Andrew Pitts - Furniture Maker, Work in Progress, Cabinet sans Sides

Click on images to enlarge and see details

I've had this idea in the back of my mind for some time to build a cabinet that floats inside a shaped stand. It's not a new idea -- there are plenty of pieces with a cabinet on or inside a stand -- but it shares on the theme of shaped legs that I started with my Chest sans Sides and continued through two more pieces, the Special Cabinet and the Filing Cabinet. I was also inspired by the prospectus of an exhibition planned for January 2009 at the Studio Gallery in Kilmarnock, Virginia. The show is titled "Black, White and Shades of Gray", and I thought this would be a great opportunity to experiment with ebonizing wood (making common woods appear as ebony wood). I spent a good deal of time sketching and then working the on the computer using my Computer Aided Design (CAD) program to get the design right. I am a big proponent of using CAD, despite the steep learning curve. CAD enables me to test a design before I build, and where sketching provides a good front/side/top view, CAD enables me to make accurate renderings of the entire piece and view it from any angle I desire. It also enables me to change dimensions, make parts larger or smaller, change the angles between parts, and so forth. The rendering above is actually not exactly what I am building! I've entered the CAD program mid stream and have reduced the front to back dimension, the depth, by 2 inches! It might be interesting to see what other changes I decide to make during the building process. In this design I have curved legs, although I'm not certain I want to leave the right legs so flared at the bottoms, and I will ebonize them. The stretchers are planned to be a lighter shade of gray, and the cabinet itself is a mix of light and dark and medium woods. I want the sides to be holly (white), but the top and bottom walnut (dark), while the doors and back are cherry (medium). I've not yet decided on door handles - I usually wait till late in the game to make that decision, acting as the mood moves me. OK, let's get building!

The setup for this cabinet was extensive. Unlike square cabinets, such as my recent Pedestal Cabinet, curved cabinets are not made by simply sawing up some stock and joining the pieces. The curved sides have to be bent laminations, and these have to be made on prepared forms. Well, to say they have to be bent

laminations is probably not quite accurate. The thing is, if I simply sawed the left side, which is the really curved side, from a thick piece of stock, or more likely a glued up blank, I would get a resulting grain pattern that would not appear as the grain on a flat piece of wood -- the grain would get a bit funky. Also, of lesser consideration there would be a waste of wood as I cut away the parts I did not need. Of more consideration would be that I would be bandsawing the curve from a 12 inch wide board, a fairly difficult task to do accurately. Also, a bent lamination would be more

stable with environmental changes (humidity makes wood move, but to a lesser extent with glued laminations). So, that was my reasoning. Well, making the forms was quite a job in itself. The forms were cut

from 1/2" medium density fiberboard (MDF) I had on hand, which is essentially cardboard fiber pressed and glued into a very stable, but somewhat weak sheet. For me, MDF is used for my forms, some jigs, and templates. I first printed out a side profile of the cabinet side, in full scale, taping the sheets of paper together and then using carbon paper to transfer the profile to the MDF, as shown at right. With that single layer of MDF (of the 24 I finally glued up into a block) in hand I determined the size of each of the 24 strips and cut them on the table saw, as shown at left. Notice that I used a saw blade guard and push stick; each of these strips was 2-3/4" wide and I did not want my

fingers near the blade, even though I was using the safest table saw in history, the SawStop! So, with the 24 strips cut, I glued them up into four packages of 6 strips each and then bandsawed

each package to the profile of the curved side. The photo at right shows me using the bandsaw for this. Now, it's nearly impossible to get all four packages cut exactly the same way, although I can make it pretty close! So I next used my edge sander to fair the face of each package, as shown at left, to get the pieces as close to each other as possible. Then, when I glued up the four packages into one 24 layer block 12" wide, as seen at right, I would only have to do a bit of touch up to make a perfectly shaped form for laminating the sides.

