Efficient electric boat

Discussion in 'Boat Design' started by Jeremy Harris, Jun 22, 2009.

  1. Jeremy Harris
    Joined: Jun 2009
    Posts: 978
    Likes: 60, Points: 28, Legacy Rep: 711
    Location: Salisbury, UK

    Jeremy Harris Senior Member


    My first post here, although I've been an avid reader for some time.

    I've been experimenting with electric power for some time, have previously restored an old yacht and now want to build a minimalist inland waterway open launch, with an efficient electric propulsion system.

    The design brief is:

    Maximum length ~16ft
    4mph cruise speed on smooth inland waters
    Light weight (car toppable if possible)
    Relatively quick and cheap to build
    Reasonably elegant
    Accommodate a crew of two
    Reasonable primary stability (my crew gets nervous in "tippy" boats.......)

    Battery endurance of ~8 hours at cruise speed
    Solar top-up and moored battery charging capability
    Varnished wood and fairly "traditional" appearance

    My first thoughts were to build a canoe, rather like the elegant Victorian/Edwardian electric canoes that were popular around 100 years or so ago on the Thames. After much experimentation I've come to realise that, although a canoe hull is efficient and easily driven, it has some inherent disadvantages when it comes to stability and deck area for solar cells, particularly when scaled down to fit my transportability requirements.

    The propulsion system is pretty much designed and built, using a cheap brushless RC model motor. This motor is rated at 2800 watts, but I've re-wired it to run at a lower rpm, fitted Hall sensors to allow it to be driven from a cheap electric bike controller and give very good low speed torque and efficiency and fitted it to a home-made drive leg that has an internal helical bevel gearbox, stainless prop shaft and 8" two blade prop. The battery is a 36V, 10Ah LiFePO4 unit that weighs just 4kg, taken from one of my other electric vehicle projects, but will probably be changed for a higher capacity one if the concept works. 10Ah should give over 3 hours cruising, so I will probably need a 20 to 30Ah pack to meet my design goal, depending on the effectiveness of the solar top-up charge system (I live in the UK, not a place renowned for sunshine...........).

    I believe that I should be able to reach my cruise speed target with a power of around 100 watts, perhaps a bit less. From all the research I've done, including reading just about every post by Rick Willoughby (many thanks!) this seems to be reasonable. I believe that a recreational paddler/rower only delivers around 50 to 70 watts on average, plus all I've read seems to show that canoes only need around 40 to 50 watts to get to this sort of speed.

    Given that a canoe may not be the best way to go, I started looking at other efficient hull forms. I've always admired the shape and style of slipper launches, plus they are reputed to be easily driven. The snag is that they don't fit with my "light weight" requirement and they tend to be broader in the beam than I'm aiming at. However, whilst playing with making some scale canoe models using the tortured ply technique I discovered that I could open out the stern and get a pleasing looking shape, with a fairly flat stern and narrow (around 36") beam.



    This would be cheap to build and light, as it uses just two sheets of 4mm ply for the hull skin, plus some framing, deck bulkheads etc. It would have tandem seating for two, set low in the hull for stability, with decks fitted fore and aft to take an array of solar cells (I'm aiming for around 60 to 100 watts worth).

    I'm hopeful that the broad stern and flat aft sections will aid stability without causing too much of a drag penalty from the greater wetted area, although I'm tempted to put just a bit more curvature into the aft sections. The transom would be just clear of the water, for low drag. DL is around 56 and SL is about 1.05, so it looks like I'm in the right ball park. I've estimated that resistance from the hull viscous drag at 4mph should be around 30N or so, and have assumed that wave making resistance will be negligible, as I'm below hull speed by a reasonable margin (LWL is about 15ft 6ins).

    This is a photo of the electric propulsion unit, before I fitted a faired leg and skeg to the lower section. The motor at the top is a TowerPro 5330-10t, that cost just $49, the controller I'm using cost $22 (most of the money will go into the battery pack and solar panels).


    I've looked at lots of designs before reaching the point of posting here, including the Bolger Sneakeasy, some sleek, fast row boat designs, all the iterations of Rick Willoughby's human powered boats (and others) as well as some of the work of proponents of the "long, narrow" school of thought, like Nigel Irens.

