I've always liked the simple lines and understated elegance of furniture from the Arts and Crafts / Mission period of furniture design (most of it, anyway). I realized early on, however, that the only way I was ever going to own anything like that (of any appreciable quality) was to build it myself as most of the quality furniture in that style would be considered "boutique" furniture..... complete with a "boutique" price of at least one arm .... possibly a leg ... maybe an eye.....
I did a bit of searching and came up with some plans sold by Wood Magazine. Although I felt their design was a good starting point, I definitely wanted a longer and deeper version of the couch (since couches are for naps, too, and I don't fit in a standard-sized couch) and I wanted to add a little of my own style - so I used their plans as a basic starting point and went from there. I ended up making the couch 1-1/2" deeper and 6" longer (notice these dimensions are multiples of 3/4"), abandoning some of their details, and adding a few of my own.
I won't be covering in-depth dimensions and details because I feel that wouldn't be fair to the folks at Wood Magazine who's plans I used as a foundation (so, buy those plans!), but I will be pointing out what I consider to be the weak areas in their design and the modifications I did to tailor the design more to my personal preferences. As always I'll include a few of the tricks, trials, and tribulations that go with a project like this.
Originally, I planned on making multiple pieces and selling some of them to pay for materials - buuuuut ...... I never got around to the selling part - lol. I ended up building two settles (couches), a loveseat, and a chair. The settles and chair are constructed from Quarter-sawn (QS) Red Oak, and the loveseat is Philippine Mahogany and Australian Lacewood. I'd also planned on using QS White Oak (the traditional material for these pieces) as I wanted to use the ammonia "fuming" process to create the final color, BUT finding Quarter-sawn White Oak in any appreciable quality/quantity has become a lot harder over the years - so I settled for QS Red Oak. Red Oak, BTW, does NOT turn a nice dark patina like White Oak does when fumed with ammonia - it turns a sickly green. Really weird.
Costs: Building quality furniture isn't exactly cheap - but then costs are relative. I spent about $4000 on materials alone for this entire project - BUT - I produced four pieces of heirloom quality furniture that will (barring an errant nuke or house fire) be around longer than I will. I didn't skimp on any of the materials - it's all top-shelf stuff - and unlike most modern furniture, these are completely rebuild-able. I would imagine that if you wanted to build something a little more budget-conscious, you could build a decent settle for around $650-$800 - but when you consider that these couches sell for upwards of $4000 - $7500 each, $4000 in materials for all of them (not counting labor) isn't all that bad.
Time: It's a bit difficult for me to estimate the time I spent because these were built over a period of around 8 months on and off. I started in the Fall and worked part way through Winter, and so I had to wait for decent weather to spray the final finish - so - a big part of that 8 months was spent with my living room filled with couch frames waiting for it to get above 70 Degrees F outside :) I think a reasonable estimate to build one of these would be roughly 60-80 hours for an intermediate woodworker.
Step 1: Stock Preparation
As with all projects that start with rough stock, the first step is stock preparation. This usually consists of jointing one side flat, then either ripping or jointing a square edge, followed by thickness planing and then ripping a parallel edge. Once your stock is clean, flat, square and parallel, you can get to grading and laying out your parts.
I ordered about 200 board-feet of skip-planed quarter-sawn (QS) Red Oak from a small, family-owned mill in southern Missouri and had it shipped to me. Surprisingly, it was significantly cheaper than buying it locally - so shopping around is worth your time. Of course, when you buy it this way, you lose the ability to pick and choose your boards - so be sure to tell the mill operators what you're going to be doing with it and let them know any minimum dimensions. I needed a number of rather wide boards for the main rails on the couches, so I was sure to let them know what I needed.
So, what does "quarter-sawn" mean? The term quarter-sawn refers to the section of the log from which the board was taken. QS lumber was cut from the center of the log, meaning that the growth rings are basically perpendicular to the board face. This generally yields a board of superior stability (they don't warp or "move" as much as rift-sawn or flat-sawn lumber) and grain character. Quarter-sawn boards are also the only boards where ray-flecking and other prominent grain features are almost guaranteed to show up - which is why they are highly prized (and priced) in woodworking. Most wood is flat-sawn (or plain-sawn) because it's just faster to produce - but even a flat-sawn log will have at least one quarter-sawn board come out of it. When picking out any kind of wood, I look for boards that have the growth rings oriented as perpendicular as possible to the faces of the board (but that's just me ;)
On a side note, I believe the Wikipedia page defining rift-sawn is *incorrect* - at least it goes against my years of experience of buying wood.
