I don't normally write a capstone entry for a project. But I don't normally get a project like this one. There are bookcases, and there are Bookcases. These are the latter. And, it's not every day that I get a project that really allows me to stretch my creative muscles, and do some real problem solving.
I'd used sliding dovetail joints before, on narrow pieces, but never really had the chance to work with the real-deal housed tapered dovetail joint in carcase construction. I'd seen this type of joint cut before, but seeing one done, and doing
several in an actual piece is a different story, and there's a lot that I
learned about just what factors need to be borne in mind during the
Starting with making the tool that allowed me to do the joint, this is a really cool technique that I've been dying to learn for years, and hadn't had the opportunity. I will revisit the dovetail plane, I think I'd prefer it if the blade skewed the other way, but I'm not going to rebuild the existing plane. I'll probably make a new version with another skew rabbet plane that I own.
This is definitely a joint that I'll be using again.
It's been a while since I've gotten to do any serious tool making. The dovetail plane and the flush cutting plane both were a lot of fun to make and use.
I'd been wanting to carve stopped flutes for some time. I've seen them on a lot of really fine work before, but hadn't had the chance to use them, or figure out how I was going to go about it. I had a few ideas on how I thought I would carve them, and some of those methods will be revisited sometime soon, but not just yet. My final technique surprised me.
I haven't had to install moldings on such wide solid panels before, and never had to take the time to work out the details of the dovetail keys, or the details of the base molding.
Lastly, I hadn't worked with inset tops made of a different material before. It wasn't such a huge technical challenge as the other details, but I will say that they made a huge aesthetic difference, and the mix of materials looks a lot finer in person than something that's entirely made of wood. And I can say that it felt really nice when I delivered the bookcases to see the tops slide neatly, and precisely into place.
Looking back, there's a lot that went into this project, and it was a real joy to do all of it. And, somewhere in the middle, my son Gabe showed up, and that's put a brand new spin on a lot of things.
This is a tool I made for the bookcase project. It's not the most inspired tool I've ever made, but it did serve its role admirably.
The germinating thought was that I was tired of the 'worms' left behind by nicks in the planer knives: I've had several instances where this small bit of wood would hit the fence on the jointer when I go to square an edge, and it throws everything out of square. There are a number of ways to deal with that, I know. But this one occurred to me, and it seemed like a good way to move forward.
The basic principle is simple: The edge of the plane gets adjusted to be exactly flush with the sole of the plane. There's a learning curve, as with any new tool or technique. But in practice, this allows the plane to slide across a surface, and remove any discrepancies without affecting the rest of the surface.
It's related, in construction, to a chisel plane. But the extra length of the body allows for a few different grips in practice. And the high pitch on the blade (57 degrees, I think) keeps it from causing too much tearout.
I don't remember where I found this particular tip... I think it was an old Fine Wood Working magazine.
Because the case is so wide, there's going to be a fairly significant amount of wood movement, so brad nails won't cut it here. The case will likely expand and contract by as much as 3/16" over the course of the passing seasons here in New England, so it needs to be able to do so independently of the molding.
So, to hold the molding to the case, I cut a dovetail groove in the back of the molding, and cut matching splines that were glued and nailed to the case. Once the splines were attached, I slid the molding into place. The front molding I glued on completely. Because it's long-grain to long-grain, it doesn't need the splines, but they certainly help with alignment. The side moldings were glued onto the front-most spline, and the first few inches of the case. The rest is only held on with the splines.
Cutting the splines doesn't have to be a nerve-wracking exercise in precision. If they're a smidge loose when you make them, you can take a shaving or two off of the mating surface, and it will tighten up the final joint. Because they're not going to be regularly visible, it's not a big deal.
In the picture below, there are two gauge lines. Originally, I thought it would be good to use a gauge line for the nails, which were set up in pre-drilled holes in the splines. Bad plan, minor discrepancies made a mess. So, there's the second gauge line, which is simply for placement of the upper edge of the spline, which worked much better.
The bookcase job called for stopped flutes on the pilasters. It's a carved detail... has to be... but 8 pilasters with multiple stopped flutes is a lot, and I was wracking my brain to come up with a viable technique. This is what I came up with.