    Although I'm reasonably practical, and have a background in aircraft design so understand the drag and power issues reasonably well, I am a complete novice when it comes to boat design so would really appreciate any thoughts people may have. Please feel free to call me crazy or offer any sort of critical comment............

    1 person likes this.
  2. BeauVrolyk
    Joined: Apr 2009
    Posts: 160
    Likes: 10, Points: 0, Legacy Rep: 153
    Location: San Francisco, CA

    BeauVrolyk Sailor

    This looks like a fun project. A few ideas.

    - I'm not certain how you're going to get the power from the motor to the prop, but the example has a 90 degree turn in it, you'll loose a lot of power in the turn so you might want to consider a straight shaft.

    - A difficulty I've had with electrics are the weight of the batteries. I'd suggest making it extremely easy to get them out of the boat for two reasons. 1) To put them in the boot of the car, rather than lifting them up on the roof. 2) To be able to take 'em home and charge them from the wall outlet.

    - When considering the hull shape, you might want to have a very narrow waterline with it rapidly expanding above the waterline, either aft as you've shown in your model or amidships. That way, when the boat is straight up you'll have low drag. Hopefully, this is most of the time. Yet, when the boat starts to tip it will pick up stability fast. Long overhanging counter with wide hips, like the model you're playing with, will do that, but they sometimes get their stability rather suddenly, when the boat puts the hip in the water. It's a bit easier on the psyche of the crew to have it more progressive than that by having flare amidships.

    - You might want to use a laser hull as a mule for what you're doing, very easily driven. It actually has many of the characteristics you're looking for without you having to build anything. Once everything is debugged you could build your nice wood hull to put it all into.

    1 person likes this.
  3. Guest625101138

    Guest625101138 Previous Member

    You may be interested in the electrified version of V11J (give it time to load):

    I got 5.6kts with a small PMSM motor running at 200W.

    If you have not seen the faux-tri concept I have been working on for a larger solar-wind boat then it is worth considering. The faux-tri will respond much better to increased power than the stable mono.

    If you build in wood it will be a challenge to be car toppable with all things on board.
  4. Willallison
    Joined: Oct 2001
    Posts: 3,590
    Likes: 130, Points: 0, Legacy Rep: 2369
    Location: Australia

    Willallison Senior Member

  5. Guest625101138

    Guest625101138 Previous Member

    The lowest drag hull for 4kts displacing 200kg will require 33W on the hull. You could double this power at the battery but still below your 100W.

    I have just ordered a 5Ah 22V lithium battery. It cost USD76. You would need two of these to get your three hours. They weigh about 800g.

    The other option is to get some of the large format cells like this:
    The Thundersky have mixed reports but if one cell goes you can replace it. You would only need 3 or 4 cells with that motor. The 40Ah would give your running time.
  6. Guest625101138

    Guest625101138 Previous Member

    Attached is the latest evolution of the faux-tri. I envisage a scaled down version of the hull but with forward facing arm chair aft of centre. Table in the middle with batteries under and aft facing arm chair forward of centre. The boat having a beam of 1.2m. Solar panel on fore and aft deck. The chairs could be used sideways as well but you would need a person on either side of the hull to reduce roll.

    A trailer would be more gentlemanly than rooftopping.

    Rick W

    Attached Files:

  7. Guest625101138

    Guest625101138 Previous Member

    Attached is a 20ft version I did some time ago. I would do it a little differently now. A boat designed for 4kts would be around 4.5m long.

    Attached Files:

  8. Jeremy Harris
    Joined: Jun 2009
    Posts: 978
    Likes: 60, Points: 28, Legacy Rep: 711
    Location: Salisbury, UK

    Jeremy Harris Senior Member

    Very many thanks for all the tips and advice, much appreciated.