Step 2: Build the Legs
The legs are built in a unique way. To ensure that there is attractive grain displayed on all sides, the legs are constructed by mitering boards along their length and gluing them up around a square core. I found that gluing them up around the core was more trouble than it was worth, so, I glued the boards up, then glued the core in place. This allowed me to pull the mitered edges together more tightly. I used packing tape to hold the boards together while the glue set up, and set the legs vertically to keep the squeeze-out from running down the faces of the boards.
Once the cores were glued in, I planed the legs to final thickness. Be careful how you sneak up on the final dimension as you need to take the same amount off of each side of the leg to keep the corner joints centered. Once at the final thickness, I beveled the edges and then trimmed the legs to final length.
To cut the mortises, I built a custom sled for my plunge router. It's a simple design - just a lexan sheet with adjustable hardwood cleats. The cleats can be adjusted in or out to allow the router to follow the workpiece - you can even cut off-center slots if required. A good tip for making sure your bit is centered, is to take a scrap that is the same width as your workpiece, center the bit as best you can, make a shallow cut, then rotate the sled around 180 degrees on the piece and see if your bit is still centered in the slot.
I had something like 54 mortises to cut, so I came up with a very simple "jig" that consisted of a few blocks of wood screwed to my work bench (see images). Two blocks located the leg position, and then two stop-blocks were clamped in place to control the length of the cut. Working this way, it didn't take much time to precisely cut all 54 mortises.
Step 3: Building the Rails
Buying solid wood that's thick and wide enough to machine out a straight, thick, rail is an expensive proposition. Thicker wood costs more, and trying to flatten and dimension a very long, wide board is not within the capabilities of my tool set at this point - so - the next best thing is to build up the rails from thinner stock. This also has the advantage of allowing you to "straighten" a board that might not otherwise be a candidate for something like a rail. I chose to glue up two 3/4" pieces of stock to build the rails. Try to find boards that are bowed about the same so that their internal forces will cancel each other out, and glue them face to face with some kind of straight-edge as a guide. In my case, I found the perfect strong, flat, straight-edge in the fence rail on my table saw - it worked perfectly :) Once the glue was cured, the rails were cleaned up and machined to final dimensions.
Next came the process of cutting the arches and tenons. Since cutting the long rail tenons on the table saw was out due to their length and weight (dangerous and inaccurate), I built another quick jig using scraps that would allow me to cut the tenons with a router. Once the setup was dialed in, cutting the tenons was fast and accurate.
The tenons on the rails actually are mitered at 45 degrees and meet inside the legs. This mitering was done next on a miter saw. Once again, a very sharp blade makes a huge difference.
The next step in the process was to cut the arches in the bottoms of the rails - I wanted curves on all my rails, unlike the original plans. I laid out the curves for both the long and short rails on template material (old closet door skin, actually) using a piece of wood and clamps - bending the wood in to about 1" from the edge and marking a line. This was then cut on the bandsaw, and cleaned up with files until I had a nice smooth curve. This curve was then transferred to the rail stock and the curve was rough-cut on the bandsaw to within about 1/16" from the line. Next, the template was attached to the rail with carpet tape, and a pattern bit was used to clean the curve up. Final smoothing in any rough areas was handled by a block plane and files. Next, the rails were given 1/4" bevels on all edges - except the upper edge of the front lower rail. The front lower rail has a broad (1-3/4") bevel along the length - this was made with a hand plane since cutting it on the table saw would have been kind of awkward, and it actually went pretty fast since I'd just replaced the blades on my Rali planes (which are awesome hand planes, BTW).
The last step in this process was to cut the dados in the top and bottom edges of the rails which will accept the spindles in a later step. I don't have pics of this step, but they were cut on the table saw using a dado-stack blade.
Step 4: Assemble the Frames
At this point, it's time to get the frames put together.
Note: One thing I want to mention is that the way I fit the square-edged tenons into the round ended mortises was that I rasped the tenons to match the mortises (in case you were wondering). It was a lot less work than squaring out the mortises, and is just as strong.
It was nice that when the rails were clamped to the legs, the time spent laying out and cutting clean tenons paid off. When I applied the clamps, the frames squared themselves very nicely. The only check I made was to ensure that the top rail was flush with the tops of the legs, and that the legs were at the same "height". This is where careful cutting and layout pays off :)
I glued the frames up using polyurethane glue (Gorilla Glue). Polyurethane glue foams as it cures, and since the tenons of the rails actually meet in a miter inside the legs, I needed to clean out that excess squeeze out with a chisel. Once again, sharp tools made quick work of this process.
