Expanding Round Table

Construction Blog: Page 15
Wrapping up Loose Ends on the Wedges and Arrows

As I mentioned in a previous page, I have been too busy to take pictures and write up blog pages. As a result, there have been a bunch of tasks that didn’t get documented. I’ll try to fill in some of the gaps now. I have tried to organize this page to be generally related to the final steps in creating the main top wedge pieces.

After I veneered the wedge pieces, I needed to mortise the slots for the pins and sockets. Because this required clamping a jig to the surface, it had to be completed before I could install the skirt boards.

I clamped this jig to the edge of each piece and plunged out the mortises. I mortised out all of the right hand sides first, and then reset the jig and router depth for mortising the left hand sides.

The jig was made from some scraps and contains two mortise template cutouts--one toward the outside and one toward the center of the table. The jig clamps to the top surface of the workpiece to ensure that all mortise slots are positioned at exactly the same height. It also has a stop-block to register from the outer radius. By building the jig in this way, the same jig was used for mortising both the wedges and the arrows.

As shown, the jig is set up for the right-hand side. To reverse it for the left-hand side, I removed the front face (vertical piece) and screwed it to the other side of the jig (the side with paint on it). Then I did the same with the stop block.

The next step was to make the skirt boards. Unfortunately I didn’t have time to take any pictures of the actual process. I used the same trammel I used for bandsawing the arcs, except I modified it to hold the router. I then ploughed out a semicircular dado to hold the skirt board.

To bend the skirt to fit the dado, I kerf-cut the back side with the bandsaw. I performed some calculations and found that the maximum kerf spacing to achieve the bend was 2 inches apart, but I wanted a smoother curve than that, so I spaced them at 3/4" spacing. This meant that the kerfs would not be totally closed when the curve was bent.

The kerf cuts are not visible because they are under the table. However, what I didn’t want was for a dinner guest to reach under the table’s skirt board and feel the kerf cuts. Using a plastic glue spreader, I filled the kerfs with epoxy resin. It took 2-3 coats to completely fill the kerfs to the point where I could sand the excess resin down to be flush with the wood surface.

The result is a completely smooth surface, and the skirt boards are stronger than the original un-bent wood. The only way to tell that the boards were kerf-cut is to crawl under the table when it is fully expanded.

My father came over to the shop today to see the finished table (yes it is finished now). He had just finished reading this page of the blog, so the first thing he did was to reach under the skirt boards to see if he could feel the kerf-cuts. Apparently from seeing this picture, he didn't think they were completely flush. Needless to say, he was surprised that there was no evidence of the kerf-cuts, and I had to dig out the original test-scrap from the trash to explain to his guests what he was looking for.

After the skirt boards were installed, I could begin lacquering the pieces. I was not happy with the brand of pre-catalyzed lacquer I purchased, and I really struggled with the finish at first. This actually turned out to be a blessing in the end, and the final finish is the best I have ever laid down in 15 years.

Because I use an airless sprayer, the volume and build is high enough that a single coat of topcoat is all that is needed for a normal finish (except for tables). I had planned on laying down 3 coats for the final finish. I can’t remember the exact problems I was having, but after each of coats, the finish was not improving, and after 3 coats of topcoat, the finish was not acceptable, and it was not leveling out.

I decided to completely sand the finish down with my pneumatic sander. Unlike a normal sanding sponge, the random orbit sander can’t follow the contours of the rough finish and will only sand down the high spots. By sanding into the glare of a bright light, I could tell when the surface was leveled when there was no more areas of gloss left.

After the surface was sanded level, it still took 3 more coats to finally achieve a good finish. Actually, the second coat was nearly perfect, but I found one flaw on one of the pieces and decided to recoat all of them. I’m glad I chose to spray this third coat, because it turned out to be a huge improvement over what I had already considered to be a good finish.

Before I could mount the mechanical system hardware to the wedges, I first needed to verify that all of the components interacted properly.

To eliminate a lot of the alignment variables from the previous table design, this design has all of the mechanical components self-contained and functionally separate from the wood components. The wood components simply screw down to the functional hardware. The black plates are what the slides and drive mechanism mount to from below, and the wooden wedges mount to from above. The most important aspect of this test-fit was to verify that the angles between the plates was correct, and that the sum of all of the angles was exactly 360.0 degrees without any gaps.

This was a fully functional test fit of the hardware and included testing the motor drive components too. I even made a video of the table opening and closing, but I haven’t processed it enough to publish yet.

One thing that this test fit revealed is that my original motor speed was still too low and not producing sufficient torque. In order to increase the motor speed, I also needed to increase the size of the capacitor on the speed-control variable resistor (see discussion on previous page). It also revealed that I had some frictional binding of the slides due to the pre-loading (anti-sag) of the steel subframe. Part of this was because the table was sitting on saw horses, and not on the pedestal. However, just to make sure this was not an issue later on, I added some foam isolators to the mounting points of the linear slide pillow blocks.

This test fit was extremely successful and showed that all of the hardware parts lined up perfectly (better than I had hoped). It also showed me that the motor had plenty of power to drive the system. When I first tried to open the mechanism, I didn’t notice that the table was too close to my lumber storage rack. When the table opened, one of the plates hit the immovable rack. Instead of stalling the drive, it pushed the entire table top away from the rack!!! (I designed the mechanical system to have its greatest leverage at the two points where it needs to close down and compress.)

After the functional test of the components, I disassembled all of the hardware and mounted the pre-assembled plates to the wedges. Here you can see the three linear slides and the riser block needed for each wedge.

The lacquer application steps for the wedges was for error correction, but the results were so good that I decided to implement the same steps as standard procedure for the arrows.

In a sense, the first 3 coats were sacrificial. After laying down 3 building coats, I nearly sanded them completely back off, but the result was a dead-flat surface, ready for final build coats.

The wedges (above) were lacquered over a month ago, but I just sprayed the arrows last week. What’s interesting about this is that this is a new picture I took with some new equipment. (I still have a lot to learn, though.) I picked up a set of “soft-box” photo strobes last week.

For the first time, I finally have enough flash power to overcome the florescent lights in the workshop. That’s why my workshop floor looks so dark and greenish in many of the previous pictures (a side effect of non-color-corrected florescent lights). Granted, I over-powered the main flash in this shot, and completely screwed up the back light, but I was trying to get some reflected glare off the lacquer surface. I’ll need to do some experimenting to get my camera and flash settings better.

Up Next: A lot of the tasks coming up will be the mundane tweaking and adjusting, but this is coming close to the end.