    The electrical stuff and propulsion I'm reasonably happy with, as I've a fair bit of experience with electric vehicles. I had about three goes at making the electric drive system, trying to minimise losses as far as possible. The first thing I found as that the 3:1 HTD belt drive reduction that I first built soaked up about 20 watts with no load. I then opted to modify the motor, by re-wiring it from delta to star to reduce it's kV (rpm per volt) to a level where direct drive might be possible. This worked very well, but using the standard RC motor speed controller, which relies on back EMF feedback from the motor to stay in sync, meant that the low speed torque and efficiency were poor. The solution to this was to fit Hall position sensors inside the motor and change to a controller designed to operate well at low speeds. The controller I'm now using is intended for an electric bike (I've built one of these in the past, too) and has the virtues of being cheap ($22), programmable via a PC, operable at voltages up to around 60V (reduces I²R losses) and is sealed in a nice alloy case.

    The gearbox losses are very small, between 1 and 3 watts as far as I have been able to tell. It has a small reduction ratio of 1.25:1, which helps a bit, too. The biggest losses come from the seal on the prop shaft. I've reduced these as far as possible by polishing the shaft, but I'm still losing about 5 watts at the moment (this may get better as the seal beds in).

    I looked at direct drive via a conventional shaft, and at the flexible shaft systems that you've used, Rick (thanks for posting all that useful stuff, BTW, it's a goldmine for people like me!). I really need to be able to lift the prop out of the water to clear weed and allow beaching, something that's not impossible to engineer with a long conventional shaft, but not too easy. The drive leg system will fit into a hinged box set into a case like a centre board, so that it can be pivoted up.

    I've thought long and hard about the weight and concluded that I need to be able to keep the hull light and make things like seats, the motor and battery removable. I'm looking at cheating and buying some removable canoe seats like these: http://www.edscanoe.com/clfoch.html as I don't think I could build them for that price, even accounting for shipping to the UK. I'd fix these with some form of quick release. The battery is very light. My current electric bike battery pack (a 36V, 10Ah one, that will easily deliver a true 360 watt hours plus) weighs just 4kg, including the home made composite box I made for it.

    That box contains the LiFePO4 cells, plus a battery management system circuit that ensure that the cells cannot be damaged from over-discharge. The BMS also controls the charge process and balances each cell to ensure maximum life. At 4kg and just a little larger than a standard house brick it'd be easy enough to lift in and out of the boat, and if I can get an average of 30 to 40 watts from solar cells it might be enough to give me the 8 hours endurance I'm looking for.

    Thanks very much for the tips on hull form, this was just what I was looking for. I'd already thought about changing the cross section from a flat bottomed "U" into more of a shallow curve, but had wondered what this might do to stability. I think I will make another model, using the same 1/32" ply (which is a scale thickness of 1/4", so too thick really) and concentrate of developing a panel shape that is optimised for this hull form, rather than a canoe. In particular, I'd like to get the transition at the aft waterline a bit smoother, by putting more curvature into the run up to the transom.

    That Nigel Irens design is very similar to what I have in mind, although I am aiming for something much lighter and a bit less than half the LOA/beam of the Elektra. If I can get the empty hull weight (less seats, motor and battery) down to around 25 to 30kg then I think I should be OK. The hull skin itself only uses two sheets of 4mm ply, so allowing for the glass/epoxy weight should come in at around 16kg or so. That gives me around 10kg to play with for the decking and solar cells, which I think should be OK. The solar cells I'm looking at are the resin encapsulated crystalline ones, that come bonded to a fibreglass substrate. I think I can probably get away with using very thin ply for the fore and aft decks and gain some stiffness by bonding the solar panels down with epoxy (might be a problem if one needs replacing!).

    I like your trimaran design, Rick, but would prefer something more traditional looking, purely from the aesthetic standpoint and to be in keeping with it's intended use (rivers and inland waterways where traditional craft abound). I would like to be able to show that it's possible to have a sustainable (i.e. "free" energy) craft that doesn't look that much different from other small watercraft.

    Thanks again for the useful advice, I'll post an update on progress soon.

  9. Guest625101138

    Guest625101138 Previous Member

    Seems you are a wunderkind with the electrics. I have a Mars motor with the Hall sensors and consider it overkill for the boat application. You do not need fine speed control because the prop slip at low rpm is quite high. Anyhow you have a nice set up.

