Sometimes delays are helpful because they allow more time for examination, and for rethinking parts of the design. And so it was with the tabernacle design. My initial plan was to use a joint that transferred all of the tension and compression loads into sheer loads on two large bolts, which I called “Jesus bots,” after the “Jesus nuts” that hold helicopter rotors in place. But then Alan, who is designing a boat similar to QUIDNON—a houseboat that sails, but for inland waterways—pointed out that my design would require extremely high precision in the way the components are fitted, or they would start to move and flex under load, and fail. Alan has lots of experience in aerospace engineering, and a keen appreciation for structural elements. What he proposed was two flanges bolted together, connected by a hinge on one side. This approach is very standard: look at your average streetlight, and that's how it's mounted. It is so standard that it doesn't require any interesting structural analysis: one can simply look it up and plug in the numbers. Since the bolts that hold the two flanges together are subjected only to tension loads, they don't need to be fitted precisely, and the math for sizing them is simple.
Alan's idea solved one problem but created another: how to fit the mast in place prior to raising it. He is used to working in a hangar, on land, when doing such things, but raising a mast on a boat in the water is a different matter. Precisely maneuvering a heavy mast on a swaying deck in order to insert the hinge pin is no easy matter. After much back-and-forth, we arrived at a solution. The hinge pin is welded onto the bottom flange—the one that's part of the tabernacle. The top flange has two slotted tabs that fit onto the hinge pin and two latches that capture it. And so all you have to do to raise the mast is get it into position and flop it onto the tabernacle. Click! After that, the procedure is simple: insert gin pole, attach hoist, and click-click-click until the mast flops forward. Then assemble and torque the flange bolts.
Another bit of very useful information Alan provided had to do with the design of the mast itself. I was thinking of using flagpoles, which are tapered, and of using them as unstayed masts. Alan tried to calculate the righting moment that would be generated by his hull in the event of a knockdown, and realized that such a wide hull would snap any unstayed mast. We also realized that there are a few other, minor problems with the unstayed design. The first is that there is nothing to prevent the mast from swinging side to side as it is being raised. The second is that the taper of a flagpole results in a sloppy fit up top, because the parrels on the junk sail have to be loose enough to be able to lower the sail all the way down. A straight, non-tapered mast would work better. Lastly, flagpoles are not the cheapest solution, which is to use 20-foot sections of 6-inch Schedule 40 aluminum pipe joined into the correct length for the mast using plugs made of smaller-diameter sections of pipe. The same pipe can be used for both the mast and the tabernacle.
And so the revised design has a mast tabernacle that uses flanges and a hinge, a mast made of straight 6" aluminum pipe, and two shrouds to at serve three functions: 1. prevent the mast from breaking in the event of a knockdown, 2. keep the mast from swinging side to side as it is being raised, and 3. take up some of the sideways load while sailing (which can be considerable, especially when sailing to windward, when the forward force is generated by two much stronger opposing forces of the sails and the centerboards acting at a small angle to each other). The design will remain unstayed as far as fore-and-aft forces are concerned, and the risk of dismasting from pitchpoling cannot be dismissed. Here, the addition of a couple of running stays, to be deployed fore-and-aft in particularly bad conditions, seems like a good idea.
There aren't too many options for shrouds on a junk rig. Since the sail slides up and down the entire length of the mast, there can be no spreaders. Nor are spreaders needed for a hull this wide. And so the two shrouds will run directly left and right between the mastheads and the ends of the deck beams that support the tabernacle at the deck, which are part of a trapezoidal frame made of hardwood timbers on which the tabernacle is stepped.
Problem solved... at least until somebody comes up with an even better idea, or finds a problem with this one.