Thin Plate Saws- More complicated than one may think

I had hoped not to sound negative or snarky in a recent ww forum post about thin plate dovetail saws. The responses I received were mostly excoriating. When the drama regarding my word choice, and whether I should have posted publicly or privately was finished, the subject changed and no one got back to the point. Since then, several folks have blogged about thin plates saws, but none have touched on what I was trying to say.

By 2003, I had completed a full working set of Seaton Chest style saws. The saws I posted pictures of at WoodCentral were the control group. I made saws of varying shapes, sizes and plate thicknesses. I played with .015, and some thinner stock. The control group was initially secret. I started construction in 2001, using these saws for close to a year before anyone knew about them.

I shared what I learned about the control group (which generated a fair bit of attention) but either because I just didn’t think of it or because no one ever asked, I never talked about the experimental saws I made. See, I thought that the 18th c saw designs were superior to Disston’s. And I wanted to proove it. I still believe the changes 19th c saw makers made to their designs were based on cost cutting and the switch from a pro furniture market to a less skilled carpentry market.

Keep in mind, I was thinking of going into the saw making business. At that time, the only decent new western saws were dovetail saws made by IT/LN and those being made in Williamsburg. IIRC, CW was buying their long blades from England (which I wasn’t wholly impressed with). I knew then that if hand tool woodworking was to survive, we needed good quality rip saws. By the end of 2003, I was convinced no one in the world besides me was interested in 18th c reproduction hand saws so I abandoned my saw making ambitions altogether.

I returned to saw making in 2007. By that time, Mike Wenzloff was making beautiful Kenyon saws and I was curious about earlier designs. So I once again returned to the shop and began experimenting with the interaction between blade shapes and handle position and angle.

In my quest to make faster cutting saws, I experimented with thin saw plates. I tried to apply “thin plate technology” to the full range of Western style saws. Though it may not appear below, I really focused my attention on rip saws, spending a great deal of time trying unsuccessfully to hammer “tension” thin blades to make them stiffer. This is what I learned;

Tenon Saws
Tenon saws are sensitive to the depth of cut. I saw the corners out. So when I encounter architectural sized (read: wide) tenons, I need a fairly deep blade. And this proved to be the limiting factor for the plate thickness.

See, a pushed saw blade is like a column. Make the column too slender and it buckles. Stubborn as I was, I stuck with these saws (and still have some). To make the thin plate saws work, you have to either stabilize the column (with a spine very close to the teeth) or reduce the load on the column. Believe it or not, you can actually do the latter.

What I did was first reduce tooth size (which limits the length of kerf to some degree because the gullets fill with saw dust and then lift the saw out of the cut). Then I reduced rake. I never thought to add fleam. Mike Wenzloff does that with some of his saws. I have no doubt it’s an effective strategy. I just don’t think I could hand file 1 or 5 degrees or fleam.

If you are following me, adding rake, and increasing tooth pitch (tpi), make saws cut more slowly. A thin plate saw should cut faster since it removes less wood. But together, a thin saw with slower cutting teeth result in a saw with not necessarily stellar performance.

Dovetail/Drawer Saws
For DT saws, .015-.020 is fine. Below .015 is where the fireworks start. I made exactly one saw with a sub.015 plate and I hated it. I’m not sure how to get such a saw to function well. I limited the depth of cut, had a glued on machined slotted spine to ensure I had good pressure between the blade and the spine.

My thin blade got distracted by the rings in the QS Eastern White Pine I use for drawer sides. I wouldn’t even consider SYP or Oaks with such a saw. Part of the problem seemed to be the thinness of the stock. The rings would twist the blade sideways, the blade would vibrate, often cutting a wider kerf tan a thicker saw would. As a result, I was convinced not only was this not a faster cutting saw, it was a slower one. Pulling such a saw should solve the problem, but I’ve had some similar problems with my admittedly cheap Zeta dozuki (yes I have one). Lastly, cutting drawer DTs is not a place where I’m looking for speed improvements.

Carcass Saws
For carcass DTs, I do want a fast cutting saw. But for these joints, I find the use of a coping saw to be a significant time saver. Anything that prevents its use, in my opinion, can’t really be considered a time saver. I found making a long shallow bladed carcass saw was a good choice. If you are looking to make a saw faster, making it longer is the first obvious choice. This is the reason why folks made 8′ long pit saws.

I had some trouble filing really thin plates. Not sure how Japanese saw makers do it. Filing thin plates is the same as sawing thin boards. The board bends under the pressure from the saw teeth, then snaps back and you get vibration. The files I was using just didn’t have fine enough teeth I guess.