    I have found the curved spring steel shaft to be very good. I have used gearboxes under water and they work fine but have more drag than the curved shaft. You could use a curved shaft with a prop and rudder mounted just behind the transom with the ability to lift up but what you have should be fine.

    I have been surprised at how little tension the tooth belts need to operate reliably. I am using a loose fit to reduce the no load loss on the belt and all shafts have roller bearings. The 3:1 set up on the 600W motor I am playing with ticked over with 2W. It went up to 6W with the 12:1 reduction but the final stage is way bigger than needed. The big pulleys are for another application.

    When you have a picture of your next model I will sketch it up and do some drag checks for you. The two things you need to determine with some accuracy is the total displacement loaded and the required initial stability. You probably need about 900mm wide to get acceptable stability.

    If you know the diameter, pitch and blade width for your prop I can give an estimate of its performance as well. You would probably do better with something like a 300 X 300mm model airplane prop. The force on the prop will be very low at your intended power level. These cost a few of dollars. Something like this:
    will achieve around 85% efficiency at 100W and 88% if the power required was only 40W. As a matter of interest it will need to spin at a tad over 400rpm to achieve the target speed.
    1 person likes this.
  10. Jeremy Harris
    Joined: Jun 2009
    Posts: 978
    Likes: 60, Points: 28, Legacy Rep: 711
    Location: Salisbury, UK

    Jeremy Harris Senior Member

    Not sure I'm really any sort of expert with the electrics, it's just that I've learned the hard way what seems to work and not work! I did spend a lot of time looking at these RC outrunner motors, mainly because I was fascinated with the combination of power, performance, size, quality, weight and price they seemed to offer.

    Thanks for the tip about the belt tightness. I was using a 15mm wide, 5mm pitch HTD belt and running it quite tightly, so that's probably where my losses were coming from. It looks from your video as if you've opted for a 9mm belt, maybe with a smaller pitch, which might be why your losses were lower. I may take another look at making a belt drive if I need to spin the motor faster for better efficiency.

    I've also got a Mars motor on another project, a converted motorcycle. It's a big and heavy motor for the power though, especially when compared with some of the bigger RC motors. The big Turnigy 80-100 is a good example, 6500 watts from a motor that still (just) fits in the palm of the hand, and only costs around $150. Knowing what I know now I wouldn't have gone for the big Mars, but would have probably just used one, or maybe two, of these big RC motors. The weight and space saving would be massive and the cost wouldn't be much different, in fact it might have been cheaper.

    Controllers were the big problem for these large RC motors, as the model people don't seem to use high voltage systems. Luckily, several people over on the Endless Sphere forum have been playing with controller kits available from China that are intended for electric bikes. The biggest of these will deliver around 100 to 150 amps at up to 100 volts, so is more than adequate to drive a big brushless motor. Even the tiny $22 controller that I'm currently playing with can be easily tweaked up to around 50 amps at 60 volts or more.

    The snag with using an electric bike controller with these RC motors is the need to modify the motor internally to fit Hall sensors. It's not a difficult job, but is a little fiddly, as it's easy to damage the motor. Here's a photo of one of the Hall sensors fitted inside a Towerpro 5330 motor:


    There are three of these, fitted at 120 degrees around the stator, into slightly widened slots (delicate work with a Dremel!). They provide position feedback to the motor controller, which allows very low speed (down to less than 50rpm) operation with very high torque.

    Here's a shot of my test set up on the bench, with the cover off the small electric bike speed controller:


    For scale, that motor is about 62mm in diameter.

