Rough Wood

Ah, spring! With winter behind us again, all those months of Vitamin D deficiency are but a lingering memory. It’s time for actual sunlight, refreshing breezes, being buzzed by daredevil hummingbirds and getting the grill ready for some serious outside time.

It’s also time for me to take care of that pile of *stuff* that had to be put on hold while I hibernated. I saw the light last weekend, and started attacking a planter project that has been at the bottom of that pile. It’s a 3′ x 3′ (roughly) planter box with a bottom, in which we will grow a range of herbs in our tiny back yard. This one has to be finished in time for early spring planting, so it’s chop-chop time for me.

After some shopping around, I found that buying chunks of fencing at the widths that I needed would do the trick for less than buying and breaking down lumber from a hardwood source. With my now-indispensible keychain tape measure in hand, I headed off to Home Depot for some fencing. I grabbed 6′ lengths of 5.5″ x 3/4″ FSC-certified redwood (not surfaced) for $1.87 a piece. Yes, please.

I also grabbed some 8′ lengths of semi-decorative strips at 1.5″ x 1.5″, something that would have been less expensive to rip down from a board bought at a hardwood dealer. I don’t have a table saw, and ripping long strips with a hand saw isn’t my idea of a good time these days, so prefab wins this round.

On one particularly nice day, my saw bench, my ryoba and I planted ourselves on the patio and started breaking this stuff down. The interesting bit is that I reworked my original plan for the planter based on the stock that I bought, and accurately built the joinery (mostly grooves) into the SketchUp model. After spending a bit of time confirming the accuracy of the model, I then used it as reference for each initial cut. I did some rework on the sizing of the grooves as I went.

This thing isn’t really complicated, but it was the first time I’ve left the numbers up to SketchUp after building the joinery to specs.

For this project I’m going full-bore. For absolutely no reason other than practice, I’m cutting mortise & tenon joinery for the frame and grooving the panels in. I think practice is a good reason, as the next project is the tool cabinet. At the very least, it will be the most over-engineered  flower pot in the history of civilization.

Rain or shine, if you will.

On a tangent, I cleaned up my cuts and did some surfacing on the components with the new workbench and it went very smoothly. I can’t describe how much less fatiguing it is to do face planing on a bench that you don’t have to simultaneously hold down with a foot…

When working with something round, in this case dowels, clamping/holding is always a weird proposition. Round stuff doesn’t like to be clamped. More accurately, clamps don’t like to work on round stuff, at least without resorting to a clamp with rigid grips. In that case you’re likely to crush the dowel at the clamping locations. You can make a frame of some kind to hold your round in place, but then you’ve covered up the areas that you need to work on.

To the left is an appliance/jig that came to me a few days back, for use in cutting a groove with a plow plane. The groove will hold a small strip of hardwood and act as a spring for the dog. More about that later. The appliance is a 1.5″ x 1.5″ chunk of doug-fir with a 3/4″ hole bored into the end. A little less than half of that hole is sawn away, leaving us with a projection of a semi-circle that is just slightly more than half of the circle. This cross section is just right for working with a 3/4″ dowel, as the dowel literally snaps into place from the top, and doesn’t need any more vertical support. As long as the pressure applied is downward, the appliance keeps the dowel from spinning.

At the far end you can see where the hole is stopped. The idea is to saw the shoulder at a depth that more closely matches with the bottom of the hole. This saw cut was a bit low, but it still worked, so I didn’t need to make another.

What we end up with is a channel for the dowel and a clean top (and side) for a plow plane to register against. A groove is cut down the center of the top of the appliance, providing a path and support for the skate of the plane (and the blade) to follow. At this point, lining up the plane with the channel in the top and pushing it against the side of the block allows you to cut a clean groove right down the center of the dowel. I’m sure that a jig can be created for doing this sort of thing with routers and other wood erasers, but doing it this way is downright pleasant. Your mileage will definitely vary.

Once the groove is cut, you can spin the dowel around in place and clamp it down to the block, as shown at right. This works because the block is providing some friction too.

Cut a cross-grain kerf at the depth of the holding area of the dog and use a shoulder plane to nibble away the waste. I didn’t bother with perfect registration for the plane in this example. In fact, I angled the plane to the left a little to undercut the holding face for better support. I recommend that you consider doing this at about 90 degrees around the dowel from the groove, rather than on the direct opposite side, so the spring doesn’t completely compress when you use the dog. It’s not a deal-breaker, but the dog will more quickly shimmy its way into the hole in this configuration.

My grooves are about 1/8″ deep, and about 1/4″ wide. After ripping a 1/8″ thick strip from a 3/4″ thick piece of scrap I had 3 springs all lined up and ready to be freed from their siblings. For tiny stuff like this I like to use a marking gauge to make cuts. It’s clean, accurate and a lot less fiddly than anything else that I’ve tried. It’s also much, much less dangerous than using machines for small work. Just set the depth of the gauge to 1/4″ and drag the gauge as if you were marking a line. Do this repeatedly, and with more pressure than you might normally use with a marking gauge, and the little wafers will eventually slip right off.

