Thread: Hardening of sawplates

Originally Posted by D.W.

Just thinking out loud here, but if you take two pieces of steel from the same coil and you cryo treat one and not the other and put them in identical saws or jigs, sharpened identically, I wonder if you could make a device that would cut wood with them. After so many minutes of cutting, you could inspect the wear and then test each in the same piece of wood, seeing if there are any differences (e.g., how many strokes it takes to get through a piece of wood) until such device would no longer cut.

A couple of years ago, I recall someone taking about thirty strokes with a rasp and then contending because they could see more wear on an uncoated rasp than the coated rasp that everyone should buy the coated rasp (which was something sold by someone they were compensated by).

Joel at TFWW here quickly put the advice to bed by suggesting that they had designed a machine to stroke a rasp several million times and that as the rasps dulled, they actually cut faster, and that their uncoated rasps lasted more than most people could ever reasonably expect to stroke a rasp in regular use.

You wouldn't have to have anything advanced, just a machine that reciprocated maybe 4 inches and attached to the saw plate and allowed you to put some amount of weight on the spine (be it a few ounces or whatever). When a saw was dull enough that it was no longer doing anything other than burnishing, you could call it good.

That would, of course, take some time, but you wouldn't have to do the same exercise more than twice, and you could use something like cocobolo to accelerate saw wear.

Hi David,

I've been thinking about how to do a relevant measurement too. One idea is to follow the industry but the devices for doing these types of measurements are quite expensive, even when on the used market.

Surface roughness testers are more affordable but the question of how to roughen a sample of saw steel in a controlled fashion arises. One way is by creating a cutting machine as you mention but I imagine that a lot of work would go into the creation and, since wood is a non-uniform natural product, a lot of wood would need to be sawn.

I'll keep thinking.

Cheers,
Rob

The first cryo-processed blade.



No distortion whatsoever, rings like a bell now.

Continuing my investigation of the effects of manufacturing procedures on 1095 saw plate steel I looked at grinding of the cut edge. The edge of a piece of scrap 1095 0.035" thick was ground using a belt grinder just to the point where some discoloration of the blue was observed. No red-heating was induced.




As before I took five measurements at each of four distances in from the ground edge. Here are the results.




The message here is simple: 1095 steel doesn't like procedures that involve heating - use a file to dress edges that will be toothed.

Rob, interesting info on heating plate. I removed (almost all the way) teeth from a hand saw I cut down (badly kinked, cost nothing, so fine to experiment). I finished with a file. The saw plate did not heat up (I checked regularly), and was never as warm as a saw gets on a vigorous sawing session. So your test results would not apply to this situation... I guess.

I cut the end off with a Makita 100mm x 1mm cut off wheel, no evidence of blueing and did not feel at all hot. I use the same wheel to cut 0.20" 1095 to length, and have not seen any blueing, and again the plate is not hot to touch afterwards (of course the tooth line is unaffected by this action). I use angle iron as a heat sink and a guide - following Ray's lead, except I am using an old cheap Black&Decker angle grinder, that has done a huge amount of work on fibreglass boats and just refuses to die.

Are you able to test a cut with a 1mm cut off wheel, with plate held in a heat sink?

Thanks
Peter

Originally Posted by Heavansabove

Rob, interesting info on heating plate. I removed (almost all the way) teeth from a hand saw I cut down (badly kinked, cost nothing, so fine to experiment). I finished with a file. The saw plate did not heat up (I checked regularly), and was never as warm as a saw gets on a vigorous sawing session. So your test results would not apply to this situation... I guess.

I cut the end off with a Makita 100mm x 1mm cut off wheel, no evidence of blueing and did not feel at all hot. I use the same wheel to cut 0.20" 1095 to length, and have not seen any blueing, and again the plate is not hot to touch afterwards (of course the tooth line is unaffected by this action). I use angle iron as a heat sink and a guide - following Ray's lead, except I am using an old cheap Black&Decker angle grinder, that has done a huge amount of work on fibreglass boats and just refuses to die.

Are you able to test a cut with a 1mm cut off wheel, with plate held in a heat sink?

Thanks
Peter

Hi Peter,

The cut I described earlier was done with a 0.065" cut-off blade, 6" in diameter at ~9k R.P.M.. Steel is a relatively poor heat conductor and so is the wood I used to clamp up the test piece. I was thinking of a rig using some brass to serve as a heat-sink. Envision two wooden jaws with 1/8" thick brass strips about 1" wide rabbeted into the wood supports such that they form a jaw flush with the surface of the supporting wood. Then use these to clamp the 1095 and cut. The brass, because of it's higher thermal conductivity, should help keep the steel cooler.
I'll try to make some time to see if using brass as a heat-sink helps.

Cheers,
Rob

Originally Posted by Heavansabove

Rob, interesting info on heating plate. I removed (almost all the way) teeth from a hand saw I cut down (badly kinked, cost nothing, so fine to experiment). I finished with a file. The saw plate did not heat up (I checked regularly), and was never as warm as a saw gets on a vigorous sawing session. So your test results would not apply to this situation... I guess.

I cut the end off with a Makita 100mm x 1mm cut off wheel, no evidence of blueing and did not feel at all hot. I use the same wheel to cut 0.20" 1095 to length, and have not seen any blueing, and again the plate is not hot to touch afterwards (of course the tooth line is unaffected by this action). I use angle iron as a heat sink and a guide - following Ray's lead, except I am using an old cheap Black&Decker angle grinder, that has done a huge amount of work on fibreglass boats and just refuses to die.

