The purpose of this project is to design and mass-produce kits for a floating tiny house that can sail. It combines high-tech modeling and fabrication and low-tech assembly that can be carried out DIY-style on a riverbank or a beach. This boat is a four-bedroom with a kitchen, a bathroom/sauna, a dining room and a living room. The deck is big enough to throw dance parties. It can be used as a river boat, a canal boat or even a beach house. It's rugged and stable enough to take out on the ocean. Kits will start at around $50k (USD). The design has been tested in simulation and prototype; full-scale production will begin next year.

Wednesday, May 22, 2019

The Rudder

Rudder assembly
Quidnon’s steering has evolved quite a lot since the original concept. Now all that’s left of the original concept is the idea that the rudder should have a kick-up blade: when sailing across shallows it should gently float up instead of getting torn off or getting stuck, and when the boat settles on its bottom at low tide the rudder blade should automatically get itself out of the way. Only now has a good solution to this problem has finally been found.

Early on it was thought that twin rudders and wheel steering made sense, but this made the design complicated and expensive. Twin rudders require a complicated steering linkage that uses something called Ackermann geometry, which is also used in cars: when turning to the right, the right wheel has to turn more than the left wheel because, being closer to the point around which the car turns, it has to follow a tighter circle.

Later on, after single-handing a 36-foot sailboat from Boston to South Carolina, I discovered that wheel steering is a bad idea and that I prefer a simple tiller. There are very few steering positions that are comfortable with a wheel: sitting behind it and standing behind it are more or less the only choices, and they both get tiresome rather quickly. On the other hand, with a tiller, it is possible to steer the boat while standing, sitting or lying down, using hands, feet and hips, or, with a tiller extension clipped on, with the inside of the knee or the armpit. It is possible to operate the tiller remotely, by tying a bungee cord to one side of it and pulling it to the other side using a lanyard.

In turn, a tiller on such a boat of this size is only workable if the rudder is a balanced rudder, with about a third of its area ahead of its rotational axis, so that the boat can be steered with a fingertip instead of your heel on the tiller pushing with all your might, as is the case with an unbalanced “barn door” rudder that rotates around its forward edge.

Further on, I discovered that Quidnon doesn’t heel enough to make twin rudders necessary: just a single rudder would work fine, and so the design was changed to a single rudder was hung off the center of the transom. But this arrangement was still somewhat problematic. First, the rudder assembly cluttered up the transom and made the boat a bit longer (which is a problem because marinas charge slip fees by overall length). Second, the pivot point of the rudder was too far from the cockpit to give the tiller a useful swing range.

Rudder assembly installed in engine well
And so the rudder was moved from the transom to the back of the engine well, where there is just enough room for it. This made it possible to solve a few more problems.

Quidnon doesn’t always need to have a rudder. It is a houseboat, and houseboats mostly just sit at the dock, where having a rudder is not just unnecessary but also rather inconvenient. The tiller tends to whip around and hit things whenever the tidal current shifts or a boat wake hits. Since it sits in the water, it tends to accumulate marine growth which makes it not work very well when the time comes to move the boat. A better solution is a rudder assembly that is easy to install and just as easy to take out again when the boat is at rest.

Gudgeons in engine well: top view; aft view

With the rudder assembly removed, all that remains on the hull are two gudgeons bolted to the back of the engine well along the centerline. To install the rudder assembly, it is turned 90º, so that the tiller faces directly sideways and lowered into the engine well using a hoist. The rudder shaft has two pintles that engage with the gudgeons. The lower pintle has a longer pin than the upper pintle, so that it can be engaged first rather than having to try to line up two pintles with two gudgeons at the same time.

Once the rudder assembly is dropped into place it can be turned to the 0º amidships position, with the bottom part of the assembly sliding under a recess in the bottom of the transom. This recess serves several purposes: it provides an exit path for the stream from the propeller; it also provides an exit path for the outboard motor’s exhaust when it is in idle (when it is in gear the exhaust goes through the propeller and into the water); lastly, it provides a space for the rudder assembly.

The bottom part of the rudder assembly consists of the rudder blade case and the rudder blade. The case is a box, welded out of mild steel and galvanized, with its bottom and rear open and forming a slot from which the rudder blade protrudes. It is welded to the bottom of the rudder shaft (a steel tube) and reinforced using a triangular gusset. The gusset has a hole in it for attaching a lanyard by which the rudder assembly is hoisted out of the engine well. The sides of the box have specifically shaped cut-outs in them.

The rudder blade is made of a 3/4-inch (20 mm) piece of plywood sheathed in fiberglass and painted. Close to the bottom of the blade is a circular cut-out that is filled with a lead disk, to ballast the blade to counteract the buoyancy of the plywood and to exert a certain downward force when submerged. The top of the rudder blade is surfaced with epoxy that’s loaded with graphite powder, to create a hard bearing surface.