Now, it was time to put the frames together. The tenons on the rails, as well as the mortises, were given a coat of glue and everything was slotted into place. Doing this by yourself is a bit like herding cats, but it is possible. Be careful about using too much glue as the squeeze out can be a pain - so - what I suggest is to put painter's tape around the edges of your mortises and the edges of the "shoulders" of the tenons so that you can just peel the tape off and the squeeze out with it. As I expected, the frames were not quite square, so a band clamp around the "long" side allowed me to pull the frame back into square as the glue cured. Clamp everything up tight and be patient.
Once the glue is cured and your frames are solid, I drilled through the tenons using the "pinning holes" as guides and drove in some 1/4" x 1" hardwood dowels. These dowels were NOT flush (again departing from the plans) because I wanted decorative Ebony "pyramid" pins instead of flush sanded round pins. I used a drill-press mortising bit and a mallet to "square up" the holes to a depth of about a half inch.
The final step was to install the "stretchers" from the front rail to the back rail, and the "lip" to which the spring frames will attach. The stretchers support the lower frame as the elastic banding can exert quite a bit of force pulling them together....
Step 5: Making the Spindles
Nothing will make you feel like a manufacturer faster than making 300 of the same part - lol. Processes like this are also where jigs pay off dramatically.
The first part of this process was to cut the long rough stock. I ripped boards into roughly 1"x1"x8' pieces. These strips were then run through the thickness planer until they were the "perfect" match for the dados that were cut in the rails - i.e. I didn't machine them to a specific dimension, I machined them to fit the dados. Next, the stock was cut to length on a miter saw using a jig to ensure repeated accuracy. At this point, all of the spindles were sanded on all 4 sides - yes, it's a PITA, but it's far easier to do it while they're out and maneuverable than when they're installed. I didn't like the idea of super-sharp edges, so I wanted to put a 1/32" bevel on all of the edges to make them more pleasant to touch, and to make them harder to damage. The best way to do this is with a block plane - so I built a jig that allowed me to drop a spindle into it, take a couple of swipes (I had the plane set very shallow) flip it 90 degrees, cut the bevel, flip it, cut it, etc. The jig was simple and made what could have been a horribly tedious task a lot less frustrating (and faster).
I decided to "ebonize" my spindles (for the Oak pieces) because I like the combination of black and dark reds. An "old timer's" recipe for making an ebonizing solution is to take about two quarts of white vinegar and drop a pad of fine steel wool into it and let it set. Make sure you're not using dirty steel wool, or oiled steel wool. In a day or two, the steel wool will dissolve leaving you with a greenish juice. Go ahead and filter out any "floaters" before using the solution. Basically, what you've done is dissolve the iron into solution. This iron will react with the tanin in the wood to resulting in a black color. This happens in nature all the time - think of the black stains you see around nails in old fences or nails driven into trees. The higher the tanin content, the darker the black will be. Just wipe the wood with the solution and let the color develop (I used a cheap sponge).
In addition to making the spindles, at this point I machined the "filler block" stock that will be used to "fill in" the spaces between the spindles. How this is used will be explained in more detail in the spindle installation step :)
Step 6: Testing the Spring Frame
The elastic band "spring frames" were one of the big weaknesses in the original design. The frames are far too fragile if you want to make a *firm* seating surface (as I later found out). The original article had very little information on what to expect from the elastic spring material, how far to stretch it, etc, so I had to do some testing to find out what worked for me.
I used 2" Elasbelt Webbing from Rochford supply for my seats. You can buy the appropriate clips from them as well, or, you can choose to forgo the clips and just staple directly to the frame (but you should be sure to use a *furniture grade* pneumatic stapler if you do). Installing the clips is a little tedious, but I found the best process was to cut all my webbing to length, then using a small anvil and a 2lb sledge, mash the clips closed on the ends of the webbing. It's a surprisingly strong attachment. The clips have a "lip" that will catch in a saw kerf cut in the spring frame, allowing you to attach the webbing with a screw instead of staples. Having done it both ways, I will say it looks nicer, but isn't any stronger or faster than direct stapling (IMO).
When I first assembled a set of "springs" to the original specifications, I went ahead and installed them in the chair frame and tested it for sag by putting the foam in place and sitting on it. It was way too soft - I sank to the stretcher right away, and knew that I needed to modify the specifications of the spring assemblies. To achieve a firmness I liked, I shortened the length of the elastic, fit more strips closer together, and cross-wove pieces in (which required additional frame bracing on the main frames). You will need to do some testing to make sure you get a seating firmness you like. In later steps you will see how I ended up building the spring frames.
On a side note, do NOT be tempted to forgo the springs and just use a plywood panel as a base - you will produce a very uncomfortable seating surface.