    The Towerpro motor I'm using has a Kv of 215 as standard, but I found that it was wired internally as a delta three phase motor. By pulling the internal wires apart, I was able to re-connect the windings into star configuration, which lowered the Kv to about 124. This is much better, but still not really low enough to drive a really efficient prop such as the one you linked to, Rick. The prop I have at the moment is an 8" alloy two blade one that I've had in my "may come in handy" box for at least 20 years. It originally came from a small, 2.5hp "mud motor", I think. As far as I can tell, the pitch seems to be around 7" or 8", based on measuring the blade angle at the 66% radius. I think this should work OK at around 1000 to 1200rpm, but probably only with around 65 to 70% efficiency if I'm lucky. It is nice and tough, though, and has a slightly swept back blade profile (probably for better weed clearance) so it may be that I just accept the efficiency penalty in return for a more robust/practical solution. Time will tell! The blade width is quite wide, and approximately elliptical in profile. At the widest point the blade chord is about 2", but it is reasonably thin, perhaps 1/4" at the root centre, tapering to about 1/16" at the edges and end. Here's a better photo of the prop, before I cleaned all the chipped red paint off and polished it:


    I like the idea of the curved shaft and may well revisit it. As you say, it could be made to pull up clear of the stern quite easily for weed clearance, etc. The gearbox I'm playing with is a bit of an unknown. I've read the dire warnings about drill right angle drives, but suspect these were aimed at the plastic ones common on eBay. I found a nice alloy drill right angle drive here: http://www.transtools.co.uk/store/prod_8708/power-tool-accessories/adaptors-and-chucks/faithfull-keyless-chuck-right-angle-drill-attachment.html for just under £20 (about $32 US). When I took it apart I was pleasantly surprised to find that is had decent ball races and nicely made helical bevel gears with a 1.25:1 reduction ratio. I ended up with a "free" keyless chuck, as well! I'm hopefull that this will cope OK with the rigours of handling around 50 to 100 watts for a sustained period, I can't see any obvious reason why it shouldn't, particularly as it's now virtually immersed in oil internally.

    Thanks for the kind offer re: the hull modelling, I may well take you up on it. If I can repay the kindness with any electrical help, just ask. I should get the new model built in the next day or two, provided I don't get distracted,

    2 people like this.
  11. mydauphin
    Joined: Apr 2007
    Posts: 2,161
    Likes: 53, Points: 48, Legacy Rep: 575
    Location: Florida

    mydauphin Senior Member

    I would get away from 90 degree gear, you are losing power there. A straight shaft is much better.
  12. Guest625101138

    Guest625101138 Previous Member

    It is great that you came along. Just your shared interest on these motors and controllers buoys my interest.

    I have been playing with Kelly controllers for the Mars and it all looks good.

    This clip shows the original set up with the 3:1 reduction before I added the 4:1:
    This is the set up I intend to use in my 1/5th scale model of the solar wind boat. I have a 10W solar panel and am quite impressed how it powers the motor. The belt is 6mm wide with 2mm pitch. I would go for larger pitch if I do this again.

    I expect your gearbox will hang in if it stays water tight. One of the things we are doing with underwater boxes is to ensure the box has a positive oil head so that oil will leak out rather than water in if the seals fail. I have found the curve shaft hard to beat when using the high aspect props. The vibration is annoying if you run an inclined shaft.

    Your existing prop should get around 69% efficiency at 100W at the hull. It will increase to 79% at 40W on the hull.
  13. Guest625101138

    Guest625101138 Previous Member

    The set up I have with the belt drive would be quite adequate to fit to a curved shaft. It could be just a length of 8mm aluminium round bar. The torque is not very much. That motor is capable of 600W and took 200W easily. Some of them are up around 93% efficiency.

    I get most of this stuff from Hobbycity. Their prices are good and fast service.
  14. Jeremy Harris
    Joined: Jun 2009
    Posts: 978
    Likes: 60, Points: 28, Legacy Rep: 711
    Location: Salisbury, UK

    Jeremy Harris Senior Member

    Thanks very much for that info on the prop, Rick, it looks good enough to go with as a first iteration, maybe I can fine tune it if the concept works OK.

    Your belt drive is much smaller and neater than mine, I used what I had available, so went with an 18 tooth motor pulley and 60 tooth driven pulley, using a 15mm width, 5mm pitch HTD belt. I also used some bearing blocks that I had which are fitted with sealed bearings, which have a lot of drag and soak up power.

    The gearbox is oil filled for just the reason you give, to stop water leaking in. I've measured the power loss on it, without the lip seal fitted, and found it to be pretty low. Best guess is around 1 to 3 watts being lost, but as I can't load the gears up properly this might be a bit optimistic. One snag with partially filling the leg with oil is remembering not to turn it upside down............... (I don't want the added drag from proper seals at the top).

