(Photos with a thick border were taken by Jody Kolodzey.)
Well, I had to get around to it sooner or later. Over the past seven years, I've bought a table saw, a router and, most recently, a sliding-compound miter saw. I spent several months building a workshop in my basement, making tweaks and revisions here and there. And I spent some time stripping paint off of furniture I found in my house, making it look a lot better than it had when I found it.
It was time to take on a real woodworking project. There were lots to choose from, when you consider how many projects had been published in the various woodworking magazines I'd been reading. But I finally settled on something really simple, in theory at least: a wooden radiator cover. And I wanted to design the thing myself, more or less: no design's wholly original, but I did want to be able to say that I'd made it from my own drawings.
A radiator cover's pretty simple. It's got three sides, a top, and some kind of grating in the front so the warm air can circulate. I made some measurements of the radiator in my living room, which showed me that my project should enclose a rectangular cube 43 inches wide, 27 inches high, and about a foot deep.
The first idea was to try a really basic, Arts and Crafts-style design. It'd be four posts, maybe an inch-and-a-half wide, and the front and sides would be made up of those thin woden slats that turn up on the sides of Morris chairs.
Now, to an experienced woodworker, this doesn't sound as though it'd be much of a challenge. But I ain't got that experience. My joints seem to come together just a hair off of ninety degrees. And I'd never even tried a mortise-and-tenon joint, mainly because I could see how I'd mess it up tremendously-- shaving the tenon too thin, misaligning the mortise so the pieces are off too damn much. But I'd have to try it, sooner or later. So I took some cheap pine down and started cutting the parts for a prototype.
So there it is. The "slats" were cut from 3/4" pine, and were 1/2" wide. (I did this to see how they'd space out. I could lay them so either of these sides was visible, and could see what spacing would be most visually pleasing. So the slat could be 1/2" wide, or 3/4" wide, and the spacing could be variable. Generally, the gaps had to be no larger than the slats: if the gaps were bigger than the slats, it didn't look that good. So for the prototype, I did 1/2" slats and 1/2" gaps. I took those boards that run horizontally, cut 1/2" dadoes in them (like a really rough hair comb), and put the slats in the dadoes.
Before
the assembly, I tried staining the pine with a mahogany stain (Minwax gel) to
see how it looked. This picture also shows the spacing issue a bit more clearly.
Yes, that is masking tape holding the stuff together.
Now I had these comb-assemblies, and I had to attach them to the posts. The best joint for the project was the classic mortise-and-tenon joint. But I'd never tried one before; and while Norm Abram can use an expensive mortiser machine, and while others can weild chisels carefully enough to smooth out their mortises, I just ain't got the skill just yet.
I decided to try this Beadlock floating-tenon jig. So why would I try this device, rather than learn how to do mortising by hand? It's pretty simple. It was cheap, so if it didn't work, I hadn't wasted much money.
As it turned out, the mortise-jig thing worked pretty well. The tenons seemed to have some "play" in the weirdly-shaped mortises, but as I said, I had to re-drill the holes and trim the tenons. But it held well enough for the prototype.
It
was time to example the prototype, and maybe learn a few lessons. The first
lesson was that, for some reason, I'd made the front face 51 inches wide instead
of 43 inches. (I know I did this for a reason, but I can't remember it.) So
the front face had to be shortened. Also, when I pressed down on the middle
of the front face, it felt more than a little fragile. Both of these problems
were easily fixed. I simply removed eight inches from the middle of the front
face, and joined the two halves behind a new post in the middle to prevent that
sag.
Okay, so it's not the prettiest solution. OKAY, so it's just a two-by-four with some slots cut in for placement.
But the prototype did show me that a decent mortise-and-tenon point was possible, so I didn't have to worry too much about misalignments too much. And if the joinery was within my abilities, then that gave me the OK to buy some decent hardwood for the project. This meant that I could be a little freer in my design; in fact, I could even abandon the Arts-and-Crafts style, which I'd adopted mainly because it was all easy angles and straightforward joinery.
I sketched out a few ideas, and finally decided that I was going to include some Greene and Greene-style touches on the radiator cover. For example, the board which ran along the top and bottom of the slats would now have a G&G-style "cloud lift." Sketches on graph paper showed me the sizing I liked the most, and I started working out how I'd actually build the thing.
Turns out a lumbermill in walking distance had some 11" wide mahogany boards. I got 24" feet of the stuff, and with the help of a friend, got it home. For what it's worth to anyone-- I got my wood at the Patterson Lumber , 47th and Woodland Avenue, here in Philly. It's within walking distance-- and that's how my friend Brenda and I hauled the crap back to my house.