Are you able to test a cut with a 1mm cut off wheel, with plate held in a heat sink?

Thanks
Peter

Hi again Peter,

Here's the tempering curve for 1095 steel.



Any heating beyond about 350 ^oC will draw the hardness of 1095 down to a level that is currently considered unacceptable for modern saw plates.

Originally Posted by rob streeper

Hi again Peter,
Any heating beyond about 350 ^oC will draw the hardness of 1095 down to a level that is currently considered unacceptable for modern saw plates.

Thanks Rob, I cannot imagine ever getting close to that temperature with a 1 mm cut off disk, I can hold the saw plate immediately after the cutting. Do you have any idea of the temp of the plate after you used your larger, thicker cut off disk (I could only find 1mm disks for 100 and 125mm grinders, 150mm may be too large for 1mm).

Having said that I may have access to large guillotines/shears, handy if doing a number of cuts for all reasons.

Originally Posted by Heavansabove

Thanks Rob, I cannot imagine ever getting close to that temperature with a 1 mm cut off disk, I can hold the saw plate immediately after the cutting. Do you have any idea of the temp of the plate after you used your larger, thicker cut off disk (I could only find 1mm disks for 100 and 125mm grinders, 150mm may be too large for 1mm).

Having said that I may have access to large guillotines/shears, handy if doing a number of cuts for all reasons.

I'll try to measure the temperature of the steel as I cut a bit of it. I don't have a contact pyrometer but I may be able to get an idea with my IR instrument.

My aims in this effort are really twofold:

1) Understand what Disston did to the degree I can and

2) Use the insights gained to try to make the best blades I can. I have a process worked out in principle but have yet to test it in practice. I just received material for building my hardening and drawing dies.

Of course I still need to do a multivariate analysis of design factors to optimize cutting but all in good time.

Continuing the train of thought of this thread I decided to do some measurements of my Disston saws. I've now accumulated a collection that covers pretty much every major known phase of saw production at Disston. I won't bore you with all of the data but here's the synopsis. The plot got a little screwed up in processing but bear with me, I'm trying to figure out how to make it better.

Historical hardness survey of Disston saws 091116a.jpg
The plot shows the average hardness of each of the saws in my collection. For the various time periods I chose a year about half way between the beginning and end, for example 1897-1916 is presented as 1906 and 1940-1947 is presented as 1944. The earliest two saws are somewhere in the late 1840's or early 1850's, these are presented as 1850.

The green line is the average of five measurements taken at the cleanest possible place under the handle. I know that the old-wives tale is that this area is harder but my work shows that assumption to be incorrect. The area under the handles is often the least corroded and corroded surfaces return hardness readings that are anomalously low.

The two red lines are the upper and lower bounds of the 95% confidence interval as calculated by multiplying the standard deviation by two. This approximation isn't exact of course but, let's face it, saw hardness measurements aren't a life or death issue.

My interpretation is: The data show a trend of decreasing variability from the inception of the company through WW1. Then variability increases through to the post WW2 period where the variability decreases again. Maybe this is why the 1896-1917 period is considered the "Golden Era" of Disston saws, better QC by skilled first generation hands who had been at the works for the majority of their working lives. By the time that things got better again after WW2 the consumer had taken a bad impression due to the sloppier post WW1 products, and or they wanted those shiny new Porter Cable power saws, not sweaty old handsaws.

The WW2 era saws are generally harder and the 19th century saws are a bit softer than are the later saws.

Some of these saws are the same saws that were showing unusually high values in the previous measurements. I've discovered that some of those measurements were high because the plates are very slightly bowed. This applies to the saws in my collection made from 1928-1947. The screw holes in these saws appear to be punched and the punching operation puckered the steel around each hole leading to the anomalous readings. A measurement taken on a surface that springs back under the indenter reads anomalously high. I took pains to avoid that problem with these data by sampling the other side of the bowed blades so that there was no or minimal springing of the test articles.

Here's some alternative plots. Found and fixed an error in the order of the latest three saws.

Historical hardness survey of Disston saws 091116c_Page_1.jpgHistorical hardness survey of Disston saws 091116c_Page_2.jpg

The plot directly above shows the variability of each saw relative to the overall average hardness, 50.7. The black lines are 2 point moving averages that may help highlight the trends in the data.
The data suggests that the period after WW1 and before the end of WW2 was characterized by increased variability in the hardness. QC problems? Ennui or laziness? Poor management?

I like that range, it really ties in with what the files say, too (not some secret files in a drawer, but what the saw files communicate to you when sharpening).

The 19th century saws feel different to me when filed. Hard to express in words but they feel like they're softer and or more friable, almost as if they're less substantial material despite the fact that the cutting of the steel isn't obviously different. The later saws feel like new 1095 steel does.

All of the saws but one tested above were smooth or bright in finish under the handle. The corroded saw was one of the 1850 era but after cleaning with 280 gr. W&D it showed a similar variability to the other 1850 saw with a S.D. of 1.21 vs. 0.96.

I would say the same thing - the older saws have a very dry feeling. Friable is a good term. Not sure if the steel was larger grained, but it feels more grainy and plain.

Rob, how many saws per one green dot?
Thanks,
Paul