Rip Saws
I found some success with a thin plate (.032) rip saw. This, in my mind, was where an improvement in speed had the most potential to be helpful. I did a few things to make these saws work:

I made them shorter than normal. They are difficult to control at 26″, and much better behaved at 24″. I filed fairly small teeth (approx 10ppi) and added more rake than I usually prefer. For very hard, thin wood, they are nice to use. But are they really faster? Frankly, I doubt it. Part of the problem is their tooth geometry. Part of the problem is just their delicacy. I’ve since shared these saws with friends and students. Some students have found them unusable.

To be fair, it’s hard to measure the speed of a saw. In real world tests I conducted, the sharpest saw always won, typically regardless of its design. And this really shouldn’t come as any surprise to you. And I can typically change a saws teeth to get it to go faster. So much of the sawyer’s experience is governed by the filing of the saw’s teeth.

My advice (if you’re asking) is to try it before you buy it. And don’t assume that because a saw has a thinner plate and therefore removes less wood, it will be faster cutting. Saws are complicated tools. Thin plates might work best in super hard woods, rosewood, ebony, etc, and especially those that are fairly homogeneous.

Adam

11 thoughts on “Thin Plate Saws- More complicated than one may think

  1. Shannon

    Adam, like everyone has said, this is a great post and a good discussion thread to boot. I have to agree with Bob that I’m not really concerned about speed in my joinery sawing and I would expand that to include cross cutting lumber. The point where I want a fast cutting saw is when ripping, and especially those first cuts from the mill. I have been working with a lot of 8/4 and 12/4 stuff lately and it is a bear to rip. I do have an Atkins 28" saw that I had Mark Harrell at Bad Axe Toolworks sharpen for me and he does employ a tiny bit of fleam along with a progressive rake. This saw cuts like lightening but as you suggest sharpening it is a bear so it is an unhappy trade off.

  2. James Watriss

    I wonder if the race to thinner plates was inspired by the Japanese style saws. I know that for me, for a long time, my favorite DT saw was actually a Ryoba. Even without a spine, it was the first thing I found in my early days that had rip filed teeth and was relatively affordable. And once I got used to the progressive pitch on the rip teeth, I was a very happy guy.

    But pull saws are a whole different animal… Thin plate works in this scenario, specifically because it doesn’t have to bear a load, like your column analogy. So it’s more likely to stay straight.

    If this is, in fact, the case, that the thin Japanese saws have inspired the euro crowd to try to chase thinner plate saws… well, cross-pollination can be a good thing. But not always.

    Keep thinking and writing.

  3. Bob Rozaieski

    Great post Adam. There are so many variables affecting a saw’s performance that focusing on only one of these variables does little service to the customer or end user. I don’t find sawing by hand to be a particularly slow task, with the right saw, except for ripping really thick stock. Certainly, sawing joinery is not where I would want to improve my sawing time. I use a 0.020 plate dovetail saw and can saw from start to baseline in 2-3 strokes of the saw. That’s plenty fast enough for me, and I don’t see how a thinner saw plate would improve much on this, especially if it would have to have more, smaller teeth and more rake than the saw I’m currently using. For a saw designed for long rips in thicker stock, maybe. But again, I’m skeptical, since other variables can also be changed that would likely offer a comparable increase in speed with a thicker plate compared to a thinner plate with smaller teeth. To me, plate thickness is of minimal importance compared to the other variables, which consequently are also much easier to experiment with.

    My experience is similar to what Mike suggests. I’ve only experimented with smaller tenon & sash sized saws, but I feel like I can use a saw with a more aggressive rake when using a lower handle angle (i.e. earlier design). When I use my large rip saw, however, which is a much later Atkins, with a higher hang angle, I have to add some rake to keep the teeth from grabbing and the saw from jamming. I’ve haven’t yet made a large rip saw in an earlier design with an identical pitch to the Atkins to really test this out (not looking forward to filing those large teeth into a 26", 0.042 plate), but I’d be willing to bet that when I eventually do get around to it, it will confirm Mike’s thoughts.

    Of course, as everyone has already suggested, the sharpest saw always wins.

  4. Jonas H. Jensen

    Interesting post.
    In Scandinavia it is my impression that the traditional saw was a bow saw, that could work with a thin blade, but I don’t know if they made western styles as well. I only have one old Bahco cross cut saw, that is the "normal style" (fuchsswanz)or fox tail or whatever you call it.
    But I am impressed with your many talents Adam, making your own saws deserves credit.
    brgds

  5. Luke Townsley

    Along those lines, ease of sharpening may rate high as the key to getting the best saw.