I
still haven't completely settled on the design. But, whatever I came up with,
I knew I'd need four posts, each about 26-28" long, and perhaps 2 1/4"
wide at the most. I cut myself two three-foot pieces of mahogany, and glued
them face-to-face, with the idea of cutting them to 2 1/4" widths later
on. The photograph at the left shows the cheap clamping system I used-- I piled
a bunch of bricks and two boxes of books on top of the two boards.
This was my First Big Mistake. If you look at this photo, you will see that the cupping of the mahogany-- along with the sheer inadequacy of several bricks and books for the job-- made for one of the worst glue-up jobs imaginable. So, I just wasted about six board-feet of mahogany-- roughly forty bucks. I put these chunks aside, and resolved to find some way of using them for something else. (I did.)
So for my second try at the legs, I decided to cut to width first, and then glue them up. I cut some lengths of wood, 2 1/2" wide, selected them for the best faces, and glued them back to back. Here, you can see that I wrapped them with wax paper to keep the glue confined. And you can see that I've learned the Greatest Lesson A Woodworker can Learn: You can never own too many clamps.
When
I took the legs from the wax paper, I found that they'd glued up nice and tight
and reasonably true. The photo at right shows what they looked like when I removed
them from their wax-paper wraps after about two days in the clamps.
Two things surprised me. They glued up fairly true to each other, which was nice: it meant that I wouldn't have to do much jointing to make them nice and smooth. But when I picked these things up, they felt heavy. They didn't feel like two strips of wood glued to one another, and they didn't feel like a pine two-by-four. Mahogany is a pretty dense wood, and these things felt like bricks.
So now I had to make'em look nice. This is where a woodworker would shoot them through his/her jointer, to true up one side. Thing is, I don't own a jointer. I did experiment with my router and a 2" long straight cutting bit, but I couldn't get it to work exactly as I'd wanted it to.
This was when I had my small revelation. Why did I have to rely on machinery so much? There's a thousand years of woodworking before electricity. Haven't woodworkers been able to true up their lumber without big, powerful, growling monsters kicking up clouds of dust? Haven't I been immersing myself in the books of Patrick O'Brian, imagining the ship's carpenters repairing the cannonball damage? Didn't David Theil have a recent segment on scrapers and handplane sharpening?
So
I took out the bench plane I'd barely used before and, after a little adjustment,
started scraping away at the leg assemblies. The results were really pretty
nice. The faces I planed came out nice and flat, and the grain really popped,
much more so than if I'd sanded the things flat.
And part of my revelation came from the actual act of planing the wood. I put the pieces on my workbench, anchored by a pair of benchdogs, and began planing. You know the physical motion, right? You place the plane down, you press on the front knob to engage the blade, and you swing your back forward while pushing with your elbows in a nice, fluid motion. And it's this action that really makes you understand what "working the wood" is all about. The wood is being shaped by your muscles, and you're not just gingerly guiding it past some whirling blade or bit. It felt good, and it felt really good when the final result was as smooth and as vibrant as the mahogany legs I made.
Eventually,
I rounded the corners, using a 1/8" roundover bit. The results, as you
see, are actually pretty nice. The legs are about as straight as can be, and
the grain has almost no real or worrisome defects.
By the way, remember Greene and Greene? I arranged the legs so that they looked a little bit like the joinery in the stairway of the Gamble house. Here's what I got.
Okay, the prototype showed me that I'd need a slew of 3/4" wide strips, 1/2" deep, for the front of my radiator cover. I cut these with my tablesaw.(I still hadn't decided to space them at 1/2" or 3/4", so I cut enough for both uses.)
Below, you can see how I made the strips look a little nicer. I cut them to a rough length, and placed them side by side, so I could sand then in batches. The photo shows the round dog-holes I cut in my workbench, and two brass dog hold that strip at the right hand side.
Now, here came the tricky part. I had the strips, and I had the posts. What about the long, horizontal pieces which would hold the strips? They'd pretty much be the pieces which "held" any degree of style. I'd decided to use Greene and Greene-style cloudlift patterns for them, and after some sketching with graph paper, I settled on a nice, simple design.
The general idea was to create a template which I could tape onto the mahogany, and use a flush-trim bit in my router table to cut the mahogany to the desired shape. The photos below show a piece of 1/4" MDF, on which I've drawn and cut the outline of my desired cloudlifts. I did this for only one end, as you see.