After cutting, the strips are sized and trued a bit using a block plane held upside down in a vise. This is a bit dangerous, but as long as you’re careful you should manage to keep the skin on your knuckles. After a bit of cleanup on the strips, they’re almost ready to glue into the grooves.

The only task remaining is to cut a slight facet at the end, which will be the portion that is glued into the groove. This angle determines how much the spring protrudes from the side of the groove, so it shouldn’t be too severe. Keep in mind that springs are not a part of the clamping operation of the dogs, they really just keep the dogs from falling through the holes in the bench. I placed a sheet of sandpaper on a granite block, tilted one end of the strip up about 15 degrees, and sanded a facet on the end that is about 1/2″ long. Time for glue!

After a bit of glue-up, the extra bit of spring is trimmed from the end, and voila, bench dogs.

The process is really quite simple with the right support. I quickly made a bunch of dogs for the surface of my bench, and a few half-height versions for my vise, as the holes do not go through the entire face. At the worst I will find a flaw in my process and have to build more, but at about 25 cents each, I can quickly, easily and inexpensively do some experimentation.

Please note that there is an earlier post with more details about construction, which can be found here. Thanks!

It’s finished. Last night I put the final touches on the workbench. After a few coats of finish on the vise it was ready to assemble and put to work. I immediately made some bench dogs. It was a true revelation to have a real vise and a relatively stable surface. It’s not a massive, stationary bench, but it’s so much more than I had. It was made with hand tools, with the exception of the use of a drill for the many, many holes.

Here it is with the vise handle removed and ready to break down. The gray discs on top are what I call “bench slippers.” They are short bench dogs with nylon furniture glides attached. When the legs are removed, and the surface is flipped over, the slippers are inserted and the bench can be slid around on wood floors without damaging them. More on that later.

My design is literally full of holes. I made a large array of dog holes in the surface. It may seem like overkill, but I expect to use random things like the Veritas “surface vise,” as well as using dogs to box in various rectangles on the surface. With only 1 vice to handle all tasks, flexibility is key.

The vise works quite well. It feels very stable and predictable. My only complaint is that it screeches a bit, even after extensive waxing. Without a drill press, the angles on the screw and the 2 guide rods are not perfect, so there are various slightly opposing forces involved. It works great though, being both smooth and steady. The external garter works like a charm. The sapele face/chop is working great and looks amazing. There are more details on its construction here and here.

If I did not have soft wood floors (bamboo) I would have bought a metal vise and been done with it. Sometimes things work out for the better, because I really like this vise (so far). If you have the means to make the bench but not the vise, buying a metal one is definitely an option. The vise is by far the most complicated part of this project.

Once broken down, which involves removing the two leg assemblies, the main surface sits flat at about 5″ tall. With the 2 leg assemblies, it breaks down to a stack about 8″ high (+ the glides).

In the end, I had purchased about $40 worth of wood for the bench itself ($30 for the poplar top, $10 for the 2x). The vise is scrap, other than the face. If you have scrap around you could use that and make adjustments to the design. I bought knobs and long carriage bolts for temporarily joining the leg assemblies to the top, but don’t think I’ll need them. I used about $4 worth of the huge sapele board that I bought for this and other projects. Add $20 for the hickory dowel and we come in at around $64 for materials.

Here’s a link to a SketchUp file of the final build. There is no joinery in the Sketchup. Dog holes are represented by black cylinders so they can be rearranged. Please keep in mind that I am left-handed. If you are not, you might want to consider the orientation of the vise.  If you build it, please let me know, as I’d love to see it out in the wild. Please note that it does flex in the short direction due to the lack of stretchers there. I may add some supports that can be temporarily bolted in, but I want to use it a bit first. It’s quite solid in the long direction due to the existing stretchers.

Now I can actually build some stuff… Next comes the tool cabinet!

It’s so close.

My design for this vise involves one giant chunk of scrap cherry which serves as both the nut (the block containing the female threads) and support for 2 huge alignment rods using the same 1-1/4″ hickory that makes up the screw. Meet me over in Part 1 for specifics on the creation of the screw itself.

The process of tapping the nut involves drilling a hole that is slightly smaller than the original dowel (according to the recommendations of the manufacturer), then using the tapping device. I again used camellia to make the threads less prone to chewiness. Note the try square used for reference, as it was used at every step. This will be significant later.

The tapping went swimmingly. Clean, quick, effortless. Something must be wrong. Next slide.

Once the nut was tapped, I reluctantly drilled a hole through the apron of the bench and temporarily attached the chop/face using the screw. This acted as a guide for drilling the alignment rods through all 3 layers (chop, apron, nut). It was much easier to do this from the outside. It’s not as pretty, but we have ways…

After more wrangling with the slightly out-of-round hickory dowel, I had 2 alignment rods passing through all 3 layers. It was starting to feel like a vise. Time for some glue.

