Insights from making infill plane (long)
Posted By: Jim Yehle Date: Monday, 1/6/03, at 5:54 p.m.
I am building an infill plane. My design sources/inspirations are from Jim Kingshott's book, Norris catalogs, photos gleaned from magazines or the Internet of actual antique planes made by Norris, Mathiesen and Spiers, and modern ones by Holtey, Knight, Shepherd, Hoosier, and Sauer & Steiner. I own a ninety-year-old Spiers 7 and have gotten my hands on a friend's Norris A5 and a Sauer & Steiner (Spiers 7 knockoff) briefly.
The basic form and dimensions follow the Norris A13, but smaller, as I believe they only offered 12" and 15" versions, which would typically fall into the size range classified as a panel plane, whereas mine has a sole length of but 9". The iron is bedded at York pitch of 50 degrees and is 2" wide. The reduced length leaves little room for the Norris' A13 front cushion, so I have truncated the back of it like on the Holtey A13. My choice of materials was steel for the sole and sides, for even though brass and bronze alloys look nice, I hate the way they stain lighter woods. I realize that using a steel sole and brass sides may minimize or eliminate this factor. Another reason for the steel/steel choice is that my fabrication method uses a 60 degree dovetail mill, and I think that this angle (30 degrees off vertical) looks too steep on the sides. I prefer closer to 15 degrees, and this is what I measured on my Spiers. I don't think that the strength is compromised in any way, it's just visual, and if both pieces are steel, then the dovetails disappear. I designed a single-thread fine-pitch adjuster and bronze lever cap, which in retrospect I would have been better off buying. I have access to a three-axis NC milling machine with an Anilam 1100 control, and a metal-turning lathe. Because of this, the design of the dovetails takes advantage of machining techniques which aren't practical for a manual (or nonexistent) milling machine.
I began by attempting to reverse-engineer Jim Kingshott's planes from the drawings in his book. As any of you who have used this book as a sole guide to building a plane know, many details are missing. I fleshed out missing dimensions and entered drawings into a CAD drafting program, which I suppose now anybody could use to infer dimensions which I forgot to explicitly state, or to print for a machinist. I also took shapes and dimensions from the sources I mentioned above. I now have a design which I notice is similar to others I have seen posted on Badger Pond in the past few months (Sandor Kovacs and Mick Doherty/Brian Brady), although I don't know dimensional details of those designs. And unlike them, my only 3D rendering of the plane will be in metal and wood. (My CAD abilities aren't those of a full-time professional, nor are my machining talents.) I'm just making this because my current project's curly maple is challenging the abilities of my assortment of handplanes, and as a procrastination aid in my efforts to avoid all that scraping and cussing.
I can post my further details when things are finalized.
To avoid repeating what followers of planemaking forums (fora?) already know, I will mention only things which I don't consider to be common knowledge.
Dovetail design: Perhaps this is obvious to anybody who has already made a set of double-splayed dovetails, but I had trouble visualising it from Kingshott's description. I started my dovetails very similar to traditional wooden ones; the tails are cut into the sides, pins on the sole. At this point, the bottom surface of the sole has notches on either edge whose angles are perpendicular to the sides. Now you make a second cut which opens the roots (i.e. inbound from the edge's outer face) of the pins by another 15 degrees, but only on the bottom of the sole. If you used a small three-square file, you'd remove a small triangle; my CNC technique involved a 60 degree dovetail mill, which I brought sideways, forming a triangular "tooth" into the base of the pin's wall (the sole being held vertically), then raised the mill at 15 degrees off vertical. This leaves the pins looking like more tails from the bottom, and a parabolic void when viewed from the front and rear of the sole. Note that if 75 degree dovetail mills were available, this would be easier and wouldn't require CNC for the simultaneous two-axis move, but I've never seen one save for carbide-tipped router bits, the use of which on steel is an experiment I am not eager to attempt.