Rudder blade detent mechanism: roller guide and rollers

The rudder blade is joined to the rudder blade case using 4 rollers, 2 on each side, that are through-bolted to the blade and ride inside the cut-outs in the case. The arrangement of the rollers and the cut-outs acts as a detent: in order to get the blade to kick up the force acting on the front of the blade generated by an obstacle has to be more than 4 times the downward force on the blade due to gravity.

The lead disk is sized so that this force is significantly more than 4 times the force generated by drag with the boat moving through the water at its maximum speed. Once this initial resistance is overcome, the rudder blade kicks up rather easily. Once the external force acting on it is removed (because the boat is again in sufficiently deep water) it floats down into vertical position and the roller mechanism clunks into place. When the blade kicks up, it fits in the recess under the transom.

Rudder blade in kicked-up position

The only necessary precaution is to avoid running aground while moving astern: the rudder blade will not kick up backwards. Most of the time the centerboard will strike bottom first, because it hangs down lower, and will stop the boat, shattering if it has to, in which case it will be time to pull the remainder of the centerboard out of its slot on deck and to drop in a new centerboard. But the centerboard somehow misses the obstacle and the rudder blade doesn’t, then the rudder may suffer a bit of damage.

If the bottom is soft and the boat is moving slowly, it will simply stop with the rudder blade stuck in the sand or the mud. If the obstacle is hard or the boat is moving fast, the bolts holding the rollers to the rudder blade, which are designed to be the weakest element, will sheer off and the rudder blade will drop off. Then it will be time to pull out the remainder of the rudder assembly and to dive to recover the rudder blade and the rollers. Then the rudder assembly can be put back together with new bolts. There is also the chance that the rudder blade will strike and damage the prop, in which case it will also be time to pull out the engine and replace the prop. In short, don’t run aground when backing up!

To summarize: the rudder assembly easily installed and easily removed when not in use and for maintenance. With the rudder assembly removed, all that remains in place are two gudgeons mounted to the back of the engine well. It doesn’t kick up unless it encounters a hard obstacle, with no amount of moving water able to displace it from vertical. Its action is fully automatic, never requiring any operator intervention. It provides for fingertip steering using a tiller because the rudder blade is balanced, with 1/3 of the area ahead of its axis. The use of the tiller makes it possible to use the simplest and most affordable kind of autopilot: a tiller pilot that clips onto the tiller. The tiller itself is of a telescoping type, with a handle that slides into its body, so that it isn’t left swinging about the cockpit when the boat is on autopilot.

After all of the various evolutions, I dare say that this rudder design is very close to final.

Monday, May 13, 2019

The Centerboard


Although the Quidnon blog has been quiescent for the past three months, there has been some good progress on completing the design, and I can now report these results and see what comments, ideas and suggestions emerge. It takes time to come up with simple and cheap solutions to complex and potentially expensive problems.

One of the problems that is now solved is how to provide lateral resistance with minimal complexity and expense. The initial concept included chine runners, which are narrow ledges that extended horizontally from the hard chines at which the bottom joins the sides, and two centerboards that hung down from wrist pins and extended from slots in the bottom in such a way that they would kick up into their slots when encountering an underwater obstacle.

The chine runners were discarded when it turned out that Quidnon’s hull, being quite wide in order to provide relatively spacious living quarters (it is, after all, a HOUSEboat), doesn’t heel enough to allow the chine runners to bite into the water. All the chine runners did was add some drag (and, of course, complexity and expense).

The kick-up centerboards worked well enough, but there is a basic problem with them: since they hang down from a hinge, they are deflected when moving through water, and this makes for erratic steering behavior. The deflection can be minimized by adding ballast, but this makes the boards too heavy. It is also possible to add tensioner lines fed to cleats that pop open when the boards hits an obstacle, but this adds complexity.

The problems with the original centerboard design didn’t end there. How does one remove them for maintenance and cleaning, and put them back in? If this required the boat to be hauled out, then that would invalidate the very important requirement that Quidnon must never need hauling out (haulouts are expensive!). The copper cladding can be cleaned with the boat hard aground at low tide and there is nothing else down there that should ever need attention. And so a plan was created for installing and removing the centerboards with the boat in the water with the help of a diver. But divers are also expensive!

And then another good question arose: why are there two boards? Well, the initial thinking went as follows. Putting a single centerboard along the centerline wastes precious living space in the middle of the cabin by filling it with a centerboard trunk. Moving it off to the side makes the design asymmetric, and that’s functionally unimportant but aesthetically unpleasing. Therefore, let’s have two of them. The flaw in this logic is that the aesthetic consideration matters not at all because the centerboard isn’t visible. Your heart is on your left and your liver on your right, but nobody will ever call you ugly because of that.

If two centerboards are too many, how about zero centerboards? Well, it turns out that having a centerboard is rather important, but only when the boat moves. It is especially important when motoring in and out of marinas, because without the centerboard the boat will drift sideways instead of turning within a tight radius. It is also important when motoring, especially upwind. It is sometimes possible to sail downwind without the centerboard, but that’s about it. But if the boat doesn't move (as houseboats often don't) then a centerboard isn't needed at all, and having one that's quick and easy to install and remove would be a bonus.