Next, I prepared the stock for the laminated sides. I wanted the sides to be as white a wood as I had, and holly fit the bill. But, I didn't want to use up my precious supply of holly for the inside layers of the laminated sides, so I chose poplar for the filler layers, using thin strips of holly on the edges. The sides were built up from five layers of wood, each 1/8" thick for 5/8"

thick sides. Using my bandsaw, seen at right, I first resawed the stock to a thickness of about 5/32" - 3/16" so that I could sand out the saw marks and end up with 1/8" thick laminates. With all the stock sawn I used my horizontal drum sander to thickness the pieces and remove all the saw marks, as seen at left. Once all the laminate pieces were prepared, I did a test clampup of the assembly using my vacuum bag clamp, just to make sure the vacuum bag would have enough umphh to bend all the layers. Satisfied with that, I set up for the actual glueup. Now, we are talking about gluing five layers around a form and clamping the whole thing inside a big plastic bag from which most the air is withdrawn causing the atmospheric pressure to clamp down on the bag which clamps down on the work. Referring to the photo at right, the bag is very heavy polyethylene, and my vacuum

source is a venturi pump using my air compressor as a source of energy. A venturi pump, sometimes called a jet pump, works on the principal that a fluid accelerated through a nozzle will experience a pressure reduction as the velocity increases. My air compressor supplies 100 psig air to the nozzle and develops 25" Hg of vacuum at the nozzle. That creates about 12 psig of clamping pressure inside the bag. To put that in perspective, the area of my lamination was about 26" x 12", or 312 square inches. Doing the math, 12 psig x 312 sq in = 3744 pound of force on top of the lamination (plus the same force pushing up on the bottom), so this is a pretty efficient clamp, and it can conform exactly to any form! But, using a vacuum clamp requires a lot of advance planning. Since the glue has a short open time, that is, once spread on the wood clamping must take place within minutes, and since there were five layers to spread glue on evenly, I had to have everything ready before I mixed my glue. As always, I use Unibond 800 for my laminations as this glue sets up very firmly, even brittle, and will not "creep", or plastically deform over time. That keeps a bent lamination bent for life. Additionally, I had to place the form with the laminates on top of a platten that would permit the air to be evacuated while keeping the assembly flat. Finally, I had to affix the layers such that they would not slide around in the bag while pressure was being applied. So .... I used veneer tape to hold all the pieces of laminate together, especially the bookmatched pieces of the face and back of the panel. Then I mixed and applied the glue with a sponge applicator, and finally I laid the glued up assembly on the form and temporarily taped it to the form, bent, so the bag would not try to get between the lamination and the form as air was evacuated. I worked fast, and successfully glued up both sides (one at a time), allowing overnight for the glue to completely cure. When released from the forms, there was only a slight bit of "springback" in the bent laminations.

Having successfully laminated the sides (whew!), I set about to make the legs. As I am prone to do, I designed these legs so that they were pretty near impossible to build (just kidding ..... sort of). They curve in two directions, and are of varying thickness along their lengths. In addition, the stretchers would have to be mortise & tenon joined to these curved legs. Now, there are probably a number of ways to cut mortises into legs like these, and in fact if this was not a prototype piece I would probably have enough experience and data to be able to cut the mortises before cutting the curve into the legs. That is, I could cut large blanks for the legs and calculate the correct angles of attachment for the stretchers and then cut the mortises on relatively rectilinear leg blanks. But, I had no experience with this piece, so my only good alternative was to glue up the leg blanks, cut the curved legs from the

blanks, then fit up the stretchers dry and measure all the angles, then use my Domino tenon machine to cut the mortises. I started by selecting white oak to be laminated into the blanks. I chose white oak because I planned to stain it black, called ebonizing, and the even color of white oak would work good for that. If I had a stock of ash, I might have used that, but I didn't have significant quantities of ash on hand. Using my computer aided design (CAD) program,

I figured the optimum size of each piece of oak to be laminated so that I

would save wood. As it was, there was going to be a lot of waste and I wanted to minimize that. After cutting all the pieces and thicknessing them, I glued them up using Titebond Original, just plain yellow glue, and clamped them with a LOT OF CLAMPS! The photos at right show this. Once the blanks were glued up, I

bandsawed each leg to approximate shape, as shown to the left, then using spokeshave and edge sander I shaped each leg to the finished dimensions, as shown at right. This was a long process! With the legs pretty much to size, I laid them up with

scraps of wood to

hold them in place so I could measure the angles for the stretchers, as shown to the left. This was the first time I could really see how the legs, with all their curves, would interact. It took a while, but I measured all the angles for the stretchers, then carefully cut the stretchers to fit and dry clamped them in place, as shown at right.

The stretchers would be connected to the legs with Domino tenons,

which are floating tenons glued into mortises made by the Festool Domino machine. The photo at left shows the machine at work, and the photo at right shows a pair of finished mortises for one end of one stretcher. It took making some guide jigs to hold the machine in exact place, but the results were worth all the effort, as you can see in the dry assembled stand shown at right. Piece of cake (NOT!).

