Microship Physical Substrate

NOTE: This discussion was written in 2000 or so.  Considerably more detail is available in the Resources section.

OK. Let’s turn all that into reality. The first issue, and one that provided many years of entertaining distraction, is just what, exactly, is going to float this whole affair. The substrate has taken on quite a variety of shapes and sizes during the comparatively linear development of on-board systems, culminating at last in the twin micro-trimarans now under construction in our island lab.

The issue of scale is a critical one. It may seem most rational for a traveling couple to occupy the same boat, and as I concluded at the close of the Fulmar epoch there are indeed some compelling arguments for this: companionship, the ability to rotate watches and share difficult tasks, avoidance of equipment redundancy, logistical simplification at marinas, and so on. But the Microship project specification also includes the requirement for unassisted haulout under human power at any time, including the possibility of limited overland transport... and any boat big enough for two people to live aboard full time (without killing each other from overcrowding) is simply too large to be moved without a trailer and tow vehicle. In addition, the whole “aesthetic” that renders this adventure philosophically congruent with the computerized recumbent bicycle that preceded it calls for small personal machines powered by pedals, solar panels, and sail.

Given all that, our key nautical issue became one of finding (hopefully) or building a pair of small trimarans. One thought was to revisit and hack the Fulmar-19, but as I discovered during the 1994 adventure, the inability to sleep on board or haul a loaded boat above tideline without 2-3 additional strong bodies led to all sorts of daily headaches... and repackaging the boat to solve those problems would be daunting. Other commercial trimarans of the same general scale, while endlessly amusing to daysail, share the same basic problems.

So... we had no choice but to build our own, hoping to violate the aphorism that the average completion time of a homebuilt boat is 135 years.

Fortunately, we’ve had the opportunity to steal from the best. The Microship center hulls are based on Wenonah Odyssey canoes, fashioned of Kevlar using their ultra-light construction method with foam-core bilges and stiffening ribs. Our own custom decks, laid up with 10-ounce fiberglass and West System epoxy over .5-inch Divinycell PVC foam core, integrate the canoes into the complex system of stresses and fixtures necessary to accomplish our goals; added plywood bulkheads at the high-stress areas stiffen the entire structure while giving us sealed forward and aft stowage bays accessed through gasketed hatch covers.

Crossbeams (A.K.A. “akas”) and outer hulls (“amas”) are off-the-shelf units from Fulmar Canada – life is too short to re-invent the wheel! The scale turned out to be ideal, so we simply boxed in the U-section akas temporarily with scrap Formica, pulled a mold, and integrated the matching nests into the decks. Likewise, we were able to shortcut the whole rigging can o’ worms by applying stock WindRider units – elegantly shaped medium-roach sails on freestanding aluminum masts, which we stepped in Teflon-anodized aluminum tubing glassed to the forward structural bulkhead. To simplify furling, we added a drum for remote mast rotation from the cockpit; the boom, mounted to a gooseneck bearing, is sheeted to a Ronstan traveler on the “arch” behind the seat.

The remaining basic sailboat components are the appendages – and again we worked around available components with the guidance of multihull marine architects John Marples and Jim Antrim. The rudders are carbon-fiber foam-core units made for us by Moore Sailboats, hung from integrated pintles via a custom hydraulic kick-up assembly and controlled via two more pairs opf Clippard cylinders by a pair of pivoting aluminum T-handles... which also carry computer interface hardware and a few other controls. The boards aren’t quite so sexy – they’re clunky used Nacra daggers picked up from a sailing shop in Berkeley (hopefully to be cloned in the spirit of the ultralight rudders after initial testing!). These are perhaps the weirdest looking part, from a boat-design perspective: in order to support the fundamental requirement of sleeping on board, they are deployed through molded trunks at the turn of the port bilge... angled forward in a completely non-intuitive way to maximize the chance of favorable retraction on grounding and to simplify manual retraction from the cockpit.

Tech note on the daggerboard: Positioning the board to provide a bit of weather helm is critical, for the opposite can get you into trouble in heavy airs by falling off into a high-stress condition that can cause capsize or dismasting if you don't instantly uncleat the main. Heading up when rudder control is lost depowers the boat and leaves you safely – albeit embarrassingly – in irons. Accomplishing this requires the below-water CLR (Center of Lateral Resistance) to be slightly forward of the above-water CE (Center of Effort). As each is the complex resultant of many factors that can themselves vary with wind conditions and reefing, I never felt too confident in our analytically derived daggerboard placement... which is why we used a temporary rudder lashup for the first on-water test (leaving ourselves one last adjustment before having to tweak the helm balance with a Sawzall) But all worked as planned, so we went ahead and finalized the rudder mounting.

