Efficient solar powered (electric) kayak

Hi everyone,

My first post. I'm adding electric propulsion to my kayak with a goal to travel at minimum 5km/h using two 50 watt solar panels (5.5 amps total). I thought I'd write up what I've done so far, hopefully for some good advice as to what to do better. And if nothing else just share my results for others who may be interested.

I'm using a MinnKota 40 lbs thrust electric trolling motor with Digital Controller on a 13' inflatable Helios 2 kayak (26" beam at the waterline).

Initial tests showed 8.6 amps while travelling at 5.1km/h on 50% throttle position. Well above my 5.5 amp goal.

Replacing the MinnKota power prop with the APC 10x6 Sport pusher model airplane propeller and a 3 inch spinner produced 5.7 amps while travelling at 5.1km/h on 37% throttle position. Much improved and arguably a "pass"...but we always want to do better!

This leads me to a couple questions:

1) Does anyone know if these motors run equally well in the "reverse" direction (by swapping the wiring to the motor, or rotating the permanent magnet)? It would open up a wider selection of propellers. MK seemed OK with this but they couldn't tell me if the brushes were timed in any particular direction so I'm not confident in their answer.

2) I've downloaded JavaProp and figured out how to use it at a very basic level.

Here's what I did.

I guessed from paddling the kayak that it's probably between 5-10 lbf need to paddle at 5.1km/h. So I entered 30N for thrust required. I then entered my prop and spinner data and the desired velocity of 1.4m/s. I guessed 600 rpm prop speed based on a maximum quoted spec of 1200 rpm for the MK motor and the 50% throttle setting.

All this produced Javaprop results of:
71% efficiency
59 watts at the propeller
Prop pitch 170mm

This seems a pretty good correlation with actual results. Especially since I had to guess at a couple parameters (kayak drag and actual rpm). Do you experts agree? If not, what else could I, or should I, have done?

If the motor can run in "reverse', I can use a 14 inch x 8 inch prop. This gains efficiency to 77%. But it increases the torque loading by 25% and PMDC brushed motors apparently lose efficiency quite significantly as the torque increases so there is a chance the propeller gains might be offset by the motor losses. However the motor may not be in the falling part of its curve at these low levels so I don't really know. Thoughts anyone? What else can I do, or should I be doing, with Javaprop?

3) How much drag does the 1.125" tube leg cause? It extends into the water about 12 inches. I held a similar sized pipe in the water at 5km/h and guessed there was maybe 1 lbf of drag. It seems this might be a candidate for a fairing to reduce drag although I'm not sure how this translates to actual drag given the tube leg sits at the rear of the kayak. Anyone care to hazard a guess in the actual drag reduction? 1 lbf would save almost 0.5 amp which would be worthwhile.

Obviously I'm learning as I go here. I'd still like to get the total draw lower but am running out of ideas. Any advice or input greatly appreciated!

 

Pmdc Motor

I have a 3/4 horsepower 24 VDC Pacific Scientific motor for an electric boat and had a discussion with their tech people about their motors. The Pacific Scientific motors are high quality high efficiency brushed motors, and come in sizes from small fractional to a couple of horsepower. I was told that the reversible PMDC motors never have a bias placement of the magnets or any brush advance/retard, so there is no matter which direction the motor runs. I suspect this information would be applicable to your motor, but that is just a guess from this mechanical engineer.

There is a reliable way to test your motor to verify it has no brush bias. Mount a common disc (pressed sheet metal pulley, or gear, or sprocket) on the output shaft to produce drag and load the motor when submerged in a water tank. Running full ahead, or full astern, a simple disc will produce the same resistance torque vs. RPM. Measuring the amps in forward and reverse will supply your answer. A disc about 150 to 200% of the prop size should be about right to load your motor.

As to the electronic speed control, I do not know if there is any applied voltage bias in that circuitry, but by just using a direct wiring arrangement you will eliminate that uncertainty. You can just use jumper cables for testing without any electronics.

My boat uses two 12 volt batteries, and with two DPDT switches we get (A) forward 24 VDC, (B) forward 12 VDC, (C) reverse 12 VDC and (D) reverse 24 VDC. Full speed has the two batteries in series, half speed has them in parallel. While this simple switching scheme does not allow infinite speed control, it has proven wholly adequate for cruising, and has virtually no losses. Full speed is usually run, but half voltage gives almost exactly half speed, which equates to 12.5% power, greatly extending range with about 20x the available nominal run time.

