After assembly, I used a rounderover bit to take an arc out of the outside corners.

Paper mold with mesh sewn on partway:

White oak mold and deckle, from corner

I’m really happy about how the deckle came out. The joint was tricky, but quite strong despite the short grain on the nub that sticks out on one end:

!
https://s3.amazonaws.com/ljimg/mn22vqy.jpg!

I realized an interesting difference in this construction, compared to most frame joinery. Most frames are joined so the two sides press on in parallel from opposite directions. For example, in a dovetailed drawer, the pins are on the front, and the two sides with tails press on the pins from left and right in opposing directions. In the joinery for the deckle, the sliding dovetail joints chase around the four corners of the frame. So the bottom right corner has the dovetail pointing right, the top right corner points up, the top left points left, and the bottom left points down.

This makes the frame harder to put together, but also less likely to fall apart.

But back to the weight: if I had to lift this out of as vat of paper pulp 500 times a day, I’d be pretty sore. So I’m thinking of redoing it all in mahogany.

My biggest problem was getting the dowel-end to be well formed. I got some great advice from the joinery forum about machine approaches, but finally ended up whittling them rough, then bringing them to a final shape with a homemade dowel-cutter formed of a steel plate with three overlapping holes (the overlaps create cutting teeth—sorry, didn’t take a picture). They ended up good enough, I think.

I’m building a more-or-less historically accurate paper mold that will be used to make paper out of West Texas cotton that we’ll print on in the Texas Tech Letterpress Lab (http://letterpress.writingstore.com/). The idea is to sell prints to alumni to help support the lab, which teaches students about historical methods of printing and helps them understand how most of the literature in the past several hundred years was created.

Setting out on this project I had the advantage of two beautiful replica molds made by Timothy Moore (http://www.timothymooretools.com/molds.html), which we bought on a grant several years ago. But these are strictly showcase models – one is even a cutaway showing several different approaches to making paper (notably, laid and wove). They were really expensive ($5k), so I’m not about to dip them in a vat of paper pulp.

Essentially, traditional paper molds are composed of a frame covered by a mesh (the mold), which is topped by another frame (the deckle) that nests over the edges. Here’s the picture from Moore’s site:

The whole assembly is dipped into a vat of paper pulp, the water streams out through the mesh, and the paper is left on the top.

Seems pretty simple, no? But On close examination, paper molds have some fascinating features:

• The mesh is made by stringing tempered wires horizontally (called “lay lines”) and tying them in place with a “chain line,” a softer wire wrapped around each lay wire to create a line perpendicular to the lay lines. Easier to see than to describe. But you can’t buy this mesh, so I made it myself, using .035” stainless welding wire for the lay lines and 15 lb test.
• The mesh is supported by a series of transverse bars, airfoil-shaped in cross section. When you line these up side by side with the skinny end of the airfoil supporting the mesh, they form little channels of decreasing volume as the water flows through—that’s right, a venturi that sucks the water out the bottom!
• The deckle, which is just a picture frame basically, is held together with a fascinating joint that combines a sliding dovetail and a mortise and tenon.

It took me awhile to figure out how to make this joint, which like some Asian joinery has features you can’t really see from the outside.

Traditionally the woods used were oak or mahogany for the frames and larch for the support bars. I had some white oak, so decided to use that. Failing to find any larch, I decided to use Spanish cedar.

All in all a surprisingly challenging project!

Essentially, I just drilled a hole in the end of the screw, tapped it for a 3/8”x16 bolt, bolted the wheel on, drilled a transverse hole right where the wheel and screw met, then drove in a spring pin. Solid!

The only problem is that they stick out pretty far. If I bump into them too often, I might cut the hub of the screw down past the now-unused holes for the bar handles and drill and tap again.

Here’s the wagon vise:

And the leg vise:

On the back of the leg vise I (and my helper) upgraded the casters:

One is the wheels supporting the leg vise. I got some small rubber casters from a big box, and as it turns out they’re just not round. So I’m in the market for some better ones (suggestions on suppliers, anyone?).

The second is the $32 Shop Fox screw handles. The screws work find – the problem is that the handles are steel, with little steel ball-shaped caps pressed on the ends to hold them in place and keep them from sliding through the transverse hold at the end of the screw. Sure enough, I hadn’t had the bench finished for half an hour before I turned the leg vise screw, and the handle dropped to its end-stop (as bar handles are wont to do), and the ball popped off the end, letting the handle drop to the floor with a clang. So it was on to the Grizzly site to buy some cast iron wheels. Hope they work out better!

The first concept was to replicate in steel the approach I took in my first Roubo-style workbench in 1991, which had a dado in each side of the wagon that ran along a walnut track screwed in the sides of the wagon slot in the bench. Here’s what it looked like using angle iron instead of walnut:

As you can see, on each side I used three pieces of 3/4” angle iron: one short piece screwed to the wagon block, and two long pieces set in a dado in the wagon slot, forming a track. Call it a poor man’s version of the nicely machined aluminum extrusion that comes with the Benchcrafted kit, but costing $5 instead of $369. The downside: once I glued the bench together there’d be no way to remove the wagon to make adjustments.

So the next iteration involved moving the track to the bottom of the bench, and using a piece of flat bar stock to form the bottom of the track (the yellow parts are the sides of the wagon slot in the bench):

Then I got to looking at the width of the slot – 3”, which is a pretty big hole in the bench. So I decided to cut a rabbet in the wagon and route in a piece of wood to narrow the gap:

But that meant I had to shift the nut down about 5/8” inch, so I added some hardwood spacers below the bench:

That’s essentially what I went with – and with a little tinkering and fitting, it works great! (Pardon the messy shop!)

