360682

This also provides a degree of understanding about the sequence of steps used by the Disston works described here: http://www.disstonianinstitute.com/factorytour.html referring specifically to the smithing ->grinding -> tensioning (as described above) -> drawing -> blocking -> polishing -> stiffening steps.

The stiffening step, apparently a liquid phase process, is still a mystery. Perhaps a precipitation hardening step?

[hiroller]

For all that we know about saws, there is still so much we don't know about how they made saws 100 years ago. We will probably never know.

[rob streeper]

Some arts truly do get lost, Wootz steel for example (maybe rediscovered?).

Saw making in the 19th century was very much a dark art and most of that knowledge was never written down. Disston began operations before there were standard hardness measuring techniques for example so there must have been a group of experienced workers who knew how hard a saw blade should be and could tell without instruments.
Handsaw manufacture has largely been left to the hobbyists as there's insufficient profit incentive for big industry. Consider that in the years since the effective dissolution of the traditional Disston handsaw production line there have been only two notable technological advances in handsaw manufacture. The first of these is hardpoint saws and the second is slotted backs.

I've learned that the lost techniques can often be re-discovered by applying effort, thought, knowledge and modern tools. Materials science has made enormous advances in the past century and has revolutionized many products in my lifetime. We sawmakers can apply some of these advances to the benefit of our craft.
I think that we can find more ways to improve handsaws, even if it's only really a hobby.

[IanW]

....The stiffening step, apparently a liquid phase process, is still a mystery. Perhaps a precipitation hardening step?

Rob, doesn't tensioning stiffen a blade? I always thought that was the principal reason for it. This little bit on the page you pointed to intrigues me:

"The blades then undergo the "Polishing" process, then through the important operation of "Stiffening." As to this latter, the different processes and hammering under which the blades have passed, has altered the arrangement of the molecules in the metal and in order to restore the desired qualities and spring they are stiffened in a special bath, which was originated and is known only to Disston."

I guess I'm an old sceptic, but this sounds like a bit of advertising hype to me. Remember the people writing advertising copy are not always technologically-minded, & it's also said that old Henry was at least as skilled a salesman as he was a sawmaker....

Cheers,

[rob streeper]

Ian,

I think that there are likely red-herrings in every technical description proffered by a company such as Disston that has a stock of trade secrets to protect. I've seen examples of such intentionally misleading information offered in discussions relating to new age saws as well.

I'm struggling to find a way to quantify the differences I can feel in the blades. First, hammered blades feel a bit snappier, despite the fact that they seem to sag under their own weight when held on side about as much as does an un-hammered blade. They also don't oscillate as much as an un-hammerd blade on the return stroke of cutting. Finally they sound different when tapped having a broader sound than does un-hammered steel. The vibrational characteristics are likely due to the introduction of the random array of nodes over the surface of the blade by the hammer blows.

Cheers,
Rob

[Bushmiller]

I agree with the sales hype nonsense. I find it hard to imagine that anything was done to the saw plate after polishing. My understanding is that this is the absolute last step before the handle is attached.

Regards
Paul

[IanW]

..... hammered blades feel a bit snappier, despite the fact that they seem to sag under their own weight when held on side about as much as does an un-hammered blade. They also don't oscillate as much as an un-hammerd blade on the return stroke of cutting. Finally they sound different when tapped having a broader sound than does un-hammered steel. The vibrational characteristics are likely due to the introduction of the random array of nodes over the surface of the blade by the hammer blows.....

Rob, I would've expected the blades to sag less after the hammering introduced some tension, so there goes my hypothesis. I'm thinking of a growing tree, in which the sapwood is under tension and stiffens the trunk against bending in wind. P'raps you aren't managing quite enough tension yet? Although I've been sporadically working at it, I certainly haven't acquired this dark art myself, so I watch your efforts with great interest. I have managed to get a little tension & changed note into a couple of experimental blades, after a LOT of gentle hammering. I don't have any way of re-grinding, unfortunately, so the last thing I want is to introduce any visible or palpable indentation. P'raps it's not feasible to get a truly usable degree of tension without causing marks, which is why Disston (& you) do the post-hammmering grind? Anyways, you are clearly on the right track, but it may take a while to re-discover dark arts that took a few hundred years to evolve. Fortunately, you have the advantage of being able to make some accurate (& repeatable!) measurements along the way....

Cheers,

[hiroller]

The availability of quality 1095 shim stock has simplified backsaw production.
We just need to convince to manufacturers to make some tapered stock for us for hand saws.
Maybe somewhere there is a great pile of discarded tapered shim stock where the rollers slipped ...

I would have expected that the final step of saw making would have been a cleaning process prior to packing and shipping. Did they ship with any protective coating?

[rob streeper]

I think the snappiness that I'm sensing is what I'm inducing. Remember, after grinding there is less metal to support the blade, and less mass to respond to gravity so the sag test probably isn't the best, or maybe it isn't sensitive enough. The blades do get measurably harder in the areas struck by the hammer and in the adjacent zones of the blade.

As to hammering, I've taken it all of the way and purposely beat the tar out of a blade, the result is amusing. This is a 22" D-8 blade with heavy over-hammering. It adopts a number of interesting equilibrium states.

360702360703360704360705360706


Maybe it could be used for cutting corners?

[rob streeper]

The availability of quality 1095 shim stock has simplified backsaw production.
We just need to convince to manufacturers to make some tapered stock for us for hand saws.
Maybe somewhere there is a great pile of discarded tapered shim stock where the rollers slipped ...

I would have expected that the final step of saw making would have been a cleaning process prior to packing and shipping. Did they ship with any protective coating?

No idea on the shipped condition. Could have been a hot oil or salt bath of some kind or maybe it's just hokum.

I can only imagine the premium that the 1095 steel makers would charge for tapered, flattened shim stock. :o