Now, with the stand assembled I put my attention back on the carcase. As you recall, I bent laminated the sides of holly with poplar core pieces. Next, I made the top and bottom of the carcase, using walnut with the sapwood intact. Walnut sapwood is very light, so the contrast of the dark walnut and the light sapwood would be striking, especially if I dyed the walnut a bit darker than natural. After cutting the stock to size and thicknessing it, I drilled dowel holes for the sides to attach and cut mortise slots for the back. At this time, I also cut the stock for the back

(maple) and doors (poplar) so that I could do fitup all at once. I was working under a deadline (I hate deadlines), so I missed photographing some steps,

but the process for this phase was exactly like the process employed in my Pedestal Cabinet, and you can take a peek at the Work in Progress writeup for the Pedestal Cabinet to see exactly how I assembled the top, bottom,

sides, and back. I also took this opportunity to cut the interior shelf and drawer pocket, assembling them with dowels, as well. The photo to the left is one of the few I took at this point and shows me drilling dowel holes in the drawer pocket assembly. With the dowel holes cut and the mortises cut, I could dry assemble the cabinet and check fitup, as shown in the photos to right. Knowing everything would fit, I was ready to stain, finish, and glue the pieces.

It is significant to say, at this point, that I almost decided to not ebonize the legs and do any staining. I really don't like changing the natural color of wood, which is beautiful as it is, but this chest was being made for a specific purpose, to show in the "Black, White, and Shades of Gray" exhibition at the Studio Gallery in Kilmarnock, Virginia. I stopped and asked my wife to come up to the shop for a consultation, and she convinced me that the black on white would be striking, so it is she who can be thanked for the final success of the piece! I purchased some aniline dye, which is water based, and experimented with ebonizing. I found that the black dye, by itself, made my wood have a slight bluish and iridescent cast, and I found that by applying a cover coat of extra dark walnut dye I could correct that. Since I was dying wood, I allowed the natural color of the wood to remain just a bit so that the stand ended up being just a shade lighter than black and with a slight brownish cast. Using aniline dye is a process that takes some practice. First, the water will raise the grain of the wood, so before dying I had to sand the wood to the final grit, in this case P400, then apply water and let the wood dry. This raised the grain, so I had to resand with P320 and P400 to knock down all the fibers. That way, when I dyed the wood there would be no more raising of the grain. Also, applying multiple coats of aniline dye does little to affect the color, but applying a coat of a different color dye dissolves the original dye and combines the two colors, as would be the case if I had mixed two colors before application. That made it so I could adjust the color of the wood as I went along, a very handy tool, indeed! With the pieces of the stand dyed, I applied multiple coats of shellac polish, which is simply very thin (1/2 pound cut) shellac I mix myself and apply with a pad, sanding between coats until I like the feel. Then I glued up the stand, which was quite a feat in itself as there are no right angles to clamp onto! The carcase was finished without dye, except for the dark part of the walnut and the edges of the back pieces, and then was glued up. If you look carefully at the photo of the finished cabinet, below, you can see that the back was made with two 'slats' of maple so that it could expand and shrink with the seasons.

The doors were cut exactly the same size as the back slats, and I mortised the hinges after I cut them, but they were to undergo an additional process. I carved a scene of reeds with a

cattail and a dragonfly, and dyed the carvings black. To do this, I drew the scene on a full size piece of paper so that I could use carbon paper to transfer the design to the wood, seen at right and then carve it. It didn't take too much artistic talent to draw the reeds, and I doubt I would attempt to draw anything more complicated, so this worked out well for me. The carving went smoothly, as poplar carves

quite easily, and before applying the black dye I lightly sealed the carving with a coat of shellac polish so the water based dye would not wick up into the fibers and muddy the look. I wanted a crisp transition between the black and the natural wood and was fortunate to endure only a slight bit of wicking around the dragonfly wings. That, actually, was not such a bad thing, as a dragonfly wing has a very fine texture that the wicking sort of simulated -- that's my story and I'm sticking to it! The photo to left shows me applying the dye with a brush. I also dyed the edges of the doors. Once the dye dried, I applied several coats of shellac polish as a sealer. Also, and I did not get any photos of it, I constructed a hand dovetailed drawer to fit into the drawer pocket. You can see the drawer with its walnut face in the 'open' photo, below. And that's it! Click on the two finished photos to see larger versions.

Individually Styled and Crafted

Fine Hardwood Furniture

by

Andrew Pitts ~ FurnitureMaker