The Ugly Duckling: Redundant By Design

by Web FishAug 30, 2012 @ 12:21pm

So we spent all this time on this section of the blog talking about everything else BUT Pilot Fish. It's time to unveil or design. Before we do, let's summarize everything we've covered so far:

  • Possible energy sources: solar-driven photovoltaics, solar-driven heat engine, wind turbine, sail rig, a combination of any two or more of the above;
  • Design objectives: structural integrity capable of withstanding the wave and weather systems energy, maintaining correct vessel orientation at any time (flip issue), corrosion, surface deposits above and below waterline, foreign objects in the water;
  • Primary design principle: redundancy;

After a long deliberation, the decision was made to go with solar power as a primary source of energy and electric motor(s) as a primary propulsion system. It's been very much a battle of attrition - it's not so much about what we like about this solution, it's more about what we dislike about all other solutions:

  • Moving parts: This solution provides the least number of moving parts (with the solid wing system coming at close second). Less moving parts - less chance for failure;
  • Efficiency: Hard to get close to 100% energy harvest rates while satisfying all design objectives, making the Power Budget issue that much tougher;
  • Availability: All other options require various levels of custom fabrication;
  • Fit with the rest of the design philosophy: Probably most important of all (as you will see below).

Since we are using photovoltaics (solar panel) as our energy source, we now need to ensure we capture as much sun-light as possible within the limitations of our 1m3 virual cube. The sun is (generally) shining from above, so we need to maximize our harvest size in the horizontal plane (for the time being we will not worry about strictly following the sun's direction). A typical boat hull looks something like this:

 

 

Producing a structurally sound hull would not be a major challenge. With the proper keel in place, we also solve the flip-over issue. But... the only area guaranteed to face the sun at any time is the deck. To achieve any reasonable speed, the Length/Width ratio should be no less than 6:1. This leaves us with ~1/6th of the usable energy. If we take a look at our budget again, we can see we are way below our goals.

How do we extract more surface area from our hull? First thing that comes to mind is mount a horizontal plane above the hull extending to the boundaries of our virtual cube (think hellipad on a very small mega-yacht :):

 

 

In addition to raising the center of gravity really high and increasing the listing (tilting) and pitching (yawing) forces to unreasonable levels, we'll and up with a 1 square meter of plowing surface (not necessarily the thing you want when trying to maintain speed with minimal power reserve). And... we just made our self-righting task MUCH tougher - flipping this monstrosity back would require a deeper mounted and much heavier ballast. Plus we are not helping our redundancy.

A platform that would have most of the benefits of the above design while minimizing the negatives is a catamaran:

   

 

We do keep all of the surface area, avoid the erratically pitching deck and add a solid dose of redundancy. The issue with this design: its very stable on the surface of the ocean, requiring more force to flip over. Wait, this is a good thing, right? Well.... not unless you are designing a self-righting solution. Wave forces will have no problem flipping it over. But designing a self-righting single-keel catamaran is (very) tricky. A quick back-of-the-envelope calculation shows that it will be next to impossible to design something like this while both utilizing all of the surface area (maximum solar power) and fitting the whole system within 1 meter vertically (righting effect is proportional to the mass of the keel bulb and the square of the keel depth, plus the angles just don't work).

At this point we spent the obligatory time to research the various self-righting solutions for catamarans. Then we spent triple that designing our own crazy ideas (computer-controlled balloon inflation, articulating arms with ballast, etc.). Not much seems to work.

As we were getting ready to give up and go back to the single-hull idea (with all its negatives it could still be a workable solution - just not a very good one) there was one final thought: instead of fighting the flip-over and spend time, effort (and boat weight/displacement!), why not... EMBRACE IT!

That's right. The main reason for us trying to solve the flip-over issue was to keep our solar panels pointing towards the sun and keeping our propeller(s) in the water. But what if we don't need to do that? Getting the sun to shine from below could be a problem... But getting solar panels on both sides of our deck would be trivial! Getting the propellers to work in the air would be a problem as well... No, not really. Easy to move the whole drive line up or down under gravity. Or even better, how about we replicate the propeller and rudder setup on both sides of the vessel and intelligently decide which ones to activate! So, we end up with something like this:

  

 

Now, we should admit, this is will not be the best looking vessel to sail the ocean... And there are still many ways things can fail. But this approach solves MOST of our design objectives:

  • Structural integrity (no small / fragile /moving parts sticking out);
  • Maintaining vessel orientation (or not!);
  • Corrosion-resistant (again - nothing that cannot be built using composites and/or aluminum alloys);
  • Foreign objects (no big keel collecting sea-weed as we go);
  • Redundancy (on so many levels!):
    • Redundant solar power;
    • Redundant rudders (as long as a rudder on one of the hulls is functioning, we are in decent shape);
    • Redundant propulsion left/right (we could ride on one motor as long as we have a functional rudder to compensate);
    • Redundant steering (steer both with rudders and left/right motor on/off);
    • Redundant  flotation (breaching a compartment in one of the hulls should have less impact);
    • And all this is fully redundant by flipping the vessel over. We almost WANT the waves to flip the vessel every once in a while if we have a problem;

We still have quite a few left to tackle, but those are now engineering challenges. The foundation is set. Time to start drawing!

See you on the PilotFish Anatomy design blog!

Failure IS an option

by Web FishAug 19, 2012 @ 05:01pm

You've probably heard before the phrase "Failure is not an option". Unfortunately, when it comes to complex projects like this, every element that goes into the system has the potential to fail. Motors burn out. Propellers get wrapped in seaweed. Hulls get breached. Cables get snapped. And the overall system is only as reliable as it's weakest link.

How do we deal with this fact of life?

  • User higher quality components

This is an obvious one. Higher quality (usually) means "more expensive". It's a trade-off. We don't want to loose the vessel because of the failure of 25 cent washer. But we can't afford to spend $250 per washer either.

  • Design with reserve

A 12 Volt motor will (most of the times) last several times longer when run at 6 or 7.5 Volts.

  • Redundancy

Make any critical system redundant. Either via back-up or by parallel systems. The chance of coincidental failure of two parallel systems is (usually much) lower than of a single one. 

As you will see next week, redundancy is the top factor in the selection of our vessel platform.