I took this MDF template, and used it to create a full-sized, double-ended template in pine. I then shaped this pine template nice and relatively smooth (rasps and sandpaper, of course). Then I double-sided-taped the pine template to the mahogany, and used my flush-trim bit to cut the mahogany. Net result: two pieces of mahogany cut with cloudlifts at each end.
(I did make one huge mistake here, and as I don't have any photos, I'll have to draw it for you someday. But, when I was routing the mahogany, the bit grabbed the inside curve of a cloudlift and tore a nice, big chunk out of my finely-worked piece. It splintered clean, however, and an overnight with glue and clamps set it back to rights. I then routed it far more carefully.)
Now, I had to cut dadoes into those cloudlift pieces. I finally decided to make them 3/4" wide, spaced 3/4" apart. The obvious strategy here was to use a dado blade in my tablesaw, and use the principle of a box-joint jig to make sure the dadoes were cut nice and regular.
The photos above show the jig I built. Actually, I had most of the jig built for some other purpose, and all I really did was glue a piece of a strip onto a piece of plywood and clamp it to the front.
The photo below shows my test cuts on a piece of plywood. I put the strips in to test a few issues. The first was to see how attractive the spacing was. the second was to make sure that they fit, or barely fit, into the dadoes: if the dadoes were too snug, that's fine, because I could always sand the strips thinner. (The plywood was handy in another respect: if the strips were a hair too wide, they would push at the sides of the dadoes, and cause the plywood to curve.) But, once my test dadoes looked OK, I went ahead and made the cuts on my mahogany cloudlift pieces.
Now it was time for the first stage of Major Assembly; putting the strips into the dadoes in the cloudlifts. This went without too much sanding or scraping. I then cut pieces of mahogany that were roughly the same dimensions as the cloudlift pieces, and glued them onto the backs of the cloudlifts, sealing the strips in place. The photos below show both the top and bottom of the cloudlift/strip assembly. So, you can see that this 1/4" backing was actually kind of piecemeal, with wood putty filling in the gaps. (It's hidden in the final product, as you probably expected.)
But
this was the first time I could get a really good idea of what the final product
was going to look like. The photo at right was the First Look at the front of
my Radiator Cover.
And I thought it looked pretty nice. The pieces intended to be visible looked nice and smooth and whole. The router mishap hadn't disfigured it at all. The strips weren't exactly grain-matched, but I didn't mind that. This was the first time I could see, outside of my mind's eye, what the final Radiator Cover would actually look like, and I actually managed to impress myself.
As
I said before, when I made the prototype, I experimented with using the Beadlock
system to create the mortises and floating tenons. That's what I'd planned on
using to attach the legs to the front assembly. I put the pieces next to one
another (left). Notice that I did not align the top edge of the cloudlift to
the top of the leg. I wanted to have a 1/2" gap to enable the heat from
the radiator to flow out into the room. (Oh, that piece to the right, the one
that's only a quarter-inch thick? I'd thought of having panels like this in
that cloudlift area. Decided against it. More about this later.)
I
made my marks very, very carefully, clamped the Beadlock jig onto the legs and
cloudlifts, and started drilling. At right, you can see what the Beadlock mortises
looked like in the ends of the cloudlift pieces. You can also see the wood putty
cruft in one of the curves, too.
So, what went wrong? Somehow, the mortises on the legs and the cloudlifts did not match up correctly. I sort of expected something like this-- when you use the Beadlock, you have to orient it on your mark in two different ways, and there's a margin for error there. But then again, I probably rushed it, and given that I had to shim them slightly to center the mortises, so let's assume that the problem was of my own creation. But the mortises and the tenons were just a hair off, just enough to make a serious clamp-up and fitting damn near impossible.
How did I fix it? I shaved and sanded the tenons down until they fit, that's what I did. The final assembly was maybe 1/4 out of square, which wasn't noticeable.
By the way, if anyone from the Beadlock company is reading this, may I make a suggestion to improve your product? Please engrave a centerline groove on the sides and ends of your little black pieces. This way, we can align the thing more accurately, and in two important dimensions. Using the centerline on the sheetmetal piece just isn't accurate enough.
At this point, I knew that I was not going to use glue, or Beadlock tenons, to attach the as-yet-unbuilt side pieces. I decided to use pocket screws instead. So, since I didn't need to maintain a finish-free surface, I decided that I ought to apply a stain and polyurethane to the Front Assembly. Minwax's Mahogany Gel Stain worked well for me in the past, but this time, I didn't want the grain to 'pop' as dark. --- gave it a warmer, more honey-like character. Thereupon followed three coats of glossy polyurethane, slightly diluted with paint thinner.