And so just one centerboard is both necessary and sufficient. It will be located off-center (to starboard) with the centerboard trunk located unobtrusively in the back of a settee in the salon, sandwiched between it and the water tank. (But we’ll still be calling it a centerboard because offcenterboard is not a word.) The centerboard trunk forms an L-shaped slot that extends from the deck all the way to the bottom (and does extra duty as a deck drain). To one side of the slot is a channel that stops short of the bottom and tapers in a specific way before it stops.


The centerboard is just a piece of 3/4-inch plywood covered with fiberglass for durability. A circle is drilled out of it near the bottom and filled with lead in order to make the centerboard heavier than water, but not much heavier. It doesn’t have to sink particularly aggressively; it just can’t float up. To one side of the centerboard, near the top, is screwed a cam that rides inside the channel on the side of the centerboard trunk. At the very top of the centerboard is a hole used to attach a lanyard by which the centerboard is retrieved. If the lanyard breaks, a boathook can be used to grab the board by the hole. The centerboard is sacrificial and designed to snap without causing damage to the hull. Making a new one is neither expensive nor difficult.


The centerboard will spend most of its life sitting flat on deck. When the boat is getting ready to move, it is installed by unceremoniously dumping it into its slot.


If there isn’t enough water under the boat it won’t go all the way down, but that’s usually not a problem. (You may need to give its lanyard a tug when backing out of a shallow berth, to keep it from catching on things.)


When the centerboard hits something underwater with the boat moving forward (as boats normally move) it deflects aft, but in order to do so it needs to ride up a bit, so that the cam moves up inside the tapered slot. It can stay in this semi-retracted state, bouncing along the bottom, while the boat sails or motors across shallows. This will slow the boat down a bit, but will also provide good steering because the boat will pivot around the board as it digs a shallow trench in the sand or the mud of the bottom.


It is not necessary to remove the centerboard when anchoring above the low water line with the intention of drying out at low tide because it will be forced up entirely into its trunk.

This completes the conceptual design of the centerboard; next on the list are: the rudder; engine mount; mast steps and bow structure. These have all been reworked, and I will be detailing the new designs over the coming weeks.

Friday, February 8, 2019

Frame Joinery Redux

Although most of the problems with hull structure have already been solved, there remained one problem that stood in the way of completing the design: how to join together the frame. It consists of 4x4 softwood (fir) timbers (3.5x3.5 finished size) combined into a box structure that reinforces the bottom the deck, the bow and the transom and provides support for mast steps. After working out a design that included a dozen different steel brackets that had to be custom-fabricated at considerable expense, I realized that I don’t like it at all: too complicated and too expensive. And so, as usual, I sat back and waited for some new ideas to filter in from the ether.

Eventually this happened. An unrelated project required me to build a rectangular frame by joining together some square cross-section sticks using L-brackets. To avoid the wood at the ends of the sticks splitting as I drove in the screws, I wrapped the ends using several turns of fiberglass packing tape. It worked just fine. The tape took up all the force that would have gone into splitting the sticks along the grain, and the resulting joins were impressively strong.

Transferred to Quidnon’s frame design, this technique will make it possible to assemble the frame using just 3-inch-wide perforated steel strips cut to two or three different lengths and bent to various angles. Some of the brackets will need to be bent to specific angles other than 90º, but this is not a complicated procedure.

The procedure for fabricating the frame now consists of the following steps:

1. Using a chop saw cut 8-foot 4x4 timbers to required lengths.
2. For each timber, shave down the first 6 inches off each end on all four sides using a planer.
3. For each timber, router off the corners on the first 6 inches on all four sides using a router.
4. Roll on a layer of epoxy to the prepared ends, wait until it “tacks up.”
5. Wrap each end in three layers of 6-inch-wide fibergass tape.
6. Saturate the tape with epoxy; let it cure.


Frame assembly then consists of matching up the timbers and the brackets and connecting them together by driving in a lot of self-tapping screws using a cordless drill. There is no need to worry about the wood splitting, and the resulting joints are strengthened by the fact that they are pre-stressed: the wood is compressed between the screws and the fiberglass. In a humid marine environment the timbers will gradually absorb moisture and swell, increasing the pressure on the fiberglass and the screws, holding them in place securely. (The choice of softwood for the timbers is critical: when hardwood swells, it generates enough force to burst fiberglass.) The strength of the joint is determined by the force needed to crush the wood fibers, which is somewhere around 6 times greater than the force it takes to split them along the grain.


There are still several more complicated pieces that will need to be fabricated: engine bracket, mast tabernacles, masthead fittings, tiller and keelboard hardware and bow rollers. These are all key elements of the design and there is no way to simplify them. But the frame joinery is now very well in hand and can be done cheaply using components that can be locally sourced in many places around the world.