I had a pretty good day in the shop yesterday and was able to glue up the door. If you have followed my previous escapades you’ll know that I bolloxed the first attempt. You’ll also know the achieved plausible deniability and was able to blame it my mortise chisel.
Yesterday I was able to chop four nearly perfect mortises and move on to cutting the weird staggered tenons. As a footnote, I’m still not happy with the Sorby mortise chisel, I had to sharpen it twice while cutting the four mortises. I probably could have pushed it through all four mortises, but the edge was knocked and rolling over and I was afraid it would push it off course or cause a split. I’ll need a better solution before I mortise more.
This is the layout I came up with for the mortise and tenon construction on the door. the rebate in the finished door is .300, which is just over half the height of the zinc channel that will hold the stained glass panel. The mortise is 1″ deep (slightly over that, actually) to accommodate a 1″ tenon – but because of the rebate the front shoulders are .300″ longer than the rear. Hopefully the picture below make this more clear than the my description…
I was concerned about how to get the rails cut to the correct length so that the door would be the correct width. My experience tells me that measuring the stiles and subtracting them from the door opening, then adding in the planned lengths for the rebate and mortise leads to tolerance stack-up and it’s not exact enough.
Instead I laid the two stiles in the door opening, and knifed in the remaining width on a scrap of fine. I cut this to length so it was just slightly oversize, and checked it at both the top and bottom of the door opening. This gives me the target width for the exposed face of the rail. Then I added in the mortise and rebate widths, times two, to get the overall length. This gave me the final cut length for the rails.
With these two story sticks I could lay out the final width, the shoulders for the front of the tenons and then measure the offset for the rear shoulders…and double check everything before cutting. In the end I’m shooting for a door that is just slightly too large, maybe .020″ to .030″, so I can plane it down and get the gaps exactly right. When I’m a little more confident in making inset doors I’ll probably shoot for an exact fit with a shaving or two to dial it in.
In the end I was successful. The tenons were slightly over thick, and a shaving or two off the back (to keep the alignment of the fronts on track gave me a nice snug fit that took a light touch with a mallet to seat (and take apart). The front and back faces were flush, the joints square and not exposed, and the assembled door lays flat on the bench with no twist. Joy in Mudville.
Since gluing up with Old Brown Glue is something more of a process than just grabbing the glue bottle (I heat up a glass of water to warm the glue) I try to do all of my gluing at once. I added walnut Miller Dowels to the case sides to reinforce the rabbet joints, and I also cut thin strips to fill where my earlier mistake led a gap at the corners. I made the rabbet at the back of the case deeper than the rabbet at the corners, which left about a 1/8″ gap. The repair should be invisible and certainly won’t cause any structural problems.