And from the bottom:

I did make a slight mistake: the endcap is about 1/4” too shallow to cover the end of the hardware. Oh well—I’ll mostly be looking at the bench from the top, not from the bottom.

Now, on to the base and the leg vise!

Miles

Fortunately, now we have digital tools so we can move quickly from iteration to iteration. Sketchup is the epitome of such tools in 3D design—as its name suggests, it’s all about making quick sketches that can at some point be exported to more sophisticated CAD/CAM or architectural information management programs. I’m still learning it, but it’s made a huge difference in my ability to adapt and implement ideas quickly.

The great thing about Sketchup is that because the investment is low, risk is low. We all know the feeling of investing a lot of time, effort, and money in a project, only to be faced by a problem or opportunity or idea at the end. What will happen if I change something now? What if I mess up all that investment? Sketchup allows you to take risks that would be unsustainable in the real world.

The downside of course is that you can spend all your woodworking time at the computer instead of in the shop. Here’s my thinking: Sketchup isn’t primarily a time- or labor-saving tool. It’s a quality-enhancing tool. It allows us to try things out cheaply and relatively quickly, with no guilt, then change them if something doesn’t seem to work out. Sketchup actually takes time to learn and use, but the quality of the output at the end of the design process is much improved.

I also like how Sketchup encourages us to design things ourselves, rather than rely on somebody else’s plans. To me, the essence of the whole woodworking hobby is the individuality and customization that comes from handwork. Unlike mere assembly, such as you might do if you buy a bench kit or a boxed manufactured bench, woodworking combines fabrication and assembly. We tend to use some store-bought parts: handles, hinges, drawer pulls, and so on. But how we put them together depends on the materials we shape by hand. We often think about hardware as holding our fabricated wooden parts together – but the reciprocal is also true: the wood is the fabric that puts the hardware in place, the integrative tissue of a piece of furniture. Brian Boggs once said he started building chairs because he could make the joints with a limited number of tools in a limited number of ways necessary for strength, but that between the joints his creativity could run wild.

The French call this “bricolage” – the individual’s assemblage of unique ingredients, sometimes in unintended combinations. A good example: Matthias Wendell making a bandsaw out of wood, using bicycle inner tubes for the tires.

What happens in this stew of homemade, shop-bought, and repurposed materials is that we get to make things that suit us. Say you buy a desk. It’s too short; you prop it up on blocks. It’s too wide: you move things around to accommodate it. It’s too tall: you buy a height-adjusting chair so you can adapt yourself to the object. Instead, what woodworkers do is make a desk to fit us.

That’s one reason making a workbench is such a great exercise in iterative design. What we’re seeking is an amiable combination between our bodies (height, reach), our activities (planing, chopping mortises, sawing, carving) and our workpieces. The purpose of a workbench isn’t just to hold work steady, but to hold it in a particular relation to our bodies and our tools.

Which is one reason why eventually we have to move out of the world of Sketchup and into the world of reality. We can draw designs all day long, but eventually we need something to hold, touch, stand beside, and use to figure out how our bodies will relate to it. The Stonehenge scene of Spinal Tap comes to mind: the band pays big bucks to have a replica Stonehenge constructed for their stage show, only to find that their request for a 18’ monolith was read as 18”. Their solution? Hire little people to dance around it. Not only funny, but an indictment of the broken communication between users, designers, and makers.

What’s great about woodworking is that we’re all simultaneously users, designers, and makers. So none of us is really building a Roubo bench – we’re building our own bench, fitted to our bodies, our needs, our values. For some, it’s a utilitarian tool; for others, it’s a display of skill or an opportunity for innovation; for still others it’s a connection to a heritage of woodworking practice that extends back centuries. Design means that we get to decide.

• It was inexpensive if you used pine. We were poor in grad school!
• I could make it with limited tools—I owned only a panel saw, a back saw, a jack plane, 3 chisels, and my grandfather’s brace, plus various old wrenches and screwdrivers.
• I could make sense of the wagon vise. (I never could figure out from the pictures in magazines how one of those L-shaped end vices fit together.)

I used 15 2×4s, four 4×4s for legs, 2×6s for stretchers, and a block of walnut I reclaimed from a tree that blew down in the back yard the year before for the “wagon” of the wagon vise. A hardware store was going out of business, and I procured a 3’ piece of 1 1/4” threaded rod and 2 nuts and washers, which I had welded together to form a flanged nut ($25 – a fortune!).

The whole process took weeks, planing each board between classes and studying. The walnut wagon for the vise was especially tricky to do by hand to close tolerances. I cut the dog hole with the brace and bit, squaring it up with a 3/4” chisel, then cut dadoes on either side to mesh with walnut runners screwed in the wagon slot in the bench. The vise corner received my first ever dovetails – ugly, but serviceable.

By the time I finished, I’d learned a lot about wood and its ways. But I never got to build much but the bench itself, as the demands of school and career grew. I eventually sold it to a neighbor before we left Kentucky, but regretted it ever since.

Call me sentimental, but now I’m remaking that bench. I’m a bit too busy (old?) to plug away with hand tools, and there’s more money available now, so I’ve built my shop up to include all the standard goodies: TS, bandsaw, jointer, planer, and lots of hand tools. But I’m still going with fir, because it’s inexpensive and I won’t wince (much) when I drop a chisel on it, the way I would with a nice hardwood. I figure I can always upgrade to a maple top if the mood strikes me. (The abstemious habits of grad school are character-forming!)

I also like to test my ingenuity, so I’m building the vises (leg and wagon) with hardware mostly obtainable from a big box hardware store. The exception is the screws, which are $32 Shop Fox jobs.

In my next post I’ll review my design process, which used Sketchup pretty extensively. Love that program!