While waiting for the rod-chop glue assembly to dry, I made the garter. Since I drilled the holes for the rods from the outside I wanted to cover up the ends of the hickory, so the garter looks more like a plate, and covers about 1/4 of the chop.

This has been, by far, the easiest part of this project.

Back to the screw, I had to work out a way to cut a garter groove into the dowel. My solution was a sort of makeshift lathe. I drilled a 1-1/4″ hole into a stop block and used the hole in the screw hub as another support. This left me with a free-spinning dowel between 2 blocks that I could clamp to the (old) workbench. After some chisel work, which was handled a bit like proper mortising, except… weird, I had a very rustic garter. It wasn’t as clean as I had wanted, but it was a reasonable way to get the job done.

There was however an unreasonable amount of adjustment involved in the fitting. Without a drill press, it’s quite difficult to get squared holes, especially at the scale that we’re talking about. A hole drilled slightly out of square will have a far greater impact when the total distance bored is 5″. When you factor in 3 rods in a tight space—all of which should move in unison—bad, scary things happen. I was able to adjust the projections a bit with some trial and error (and a rasp), but a drill press would have alleviated this. Time to get a brace and do some practicing. Workbench first, tool cabinet second, bit brace third.

I am closing with a photo of the vise as it stands (no pun intended), from below. It is rough and loosely assembled, waiting for glue to dry before proper shaping and refinement. From this angle you can see many things not complete: the hub needs shaping, the handle needs caps, the chop needs a good shape, the nut/support block needs to be cut down to final size.

Once the glue dries I’ll be back in awesome fun-time zone. This part was very technical, which is not awesome fun-time zone. I can’t wait to have a real bench, and a real vise. The first responsibility of the new bench will be to help me make accessories for itself—which really feels like a goal accomplished.

As part of the knock-down workbench project, I’ve been scheming an all-wood vise. Considering that this bench needs to fit under my bed, and will regularly come into contact with our bamboo floors, I don’t consider a metal vise to be a viable option.

So I bought a manual threading kit and some 1-1/4″ hickory dowel, which was the only 1-1/4″ that I could find locally that wasn’t avocado, and called in the attack.

The threading kit that I bought from Highland Woodworking worked reasonably well right out of the box. I expected to have to sharpen the cutter at least, but it really did a number on the hickory. The directions indicate that a soak in tung oil (or something similar) is often needed to soften and lubricate the fibers. I chose to use camellia, as I have an abundance of it . The other options are much more of a mess, and need to be neutralized or cleaned off after the soak. Too mucky. A light slather of oil and the threads magically appeared from the backside of the threader…

In reality, that’s not how it went at all. As it turns out, the dowel was slightly oval, by a mere .002″ or so. In a softer wood, you might be able to cram it through anyway. In hickory, not so much. Hickory is about 30% harder than maple, for the record. So I attacked again, this time with sandpaper. After a lot of sanding and precise measuring (man that stuff is tough) I had the seemingly minor oval tamed, and started with the threads.

They are not perfect. They are a bit chewy in areas where the layers of earlywood and latewood are thin and close to the surface. This may be where some oil soaking would help. These patches of compromised thread are generally linear, jumping through multiple turns of thread. They are only about 2 degrees of a full turn around the screw, so they shouldn’t affect the holding strength at all. I’m totally fabricating the number, but imagine two continuous feet of tongue-and-groove (on a smaller scale) with about 1/4″ missing. Now imagine trying to pull the boards apart along the faces. They would not go anywhere.

I’m thinking that the side of this bench will fail long before this vise, and I may have it for many years to come, long after retiring the new bench. In a little cartoon in my head, I tighten the vise so much that the apron for the bench just pops off and the bench falls limply to the ground with a workpiece still sitting on top. Cue a little puff of smoke.

I don’t have a decent vise, in which to hold the threading contraption or the tough 18″ of dowel. At times it felt like I was making a circuit in a gym—to the floor, using the shoes for grip, to the old bench using the weird thin surface-type vise, back to the floor. Eventually, as I approached 14″ of completed thread, the threader stopped cold. I hadn’t perfectly rounded a very tiny section of the dowel. After losing some of the callouses that I’ve worked so hard on lately, I backed it out and sanded a bit, then finished.

In the end, I had a stick with threads on it, and a workout.

Once that was resolved, I bought some wood for the chop (or face) and moved to actual planning. I’m going to do an external garter, and I am now devising a way to cut the garter groove around the screw without a lathe. Next up is drilling the holes for the screw, the guide rods for the chop, and the seating of the screw into the hub. If I’m more than a degree or two off, my vise will tend to bow my workpieces. Here’s where a drill press would enter the picture, if I had one…

To be continued!