Step 7: Corbels and Arm Rests
The "corbels" are the decorative "flying buttresses" that support the broad armrests on these pieces. Making them was pretty straight forward. First, a template was laid out on tempered hardboard (Masonite) and smoothed to final shape. This template was then used to both trace the piece onto rough stock, and used as a template for routing the final shape. By "nesting" the pieces during layout, I was able to reduce waste. I always make extra pieces in case I damage something inadvertently - and more often than not, I end up using the extras. The shape was rough-cut on the bandsaw, and finalized using the router table and a template. This piece is thin - which is dangerous on the router table - so I built a jig to keep my hands away from the bit. The rough piece was attached to the jig via screws through the back edge (which will later be slotted for biscuits) and it was nice having such a solid mount to work with. Less vibration yields cleaner cuts.
Once the pieces were routed to shape, the edges were sanded and then bevels were routed on the edges. Finally, a quick jig made from part of the cut-out scrap and some blocks was screwed to the bench, and biscuit slots were cut for a #20 and #10 biscuit.
The arm rests are mitered at the corners, joined with biscuits, and pinned with dowels at the outer "point" of the miter. Then, all the corners are given a 1/4" bevel on the router table.
Step 8: Installing the Corbels and Arm Rests
The corbels are installed with yellow glue and biscuits - a #20 at the top and a #10 at the bottom. This is a case where the yellow glue is good because the moisture in it will cause the biscuits to expand into the joints and form a tight fit.
A dowel hole is drilled at the outer edge of each corbel where it supports the arm rest. These dowel holes must be matched in the arm rest (it's a blind-dowel) - so having a set of dowel hole centers is extremely valuable at this step.
Now, while the face-grain to face-grain glue-bond between the top rails and the arm rests is a strong one, I prefer to add a little mechanical connection as well. I did this by drilling counter-sunk holes in the dados of the upper rails and running screws into the arm rests from underneath. If you choose to do this, be extremely careful about sizing so your screws don't punch through, or "telegraph" and put a bump on the surface of the arm rest.
Step 9: Applying the Finish
The finishing process was kind of organic as I was adjusting colors as I went, but it went something like this:
- I started with a custom mix of NGR (Non Grain-Raising) dye that was mopped on with a soaked rag.
- The dye was followed by a wipe-down of denatured alcohol to even out the color
- Once the dye dried, I followed with two rub-coats of dark Watco oil with a bit of additional tint to it (don't remember the exact mixture). This coat was allowed to cure for a few days.
- The surface was lightly sanded with 220 grit sandpaper and wiped down with microfiber dusting cloth (which I like better than tack cloth).
- The top coat, as mentioned, was sprayed-on Satin Super Kenvar catalyzed varnish from Sherwin Williams - three solid coats. Because I had fit the spindles so precisely to the dados, I set "masking blocks" inside the dados to keep finish from building up inside them. This also made sure that the glue used to glue the filler blocks in place would have a wooden surface to bond to instead of trying to bond to the slick Super Kenvar finish.
A little trick: When spraying something this big, you have to move and you have to move fast to keep your edges wet. Sometimes, you get runs in your finish - it happens. Instead of sanding these runs out - which is very hard to control, here's a trick I learned painting cars: Take a utility knife blade, or alternately, a single-edge razor blade. Clamp the blade in a vice and use either a burnishing rod (like you use to form a burr on a cabinet scraper) or the shaft of a screwdriver, hold the rod at about 45 degrees from perpendicular to the blade edge and gently run it along the edge of the blade a few times. What you are doing is "rolling the edge" over so that it is almost perpendicular to the blade. You can then take this blade and while holding it at about 45 degrees to the surface of the wood, gently drag it over the runs. The blade will shave very fine layers off of the run and will allow you to have extremely fine control of what you remove and what you leave behind. What you are essentially doing is making a tiny cabinet scraper to shave away the run and only the run. Once you get a feel for this technique, you'll use it all the time - guaranteed ;)
I chose to pre-finish the spindles because frankly, trying to spray 300 spindles spaced 3/4" apart and get good coverage on all four sides without runs sounds like a formula for disaster (if not insanity). It took me a few tries to figure out the best way, and it turned out that once again, a little time making a jig paid off in a huge way. At first I tried laying out the spindles flat and shooting one side at a time. That plan went out the window when Mac, my dog, decided right after I'd completed spraying a coat of finish, that it would be great fun to grab hold of the drop cloth on the table and try to do the "rip the table cloth out from under the dishes and leave them all on the table" trick..... buuuut Mac is a dog and sucks at magic tricks .... so he ended up pulling over the entire table and dumping all the spindles onto the sandy concrete and dirty grass.... where the Super-Hard, Super-Kenvar quickly cured. I spent the next two days sanding crap out of the finish of 150+ spindles ..... Mac was put into protective custody.