The flutes I cut on the router table, using a plunge- and stopped-cut. I stopped them before the top of the flute stops.
The stopped part needed a rounded over appearance. I needed a good way
to rough this out, and what I finally came up with was to use an ATB
blade on the table saw, barely protruding through the table insert, because of the tooth bevel: It allowed me to get
in, and basically rough out each side of the stopped part, one side at a
Some of the cuts were stopped cuts, some were plunge cuts... getting the blade inside the already cut flute, and cutting outwards.
The last part was to terminate the stops with a carving gouge, and to round them off.
The whole point of making the dovetail plane was to use it on this particular job: 4 large, fixed shelf bookcases. The carcases are single-plank wide cherry, with a finished dimension of around 16".
I used the same technique I'd used on the small footstool to do the joinery here, but with 24 joints to cut in total, (4 cases, with three shelves, 2 joints per shelf) I got a lot more practice, and a little more understanding of the process, to boot.
To make the process faster, I actually ran half of the joint with the same router I used to run the grooves. I did the final fitting of the top side with the plane. I have some funky Festool attachments that allow me to do this, but I'm sure it could easily be set up on a router table, too.
Fitting the joint, first step was to clamp the board down flat, and start working the end of the board. The goal is to have joints that don't have gaps showing at the front. Gaps can always be filled, but I like it better when I don't have to. I started the process with calipers, trying to match up the male end of the joint to the female end.
Eventually I just moved to fitting the ends of the joint to the case. That is, plane the front to fit the front of the case, and plane the back to fit the back, and then work the mid-section until the board slides home. It's important to note that the board has to be flipped for this, because the left side has to be fitted to the left side, and the right side has to be fitted to the right side, before the board is actually able to slide all the way in. The idea is to get that part fitted first, and then leave it alone.
You can see here that the fitted boards are not all the way in. This is done on purpose. Because the joint will wedge itself in place when it's glued up, wedging everything together now isn't really helpful... you'll just have to beat up the work to get that board back out. So, getting the board to slide basically to that last inch or so is pretty typical.
On the clamp in the photo, I encountered a couple of times where the boards had started to bow on me. It could be because a few days had gone by since I'd milled them down, cut dovetails, etc, and they'd moved a little bit. It could be because the wood reacted slightly when I ran the dovetail grooves. I'm not entirely sure. I know that some folks insist on doing carcase joinery the same day the wood is milled, because it helps to stabilize everything, and keep it in plane, but with 4 cases to do, with the width of the material, that didn't really seem like an option for me.
That said, there were a couple of the shelves that were slightly cupped. (about 1/16"- 3/32" across 15.5") I figured it would be fine, in the long run, because it would be held flat in the case, and would stabilize over time. But now, I think that doing final milling before fitting this kind of joint makes sense. Clamping a board flat while you plane it, AND fit it, AND try to do assembly is a lot of extra work. Working with flat stock is much easier. And assembling it is easier.
Last word of advice: be SURE to get those boards in to that last inch. I got sloppy on one that was slightly cupped... I figured 2-3" was fine, that the cupping would flatten out, and everything would be fine. This is the part where I confess to the fact that it took 2 I-beam clamps, with pipes slipped over the handles, to crank that shelf into the case all the way. It creaked loudly, like an old door, for about 5 minutes while it slowly shifted into place. I was sweating bullets the entire time, thinking that for sure, the thing was going to explode. Only push in that last inch.
The first step before joining a case is to smooth the inside surfaces. On this project, the sides are all pretty big, and I was in a hurry, so I opted to sand instead of plane them.
I've tried various sanding pads, designed to hold the work to a surface. I was using something like this, but once it started getting dusty, it stopped working. On a whim, I pulled out a chunk of leather that I've had for a while, and clamped it to the bench top, stretched out.
This is the best sanding pad I've ever used. I don't know or understand why leather holds as well as it does, but the Festool Rotex is a steam train of a 6" sander, that switches between rotary and random orbit modes, and the wood held to the bench just fine. Better than fine, actually... it felt like it was being held in place by a magnet. Very impressive...
Next, I need to mount some of this stuff to the inside surfaces of my pattern vise.