After some fiddling and adjusting, I created 11 good wall sections. The first one turned into a test piece, since I got the Y and Z axis alignments wrong. So I had to cut another section from the board, the lower one in the second photo. Then it was just a matter of standing with the dust system hose in place to reduce the chip mess. Each section took about 15 minutes.
Tomorrow I’ll do the tenon side. I expect there’s be a bit of adjustment on the first one of those as well.
Yesterday the 2 long half-inch cutter bits arrived. Today I worked on the machining fixture. I had already added a reinforcing platform to the front. Today I machined it to height to accommodate the 16.3636 degree slope required. Then I carefully aligned the fixture on the CNC spoil board so that the X axis is as exactly inline as possible. The compression clamp on the right will ensure that the left of the workpiece is properly registered. I run 2 screws in from the left to be sure.
Something was off. I did manage to get the ring sections joined with the help of a strap clam and a big rubber mallet. But they didn’t want to close up, so I left everything clamped to the table and called it a day. It was New Year’s Eve so to time to go “celebrate”. I opened what I believe to be a 20 year old bottle of Mumm’s Champagne. It wasn’t very good and the remaining half will be dumped.
Today I looked carefully at the parts and the jig. Some sections had broken off the parts. As far as I can tell, the X axis of the fixture was slightly off parallel to that of the machine. It’s also likely that my XY0 was a millimetre or so to the right of where it should have been for a symmetrical cut. I added a reference board to the fixture which I will machine to a height to use for alignment in all 3 dimensions.
The other issue was that since I hadn’t allowed for the joint length, the wall sections are shorter than I had planned. So I cleaned up the widest 8/4 board I have and cut 11 new sections. This board is about 190 mm wide, so by the time the sections are cut they should be near the 170 mm that I want. That will get me back to the 550 mm or so finished diameter. I also added a bit more relief to the mortise pattern to make joining the pieces easier.
While I was doing the last few cuts I thought the bit was sounding slightly strained. It’s at least 10 years old and has seen quite a bit of use. So I ordered 2 new ones at Amazon. One’s an Amana 2″ and the other a Freud 2.5″. They won’t be here for at least a week, making this an even more long-term project.
Each section takes about a half hour on the CNC. Today I had a shorter day, being Wednesday with noon commitments. I stopped at 15:30 when 6 were done. I’ll do the other 5 tomorrow and do a a test assembly. So far nothing broke and there were no injuries.
Every pass gave additional feedback for further adjustments. First it was clearance, then positioning, then depth of cut. The -0.375 mm clearance works well. The tenons still protrude because the flattening cut is of course at half the included angle. I decided just to skip the flattening pass and trim the extra with the band saw or plane before sanding. The other consequence of the adjustments is that the usable section length is shorter than my original plan. That just means that the fountain diameter will be less than planned by about 100 mm.
4 test cuts later, and I have a joint that works. I started with a -0.125 mm clearance on the mortice, moved to -0.250, then tried -0.5 mm. The last works, but is now a bit loose. I’ll try again tomorrow with -0.375.
I also completely redesigned the joint. The other one was adapted from my Raven Desk joints, which were cut with a 1/4″ cutter. This time I have to use a half inch cutter because of the depth of cut. This latest joint is based on circles with a 15 mm diameter, assuring lots of clearance for the tool.
One of the cuts dislodged the work piece in the fixture and I hit the Stop button as fast as I could. After that I ran screws into the workpiece from both sides, in addition to the Bessey clamps. The screw hole areas will probably be cut out anyway, both top and bottom.
I just realised that if I cut from the outer side instead, I can do a half-blind mortice and not have that gap at the joint., OK, back to “the drawing board” ie. VCarve tomorrow to try that approach.
First test of the fixture is a partial success. Nothing broke. No-one was injured. The joints don’t mate. I left the -0.125 mm allowance that has worked well in the past. I’m thinking I didn’t rotate or reflect the joint design correctly. Since this is my first asymmetrical joint, that’s probably the reason. The other good news is that the angle is exactly correct, within measurement error.
Man, it sure was nerve-wracking though. I don’t like long bits or deep cuts. I reduced the feed rate to 40 IPM with a 6 mm pass depth, so it was fine. But still, had the heart rate elevated.
After thinking about this for several days and partial nights, I realized that adding fancy joints to the perimeter wall would add a lot of interest. The problem is that the joints will have to be at an odd angle, that angle dependent on the number of sides. I decided not at all arbitrarily to make 11 sides. I’ll leave the reason “as an exercise for the student”. In any case, I laid out a design for a machining fixture in VCarve and made it out of 18mm Baltic Birch. Now I just need clamps to hold down the work piece. Oh, and to design the joint file. Unfortunately Tailmaker’s software will do 90 degree or flat joints, not odd angled ones. Hence the fixture and a custom design.
I started with a composite image courtesy of the NASA Lunar Reconnaissance Orbiter, used a Fusion 360 plug-in to turn it into an STL file, then imported the STL into Vectric VCarve Pro. From there I was able to generate the tool paths for the CNC. The test piece is made of MDF. The white cylinders are a place-holder for the 7 that will eventually form a fountain. With the recent confirmation of water on the Moon, I think this is a timely project. I’ve since found a better composite image from the NASA Deep Space Program Science Experiment (Clementine).
Recently I took delivery of $900 worth of Easter White Cedar (stacked in the background) and 8/4 Hard Maple. The cedar will make 4 Kitchissippi chairs. I estimate I’ll need 3 of the Maple boards for the LSP. The remainder are on the floor in the left background. Some of them will make the base for the “Nahrung” sculpture. Now if I could figure out a way of incorporating any of Tailmaker’s software, the project would be complete. Maybe a jointed base?