Anyway, once that fiasco blew over, I figured out a way to spray two sides at at time, and this worked out MUCH better anyway - so in a way, it ended up that Mac did me a favor. The Universe works in mysterious ways.....
I also pre-finished the "filler block" material to match the frames at this point, after which they were cut to length as spacers between the spindles.
Step 10: Installing the Spindles
Installing the spindles is pretty quick - it kind of reminded me of weaving for some reason. Anyway, the spindles are tilted into the dados and stood up (with a bit of effort) and the filler blocks are set between them in the top and bottom rails. I placed all of the filler blocks on their sides at first because I wanted to get the spindles and filler blocks in place, then measure the slop at both ends so I could custom-cut filler blocks for the ends that would center all of the spindles in the frame. Once I had the end blocks figured out, I glued in the filler blocks. Again, I did my own thing with the filler blocks: first, I slightly tapered them so that they would "wedge" into the dados tightly and hold themselves in place while the glue dried (maybe 2 degrees of draft), secondly, I made them slightly taller than the depth of the dado so that they would sit up proud and form an additional shadow line. I rounded the top edges of the stock pieces to give this a more finished look.
The one "trick" here is the installation of the last spindle in the row. If you try to just install it as you have the others, you won't get them into the frame. The way to do it is to install the last 5 spindles into the dado without spacers, slide them toward the inside, tilt in the last spindle, then slide the 5 spindles back apart and install their spacers.
Step 11: Craaaack! a Minor Disaster
I found out the hard way that the frame specifications in the original design weren't nearly strong enough for the kind of seat firmness that I was after. When I first got all the elastic in place on the frame, it was already starting to warp alarmingly, but I figured that I'd just keep going (after all, these were plans I'd paid for and other people had used successfully, right?). As I was pulling the fabric up around the frame to staple it in place, the "support rods" that held the frame apart suddenly started to make cracking sounds so I stepped back just in time to have all three of them shoot out like arrows and the frame implode. Ouch.
I think that the design is adequate if a person stays with the original elastic webbing layout and specifications, but for anything firmer, you'll want to build a better frame.
The new frame I made was substantially beefier and made from Hard Maple instead of Oak. It still warped to an extent under the force of the elastic, but it was substantially more stable.
Step 12: Weaving the Elastic Springs
- Clamp your frame securely to a heavy bench
- Install one side of the strap with a screw.
- While wearing heavy gloves - Grab the opposite end's clip with a pair of pliers or channel lock
- Get a quick-action clamp ready to hold the clip down
- Pull the end across the frame, tilting the clip up so you can hook the lip of the clip into the kerf in the frame
- Quickly clamp the clip down so that it can't flip out of the kerf.
- Drill a pilot hole and install the mounting screw.
Step 13: Upholstery and Installing the Seat Cushions
One nice thing about these designs is that the upholstery is pretty simple overall. The seat cushion is basically an open-ended box that gets stapled to a wooden frame. I went with soft-backed cushions as opposed to the hard-backed (but easier to build) cushions that the original plans called for. I think if I was to do it over again, I'd use a more loose-fill style of back cushion - something that could be squashed around to accommodate more personal seating preferences. I suppose I can still do that :)
I used a very high quality medium density urethane seat foam. If you buy cheap foam, in a very short time, it will lose it's ability to support weight and will feel crappy and cheap. This is one place where I'd highly advise against using inferior materials - the foam is important.
The fabric I used was genuine Toray Ultrasuede shipped out of Japan. Not cheap, but it's amazing fabric. One measure of durability of fabric is a thing called the Wysenbeek Double-Rub rating. Heavy demin rates around 50K, Top-Quality upholstery fabric rates at around 100K... and Ultrasuede rates at over 220K ..... like I said, I like durable furniture :)
I sewed the upholstery on a cheap Kenmore sewing machine. I think it's far more important to have the right needles and thread than it is a fancy machine. Be sure to use a smooth, heavy, nylon upholstery thread and a needle rated for that heavy thread, and sewing heavy fabric goes pretty smoothly.
To install the seat cushions, I clamped the cushion frames into the main frame and drove screws through the mounting strips and into the elastic frame. The clamps make sure that the frame is lying flat in it's mounts when the screws are driven home.
Step 14: Some Final Shots
I'm really happy with the way these turned out - they pretty much met or exceeded all of my expectations. Now, to build the rest of the living room furniture ..... the projects never end....
Thanks for your time. I hope you found this interesting if not educational - and should you decide to build your own I hope this instructable will help you avoid some of the pitfalls ;)