 

Pmdc Motor

With respect to the drag produced by the cylindrical pipe which supports the trolling motor pod, significant improvement can be had by making the shape similar to an airplane wing with an aspect ratio of about 3:1. The drag coefficient of the cylinder is in the range of 0.5, the symetrical wing shape will be 1/10 to 1/20 that number.

The resistance force in pounds is equal to: F = Cd * Frontal Area * Fluid Density * Approach Velocity ^ 2 / 2. Proper units are required here, Force is in Pounds, Cd is a dimensionless number, Frontal Area is in square feet, and velocity is in feet per second. The density must be expressed in Slugs per cubic feet. We usually express density in pounds mass per cubic feet, and to get Slugs per cubic feet divide by 32.2. For water with a density of 62.4 LBm/ft3 = 1.94 Slugs/ft3 = 1.94 LBf-sec2 / ft4

This will tell you how much streamlining that member will reduce your overall drag.

As far as loaded hull resistance goes, freeware like DEFTSHIP will give you an estimate, but you will need to model your hull. Better results if you can be towed at constant velocity (near your desired speed) with a long string, and measure the string pulling force. I would think some of the experts here who have examined kayaks could give you a very close number for your hull drag vs velocity.

 

The drag coefficient of a cylinder is a bit over 1 at that Reynolds number. Thus the drag is about 10 N (2.2 lbf). With say NACA 0012 - 0030 you would get frontal area drag coefficient of about 0.07 and thus drag would be 0.6 N (0.13lbf). It doesn't really matter how thick NACA profile you choose, since increased area is quite well compensated by higher Reynolds number.

If you don't want to do a real profile, just an ellipse of 3:1-6:1 will give a drag coefficient of 0.4-0.25 and reduce the drag to 2-3 N (0.5-0.7 lbf).

 

DC motors with permanent magnets reverse direction by reversing polarity. They are equally efficient either way. A 50W solar panel will only produce 5.5A in a stationary installation for a few minutes a day and maybe a couple of hours on an installation that rotates to follow the sun. The rest will be less. On a boat, you should consider maybe 50% efficiency. It will never be optimally oriented so consider 2.3A for three hours, with less after. That will be on a clear sunny day only.

 

See Efficient Electric boat MCdenny posts 10-7-2009 post #252 on motors. You will probably need at least a small battery backup with solar power for safety reasons and because of what is pointed out by gonzo's post.

PC

 

Motor Timing: Thanks for the input on the motors everyone. I did a quick test in air with symmetrical results. However it only loaded to 4 amps which is 10% of the maximum rated so I will do a better test when I have time. But so far so good. Thanks for the suggestion.

Of course all this assumes the brushes are mounted mechanically neutral from a brush wear viewpoint. Small r/c motors often have brush mounting that favors the preferred direction to improve wear/longevity. However from the youtube videos I've seen of the MK brushes this isn't the case here. If I'm really concerned I can always take mine apart to verify but I don't currently plan to.


Pipe drag: The graphs I've seen suggest Cd of 1 is a reasonable number for the pipe so with that assumption I get 8.5N which is significant as everyone has pointed out. I don't believe a proper fairing will be that hard to make from wood so will give it a go and see what happens. Given I may have over-estimated my solar panel this is especially useful info, thanks!


Solar Panels: My 5.5 amp assumption is already derated from 8.3amps (100 watts is 8.3 amps at 12 volts) but perhaps not enough then. If it ever stops raining, I will do some tests to refine these numbers since I now have the panels.

One issue is these panels produce maximum power at 17.6v. Also the power peaks at different voltages depending on solar conditions. I had originally considered feeding the solar panels directly with no controller and no battery but it seems a smart controller (maximum power point tracking) will be needed to extract the most from the panels over varying conditions throughout the day. I've never used one so don't know if they require a battery to be connected to function properly.

I'm not adverse to a small battery as it ameliorates the vagaries of solar but I had originally planned on treating the solar power similar to wind...cruise if you can...paddle if you must! LOL

 

Portacruise...thanks for the link! That post #252 is 100% relevant. Nothing new under the sun! LOL.

I was especially intrigued by how much his fairing and spinner gained. I see he used a NACA0025 profile so I'll try that one as it meet the criteria discussed so far.

And it seems we used the same propeller (not surprising since it was the only suitable "pusher" prop available at my hobby store). Maybe even the same spinner! (again perhaps not surprising since it also was the only one at my local store that fit properly).

I'll have to read the rest of the thread. Especially to find if he tried larger "tractor" props and running the MK motor in "reverse". Again, thanks for pointing out the link.