    Anyway, nice post.

    You have started me thinking some more about how this would affect a frame saw built for resawing. I still haven’t made one. Maybe this winter…

  6. Mike Siemsen

    Adam,
    Nice discussion you have started. I believe that the hang of a saw is directly related to the rake. A saw with the handle hung low like an early English saw probably has a 0 degree rake angle, where a later Disston would have the handle hung higher but have between a 4 and 8 degree rake. We can only push so hard, and everyone has a different physical properties. We must find the saw that best suits each of us. I must concur, the sharper saw will always win out.
    Mike

  7. Adam Cherubini

    Matt,

    It was my unscientific preconceived notion that 18th c saws would outperform Disston’s. But i never proved that to myself. As I wrote, the best saw was the sharpest. No doubt Disston had access to superior steel. The question is, are we better off copying Disston’s or Kenyon’s designs? I’ll have to write more about hang angle angle and down force.

    Adam

  8. Ryan McNabb

    Thanks Adam – very helpful. With so many options, and not a little controversy, it would be very interesting to know your favorite saws, in all their details. I sometimes get confused by all the possibilities and just want to know what works best for a given job. Favorite large tenon saw, favorite general rip, favorite general use dovetail saw, etc. Assuming of course a pretty strict 18th century aesthetic.

  9. Rob Porcaro

    Adam,

    Good stuff. I think a fair general inference one can make from your post is that saws must be assessed by their performance more than by theoretical expectations based on saw plate thickness, etc. I agree!

    Whether it’s saws, planes, sharpening, or any woodworking topic, it’s too bad how often the discussion is in the abstract, based on what should happen with this or that bevel angle, etc. Yes, this stuff is surprisingly complicated since so many subtle factors affect what really happens when steel meets wood.

    Regarding saw plate and kerf thickness, I would add that there is often the assumption (such as in tool catalog descriptions) that a thinner kerf is not only faster but also makes it easier to accurately follow the layout line, for example, when using a dovetail saw. I use thin Japanese saws as well as excellent Western saws such as Gramercy’s. There are many factors which affect how well a saw can be made to track the line, especially the cleanness of the kerf and the ability of the saw to cut without flutter. I do not find that a thin kerf, of itself, makes it easier to track a line.

    Thanks for the insightful post, Adam.

    Rob

  10. Jerome Bias

    Perhaps it is an aside, but it seems that in the quest to get the fastest cutting saw we are most often focused on who has the thinnest plate. Which can be a quite racy item.

    Perhaps an equally important issue is the geometry of the blade, the hang angle of the handle, and the overall manner at which the saw forced the user to approach the item being sawn. It just seems that a saw may be great at showing off the speed of a single cut, but when you are building something and the saw makes your wrist hurt, miss cut half of your joints and inspires you to discover that you have the vocabulary of rather salty sailor then it is of no use to you.

    I am interested in knowing when you cut the corners out, when you are cutting tenons, how does the shape of the blade and hang angle effect how you approach this or effect your accuracy?

    I would also be interested in hearing your or other folks experience with long saws and hang angle. It seems that some saws only work well for me when I have been to the gym and have kept the tricpt muscles well toned and others with lower hang angles (often British or 18th cent. saws) are just easier to use. I can get by with fewer teeth per inch, but less muscle usage. ehh.

    Jerome

  11. Matt Cianci

    Adam
    Awesome post….very enlightening…thank you! As a novice saw smith myself I am absolutely obsessed with learning all I can about saws, and you are certainly a great teacher.

    That said, I was most intrigued by your assertion that 18th C. saws are superior to Disston’s…I have thus far been convinced that the gains made by Disston et al. in the consistency of steel manufacturing and quality control of the material far out weighed any loss from the individual saw makers of the 18th and early 19th C. I have been reading up on your various sites and am interested in you belief that the more traditional handle gives a greater control of the saw…I never gave much thought to the hang affecting accuracy in this degree.

    As it were, I am fortunate enough to have an original Kenyon hand saw in my collection, and though it is absolutely the centerpiece for its historical value, I never considered it much of a user…its not nearly as comfortable to hold as my own custom saws or nicest Disston’s.
    But perhaps your statements will make me revisit the style (I would never use my Kenyon saw or return it to usable condition,but I may perhaps copy it.)

    Thanks again and keep up the great work!

COMMENT