Peening: I had available three implements--a 16-oz and an 8-oz ball-peen hammer, and a 5/16" roll-pin (round-nosed) punch. I found the larger hammer useful for riveting, especially where fine control wasn't needed, given my aim with a hammer. (My wife has a green thumb; mine is black.) The 8-oz size was better for peening the dovetails, and I found the punch extremely useful for precisely- aimed whacks, such as coaxing metal into the very inside corner of the dovetails. I jointed and glued up 2x4s to make peening buck, then planed it exactly square and a few thousandths shy of my inside wall-to-wall dimension, notched it for the throat plate, and screwed the sole down at the front and rear using big washers and drywall screws. When peening, I placed C-clamps (or G-cramps if you prefer) fore and aft of where I was working, maybe three or four total, cinched really tight, and had no problem with things moving. Furthermore, I just used my workbench, not an anvil. Shepherd has an elaborate way of affixing the sides to a hardwood buck, but I don't see why all that is needed. Maybe it's because I managed to hold tolerances on the dovetails to a few thousandths, so there wasn't any place for things to go anyway. It was extremely beneficial (this being my first dovetailed plane) to make some practice dovetails and rivets out of scraps. For the (3/16") rivets, I made a guard for my numerous errant hammer blows by drilling a 1/2" hole in thin (maybe .050? I should measure it) steel, and centering it over the rivet. This really reduced the amount of filing and lapping later. Don't countersink too much for your rivets! No more than about .20 outer diameter (for 3/16 rod) is enough. When I had finished peening, I tried to remove the body from the buck, but the sides were so now squeezing very tight. The springback was .020 per side, so it would have been wise to shim by this amount (just like Wayne Anderson advised, a week later), or make my buck's sides angled slightly.
Tooling: My collection of files was pretty lame at the start of this project, and in the process of extending it, I discovered a few which really came in handy. They are:
Filing/lapping: Clean your files often! Unlike wood, chunks of swarf stuck in the file can make serious gouges in your nearly-perfect surface. The Grobet Multi-Kut is very good in this regard--it has deep "gullets" to carry metal shavings out of harm's way, but still cuts fast and smooth. I just heard of packing the file's teeth with chalk--is this the reason for that practice?
- Grobet Multi-Kut [sic] 8" (I got mine from Reid Tool)
(I believe Nicholson "Magic Cut" files are similar.)
- 8" and 6" Warding file (handy because they're so thin)
- Auger-bit (or another safe-side and safe-edge) file
- Swiss-pattern files as fine as #4 cut
- Nicholson #50 rasp, for infill
Order of construction: Rivet the throat plate onto the sole before attaching the sides. Don't make it too wide--it's a pain getting all the way into the corners when perfecting the bed. Remember, its maximum usable width is the width of the iron. I made the slot for the iron after the sides were on, and that wasn't clever. Getting your tools way down into that gully is a challenge. I needed long end mills to reach, but you want small diameters to minimize filing the rounded corners square. These two are mutually exclusive requirements.
Infill: Still working on this. It's a closed handle design. Open-handled rear infills typically use one hunk of wood for the handle, and one for the front, with a big mortise for the base of the handle. But with a closed handle, it seems to make sense to continue the handle piece all the way forward to the iron, and sandwich two sides onto it. I see no disadvantage doing it this way. The pins should hold everything together when the glue joint eventually fails. I would like to use some nice dense exotic wood for this, but tracking down a 10/4 chunk of 5" wide stock to make it all out of one piece is not easy. Besides, I would end up making a lot of very expensive sawdust. As far as the shape, I scaled up photos and drawings of many infill planes, printed them, taped them onto scrap 1" cherry, roughed them out with bandsaw, forstner bits, jigsaw and roundover bits, then smoothed and refined them with rasp, file and sandpaper. This gave me a feel for the shape I wanted. On the final version I'll use 1-1/8" thick stock, because it's more comfortable, and there's more room to house the adjuster. The open handles give a broader opening for larger hands. I like to at least have the option of putting all four fingers (I haven't cut any off yet) inside the handle. With a York pitch angle, it opens up the space at the rear, so you don't need to hang the handle way off the back of the sole, but it's still best to take the bottom of the hole nearly to flush with the sides to buy more room (n.b. Sandor). Now how about grain orientation? As near as I can tell, tradition has it running straight fore-to-aft. This doesn't make sense to me for the front bun/cushion, especially with it sticking way up and having a narrowed "waist" like mine does. It seems better to angle the grain up about 45 degrees. Am I just showing my ignorance? You can not rely on your infill wood not to shrink. In Utah's dry climate, I am always aware of this. Infill rivets ("cross pins") need to have metal sleeves around them to hold the sides square and parallel when the wood shrinks. Kingshott and Holtey know this, and Shepherd should add them. (I believe Sauer & Steiner planes have infill shrinkage problems also.)