    The Kelly seems to have a good reputation, and is certainly the best value at the moment, as far as "big" controllers go. Is your Mars motor a brushed or brushless one? The one in my motorcycle is an ME0709 brushed motor, which was chosen because it's low Kv allowed the use of the existing rear wheel sprocket. I'm using an Alltrax controller with that, but it's big and heavy though, just like the motor.

    The controller that I'm using to drive my Towerpro RC brushless motor is available on eBay, here: http://cgi.ebay.com/48V-350W-brushless-controller-for-E-bike-scooter_W0QQitemZ260323431675QQcmdZViewItemQQptZLH_DefaultDomain_0?hash=item3c9c7c54fb&_trksid=p4634.c0.m14&_trkparms=|301%3A1|293%3A1|294%3A30 and is easy to modify for increased performance, if needed. The same vendor, Keywin Ge, can supply kits of parts to build bigger controllers, capable of handling in excess of 10kW, for a very reasonable price. I purchased two of these big controller kits, without the power FETs, for $130 US, delivered to the UK. Even allowing for the high cost of the FETs, each finished controller will have only cost me about $90 to $100 US each, so still something of a bargain. There is an interesting thread on the Endless Sphere forum discussing these controllers, although it's dominated by a chap that's building custom units for those who don't want to do the DIY thing: http://endless-sphere.com/forums/viewtopic.php?f=2&t=10312

    For small electric boat use, the off-the-shelf brushless controllers from that eBay vendor are fine. As well as the small 350 watt controller (which can easily be modified to deliver over 1000 watts safely) he has some bigger units in his store, capable of running much higher power levels. The real thing that stops these being ideal as an off-the-shelf solution for driving big RC motors is the need to fit Hall sensors to the motor. The upside of using them is that they interface to the real world in a nice way, just like the Kelly. You can use a potentiometer or 0 to 5V signal for the throttle, they have built in emergency stop connections and a reverse switch, plus they can be programmed fairly easily using a PC and a serial cable. They also have the ability to regenerate, using the motor to charge the battery if the boat was moored in a fast-flowing current. They are certainly much easy to drive than the RC speed controllers, that need a servo tester to be hooked up as a throttle and don't normally have a built in reverse capability.

    I could probably write an illustrated guide to modifying an RC outrunner motor to work with these cheap controllers if it would be of interest.


  15. Guest625101138

    Guest625101138 Previous Member

    This is where I get my timing belts:
    They give very good service. I expect a bit far for you but you can get an idea of prices here and they have good design information.

    The controller I purchased for the Turnigy SK motor has continuous rating of 60A:
    It gives surprisingly good speed control for open loop operation. The safe voltage is under 30V going by the feedback.

    My Mars motors are PMSMs. I tried the low cost open loop controller with one but it just cogged. I have not tried to program these controllers. I do not know how clever they are but they seem to learn. They do not have inbuilt reversing. If I change the phases it takes a while for the motor to run smoothly at low rpm so there is something going on. I would not bother with hall sensors after using these controllers. Boat applications are quite different to road use for speed control.

    Another place I get nice stuff from is Smallparts:
    They have operations in US. I have not checked for UK. You can get neat little bearings - stainless with flanges for example.

    I want to try to make a generator with my Turnigy SK as well. I am currently making wind turbine blades. I will test them on a mechanical drive with underwater gearbox for a bit of fun. It is something I have intended to do for a couple of years just have not got around to it. I did get close but damaged the last set of blades I made while playing around in strong wind. I could not stop the turbine so ended up running the blades into a bush in the back yard. I will make sure I have a proper method of support before I test the new blades.

    You may find a small wind turbine better value than solar panels. I intend to use both for energy collection. Something like this:
    Or even smaller. They have blades as well. The outrunners would enable very low weight generator.
Forum posts represent the experience, opinion, and view of individual users. Boat Design Net does not necessarily endorse nor share the view of each individual post.
When making potentially dangerous or financial decisions, always employ and consult appropriate professionals. Your circumstances or experience may be different.