EDIT: Rereading his post, I see by the prop in the picture that he *is* running the MK in "reverse". And it seems he tried dozens of props before settling on the 10x6 ratio that I stumbled upon as my first one. That's a bit disappointing actually but he appeared to be optimizing for a much higher power/speed point than me so our final results may yet differ.

 

If you use a long thin foil, e.g. NACA 0012, it will help as a rudder.

 

I was going to comment on this earlier but forgot... I think this is a great, practical idea. My idea was to connect the solar panel directly to the motor without any solar controller, motor controller or battery. Speed control would be infinite and accomplished with a wave of the hand over the panels (creating appropriate shadow). I may not have enough power however and require a battery but we'll see how it goes.

For my purpose, batteries fundamentally change the dynamic. I've put sails on my sea kayaks. There is something magical about paddling with sails and the wind picks up. Things always seem faster when you're down at the water surface among the waves. Inevitably the wind dies of course and you're back to paddling but the experience just makes you eager for the next opportunity to harness the wind!

Since sea kayaks are paddling machines, we think nothing of paddling 20km. My idea for the solar drive isn't to turn a sea kayak into a power boat. The sails or solar drive is there for "fun" factor and the serendipitous rest opportunities. My wife, however, seems to be liking the battery idea... LOL

 

Trolls are made w/ cheapest possible brush / rear bush bearing motor guess 70% efficient. The large blunt front disturbs flow on trolls. An electronic commutated motor geared down, with tiny flex shaft might be the better way using a slow turning 14 prop?

PC


QUOTE=ElectricKayak;729800]Portacruise...thanks for the link! That post #252 is 100% relevant. Nothing new under the sun! LOL.

I was especially intrigued by how much his fairing and spinner gained. I see he used a NACA0025 profile so I'll try that one as it meet the criteria discussed so far.

And it seems we used the same propeller (not surprising since it was the only suitable "pusher" prop available at my hobby store). Maybe even the same spinner! (again perhaps not surprising since it also was the only one at my local store that fit properly).

I'll have to read the rest of the thread. Especially to find if he tried larger "tractor" props and running the MK motor in "reverse". Again, thanks for pointing out the link.

EDIT: Rereading his post, I see by the prop in the picture that he *is* running the MK in "reverse". And it seems he tried dozens of props before settling on the 10x6 ratio that I stumbled upon as my first one. That's a bit disappointing actually but he appeared to be optimizing for a much higher power/speed point than me so our final results may yet differ.[/QUOTE]

 

Pulleys, gears and sprockets? Spoken like a mechanical engineer! LOL. I don't have such things especially ones 15-20 inches in diameter....so I used a 4 inch piece of wood which loaded the motor to 37-40 amps (maximum spec is 42 amps).

I tested forward and reverse going from 0% throttle to 100% at intervals as market on the throttle. I actually always used the forward position on the throttle and achieved reverse by swapping the wires between the controller and motor.

I also reversed the wood and reran the tests. These indicated a small asymmetry in the wood around 2 percent.

I then jumped the battery directly to the motor (bypassing the controller) and redid the full power tests.

Overall forward is slightly stronger by a few percent so it does seem to be forward biased. I've attached all the results. I understand operating it against the bias increases arcing but I don't know how much.

Question: Does brush timing affect efficiency or only power/torque?

It's probably easy to adjust the brush timing. Loosen two bolts and rotate the end piece (I think). But maybe getting it this close is actually hard. I really have no idea. My sense is that it isn't worth fiddling with but I'm open to advice on this.

 

I think you have all the answers but just need the confidence to try it. Make the fairing, fit the RH (tractor) propeller (you should be able to reverse the connections between the speed control and motor to avoid always using "reverse") and reap the rewards! But if it's any help, I think that a MPPT solar controller and a small battery are necessary. And have you considered a lower drag rigid canoe/kayak? And can we have a picture?

ps just remembered, you can reverse a MK by removing the body containing the magnets, rotating 180 degrees and refitting.

 

Yes everything is ready to go...just waiting for the dratted rains to stop!

I've made the leads from the controller to motor quick connects (standard MC4 solar connectors actually) so I can experiment. And yes rotating the magnet section 180 degrees will reverse the motor but no need given swapping the wires is so easy.

 

It has been difficult for me to measure low speeds around 3 knots in open water. Even the slightest breeze or current makes quite a difference. Also I've had to find ways to keep the watts draw constant for a fair test between different props because volts and amps both drop as a battery discharges. GPS readings don't seem to allow for situations where you are fighting even a slight breeze. I use a sailboat knot meter which seems to compensate for currents but not wind: http://www.knotstick.com/ A large indoor pool and stopwatch might be the way to go, if one could find access...

Hope this helps.

PC