Beyond the aforementioned sailboat components, all that’s necessary to get the boat on water is a place to sit, a deckload of rigging components shuttling braided line to and fro, and a touch of wind. But many more design issues contributed to the shape of the physical substrate...

First, the on-water bivouac requirement dictates bulkheads about 8 feet apart, conveniently matching the spacing of our crossbeams. The adjustable seat, based on a webbed recumbent bicycle unit made by Ryan, can be flipped back onto the afterdeck to allow a camping mattress to be flopped into the bilge. To increase the comfort and security of on-water sleeping, a formfitted screened canopy is hung from the arch and stretched over the cockpit to the Lexan windshield, keeping out rain and critters -– while a Nicro powered vent circulates fresh air through the coffin-like confines of the hull.

Another essential requirement is human-powered haulout – easy with a kayak but prohibitive for even small multihulls burdened by touring load. Since open-ended wandering precludes land-based trailers and tow vehicles, each boat carries four Seitech wheels... lovely 16” diameter pneumatics with red plastic hubs and uncorrodable Delrin needle bearings rolling on a 1” shaft. These deploy like aircraft landing gear from their nesting place under the solar panels, effecting the transition between retracted and extended states by pivoting around the planes of the crossbeam bulkheads via tensioned lines under control of levers at the gunwales (this complex but lightweight solution reluctantly adopted after much analysis of seductive but impossible hydraulic approaches). By dropping the wheels while approaching a ramp or beach, the boats can be dragged to a campsite or out of storm-toss’d seas without logistical dependence on a chase vehicle.

Active steering of the forward landing gear turned out to be necessary, so there are hydraulic cylinders on each strut coupled through a stainless steel cable assembly to rotating collars, which in turn couple their rotation to the spindle bearings through scissor assemblies. To haul the boat, I simply strap on a harness and start pulling, with my hand on a winch handle protruding from a hydraulic system at the bow. This consists of a cam follower that implements the Ackerman steering function (to minimize wheel scrubbing, which is a non-trivial problem), as well as an extra cylinder that biases the system in such a way that the front wheels become pigeon-toed... effectively holding the boat on a hill.

The landing gear suspension has 8” of travel and can theoretically handle 4G shock loads, cushioning the impact of a hard surf landing with stacked bumpers of DuPont Hytrel from Miner Elastomer (I sure hope we never get to test that feature!). The development of this system, with custom machined front-wheel spindle bearings, TIG welded tubing, deployment system, hydraulic steering, and countless unanticipated details, took over a year and many thousands of dollars... but it gives the amphibian Microships a capability unique in the boating world: unassisted launch and retrieval as long as there’s a ramp or reasonably accessible beach available. A simple fixture built around the boathook plugs into the mast step and allows the 21-foot rig to be carried atop the folded boat, with the second support point being a foam block plugged onto the traveler on the arch.

Much of the remaining “deckage” supports the fixtures necessary to meet the needs of our electronics systems – retractable solar panels and their waterproof interconnects, a thicket of antennas, video turret, data collection tools, and so on. Most notably, the control console of each boat is a sealed unit packed with computers and communication gear, bolted to a nest occupying the forward half of the 8-foot cockpit area and protected by a foam-core cowling that carries the graceful lines of the boat much better than a funny box covered with hinged and gasketed panels. This was the most critical packaging problem, as anything else on board (other than the Macintosh laptop nestled in a Pelican box in the aft hatch) can probably survive an encounter with salt water... but not the dense collection of surface-mount printed circuit boards that satisfy our primal gizmological urges!

The basic skeleton of the console is a complex frame of molded fiberglass angles, filling the space defined by cowling, deck, and pedaling envelope as fully as possible. Every surface plane of the resulting microtrideckahedron is surrounded by flanges on all sides, supporting hinged panels of thin foam-core fiberglass. The exceptions are the baseplate, which is aluminum with stiffening ribs and welded mounting bosses for heavy hardware, and the “front panel,” which is a single unbroken expanse of surface abrasion-resistant polycarbonate behind which are mounted a dense array of LCD panels and other user displays. The foam-core panels support all the circuit boards, which are mounted on little standoffs screwed into flat pads that are simply glued in place.

Why foam core? The trick to this whole thing is not just careful sealing, but pressurization (at an estimated .5 PSI above ambient pressure), and the .062” aluminum I used on BEHEMOTH would bulge like a balloon. It is impossible to achieve a perfect seal in something like this, and without running the whole box at elevated pressure, thermal cycling would result in constant breathing through the leaks... contaminating the interior with inhaled salt- and moisture-laden air. But in a pressurized box, the inevitable leaks just outgas, making them easier to find while allowing the control system to diagnose its own enclosure integrity by occasionally interrupting the pressure source and watching for a pressure drop uncorrelated with temperature change.

Canoeing just ain’t what it used to be...