I
decided that, on the sides of my radiator cover, I wouldn't bother with cloudlifts
and strips. The sides were only a foot deep or so, which meant that the curves
of cloudlifts would be kind of tight. Instead, I decided to go with something
closer to actual panels, which would throw more of the warm-air flow
to the front of the radiator cover, and thus into the room.
The design for the sides was really simple. I cut four pieces for the Tops and Bottoms, roughly a foot long, and an inch and a half high. I cut a 1/4" dado down the middle, into which the panels would rest. I then decided that the side assemblies would hold three thin, 1/4" thick panels of mahogany-- one wide one and two relatively thinner ones. (I made little mahogany spacers to sit in the dados to keep the panels spaced properly.)
I then stained and polyurethaned all of these pieces separately, and used a Kreg pocket screw jig to attach them to the backs of the legs.The photo at right shows what the side of the finished cover looks like.
And
where did these 1/4" thick panels come from? Do you remember that botched
glue-up of the legs? Those three-foot-long, eleven-inch-wide pieces of mahogany
I'd glued to one another so ineptly? Well, I fed them through my tablesaw, slicing
1/4" thick panels that were up to 3-4" wide. So, what was once a loss
of 6 board-feet of mahogany was now a loss of maybe one.
I also added a four-foot-long board to run across the back, making a complete rectangle around the radiator. This would help support the top, when I built it. The photo at the left (top moved to the side) is a close-up that shows how the Kreg pocket screws amount to so much of the joinery.
The top of the Radiator Cover was really an improvisation. My original plan was to buy a nice, big, green piece of marble or fake marble, and simply lay it across the four legs. But that turned out to be extremely expensive. And I didn't have enough mahogany to make a solid-wood top. Just as well, I figured: I expected that the heat from the radiator'd just cause a solid-wood top to warp and cup.
I finally decided on a really simple design: a "picture frame" made from mahogany, but the central surface would be made up of tiles. I had enough wood to make the four sides of the frame; two four-foot-long lengths, and two one-foot-long lengths, all about three or four inches wide. I could use pocket joinery to assemble these into a large rectangle. In the middle, resting in a groove in all four sides, would be a floating panel of 1/8" plywood. Resting on this panel, framed by the mahogany, were four grey-green ceramic tiles I picked up at the Home Depot for about $1.50 apiece.
And once this assembly was stained and sealed, I laid it on top of the frame I'd built. And it was done.
And here it is. Keen-eyed readers will notice that the top isn't centered to the frame, but that's OK: it's not fixed to the frame, but just sits on top. You'll also notice that the front cloudlift piece doesn't reach the top, but that's because I wanted a half-inch for airflow there.
You can't do something like this and not learn a lot. At the most basic level, I learned that using hand tools can be more satisfying than relying on expensive machinery and complicated jigs. I'll probably be investing in hand planes, scrapers, and sharpening systems, and I may be able to put off buying that dust collector for a while. (Right now, I use a shop vac and 2 1/4" hoses.)
As for joinery, I found that the Beadlock system works fairly well, but it's not foolproof. (As I said above, centerline marks'd be a big improvement for the Beadlock.) So, for the time being, I'll stick to the pocket screws and the Kreg jig, which works wonderfully for most purposes.
One lesson, which sounds fairly cynical, is this: the wood you use, and its finish, probably do more to "sell" the final product than anything else. The cloudlifts and curves and stuff were nice, and I liked doing them, but the finish and the poly really made this thing pop. So I am probably going to pay a lot more attention to Michael Dresdner's columns in the future. Finishing this was the toughest part of the project, for lots of reasons: uneven poly layers, messiness, doing the parts piecemeal so poly'd surfaces didn't adhere to anything else. And it meant letting layers dry for 24 hours each. (Personally, I'd love to find a finish that gives red oak or pine that wonderful honey-crimson color. The Minwax line doesn't seem to do it for me. Any suggestions?)
But the best lesson here was this. It was just good to learn that I could make something that was actually impressive, and that if I did this again I could probably turn out a project that'd be close to a professional level. I was also happy to realize that I did this without anything like a formal blueprint or design. This project was almost improvised, believe it or not. (Sorry, no plans are forthcoming. If I had plans, I'd be glad to share'em, but I just didn't make any.) So now I know that I can tackle other projects without too much worry that they'll be wastes of fine wood and precious effort.
So, what's next? I'm not sure. But right now, it's a toss-up as to whether I'll try doing a nice big bookcase in my living room, or a nice Art Deco surround for the massive 72"-wide video screen I'll probably never buy.
Copyright 2000-6 Brian Siano
(unless otherwise noted)