In the previous post I mentioned I was going to discuss the rabbet detail at the edge of the doors where they meet. To create the overlapping rabbeted edges without adding any stock and making it appear as an afterthought, the rabbeted edge profiles are formed from the door stiles themselves. Since the inner door stiles have been designed to be, in this case 1 1/2 inches wide, rabbeting one or the other would reduce this width of stile by the width of the rabbet and in the process cause the door stiles to look a bit off. I also want the right door to overlap the left door with the assumption that the right hand door is typically the first one to be opened. The techniques I use is to create the left door middle stile at 1 3/4 inches and keep the right door middle stile at 1 1/2 inches. The overlapping, complementary rabbets, once created, will leave both stiles at 1 1/2 inches width. Design dilemma solved!
The primary reason to have these overlapping rabbets is both a form and function issue. Wood doors tend to expand and contract with seasonal change, although much less with this frame and panel design, but nonetheless there is a small gap that narrows and widens where the doors meet. The overlapping rabbet handles this very well, providing wood behind the small gap resulting in no glaring gap between the doors and the doors interlock. These two features justify this extra step. In the photo above the rabbet detail can be seen and the door stile widths are once again of the same width, or at least the visible parts are.
I've gone ahead and rabbeted the door panels to fit the groove surrounding the inside of the individual door frames. I perform the bulk of this rabbeting on a router table and make the last passes with a small shoulder plane to have the door panel fit accurately within the groove. At this junction, I have the tiger maple door panels rabbeted and the door frame components prepared. There are alternative methods of joining the stiles to the rails such as dowels, a slip joint, or a half lap joint. In these cases, there would be a need for stopped grooves in either the rails and stiles or in both. I decided to use the traditional mortise and tenon with haunched tenon to join the rails and stiles, this allows me to have the grooves running the full length of the frame members.The panels were purposely left a fraction of an inch proud of the surface of the rails and stiles when I prepared them earlier. As a sanity check, I intend to fit each panel in its frame as a dry fit and then determine how much more I need to reduce the thickness of the raised part of the panel. This is an extra step, possibly unnecessary, but it leaves me with peace of mind that the panels are exactly flush with the door frame members.
Someone earlier asked to have more information on the rabbet detail where the doors meet, something I referred to in an earlier post. I'll describe it in my next post and also post a photo of the detail. In the photo above, I have the doors glued up with panels in place. Part of the design was to have as small a reveal as possible between the panels and the door frame rails and stiles, this to give the effect of a thick black shadow line surrounding the very light maple panels. I'm very satisfied with the result and from a few feet away this effect is obvious. It was necessary to get the reveal just right and spaced correctly all around. I'm going to spend some time now trimming the doors as I had left some corners with a fractional overhang as a safety margin.The combined width of the doors is wider than the sides of the armoire and I'll explain the logic behind this in my next post along with a paragraph about "dynamic design". The armoire is coming along nicely and beginning to take shape.If you look out the window just behind the workbench you can see snow. Yes, we had an early winter snow storm up here this past day. Nonetheless, it was nice and toasty in the studio today, something about a pristine snowy setting that makes being in the studio that much more enjoyable.
I'm preparing the tiger maple panels prior to assembly of the door frames. This type of highly figured wood can be quite a challenge to handplane to a mirror finish. The wood can easily tearout if the angle of attack of the plane iron is not quite right or if the blade is not sharp. There is also technique involved. In the photo I'm using a high angled handplane which more or less falls in between a handplane and a scraper. I choose this tool to eliminate any possibility of tearout with a degree of certainty. After this operation of smoothing the board, I use a scraper to burnish the surface and eliminate any small streaks from the handplane operation. You can see the degree of curl in the tiger maple, since the curls are fairly tight, the wood borders on fiddleback maple quality. The curls are not very pronounced at this point except in the correct light, but once a clear, deep finish is applied, the curls will pop and appear quite dramatic. I'm taking my time with this step to prevent any tearout from occurring as the two panels , although replaceable, are fairly important at this stage.My next step after preparing the surfaces of the panels will be to rabbet the outside edges of each panel. The rabbet will match a groove in both the rail and stiles of each door panel. The reveal around the edge of each tiger maple panel and between the panel and door rail or stile needs to be uniform all around, a bit of a time consuming step to get just right.I'm going to place the panels against the individual door frames, to have the door frame overlay the tiger maple panel, as this will allow me to adjust the graphics of the panels within the door frames for optimal effect. Once I've selected a nice pattern to capture, I'll scribe the outline on the panel and create the edging with this outline as a reference. Hopefully, in my next installment I'll have the door frames completely assembled and glued up.
After discussing with the client, they agreed to give me artistic freedom on the door design. I spent some time working on the frame components of the individual door frames , and I've progressed as far as fitting the door frames into the carcase for test fitting. The door frame components, rails and stiles, are joined with conventional mortise and tenon construction. My natural inclination is to have the stiles on a door frame extend top to bottom of the door and I usually don't think twice about this. Borrowing a page from James Krenov and one of his older designs, I decided to instead extend the top and bottom rails the full width of the door frames instead. The lines become horizontal now rather than vertical. I'm also going to set the vertical stiles back a fraction to provide a shadow effect and give the appearance of horizontal lines running continuously across the top and bottom of the doors. It is preferable to make a mock-up of the door frame configuration but I did the next best thing and drew the new design and applied it directly to the carcase of jewelry armoire as in the previous photo. I'm taking a blind leap of faith at this point and basing the success of this modified design on a photo of an existing, but different cabinet along with my modified drawing.
At this point, I like the lines of the door frames and panel area and have decided to proceed and prepare the figured maple raised panels. I need to be judicious with this step as the reveal around the raised panel and the door frame needs to be uniform on four sides of the panel. I'm also implementing a cool rabbet between the door frames to have them overlap and not reveal a gap when the door frames expand and contract with seasonal change in humidity and temperature. A photo of the figured maple door panels will be posted next, after I rabbet the edges and begin to fit them into the door frames. I'm kind of excited at this point and am looking forward to what the completed doors on the carcase look like. It's easy to rush things when anxiety sets in, and patience is a great virtue during some of these delicate and accurate fittings. Steps need to be followed in the correct order before the doors are glued together. Early on when the sketching of this armoire began, one of the proposals was to have one piece veneered doors. A good design alternative, it allows the full width of the individual doors to feature a nicely figured veneer without frame components taking away from the space. Another small advantage is the increased dimensional stability a multiple ply substrate with veneered surfaces provides. I'll be exploring this technique in future designs.
In the previous post I mentioned I was going to discuss the rabbet detail at the edge of the doors where they meet. To create the overlapping rabbeted edges without adding any stock and making it appear as an afterthought, the rabbeted edge profiles are formed from the door stiles themselves. Since the inner door stiles have been designed to be, in this case 1 1/2 inches wide, rabbeting one or the other would reduce this width of stile by the width of the rabbet and in the process cause the door stiles to look a bit off. I also want the right door to overlap the left door with the assumption that the right hand door is typically the first one to be opened. 
