Most efficent propulsion?

[Drunkskunk]

What is the most efficent form of motored propulsion for a low speed, light weight boat?

I've been planning an E-Kayak for years, but have limited time, and each year my plan gets revised. Its still just a dream. But one of the things I've been wondering about is the drive.

A trolling motor is the easiest answer. its an electric drive designed for exactly what we need. But the efficancy is way low. And I'm not convenced a spinning prop has the potential to ever be efficent.

I had a friend who was the 2nd owner of what had once been the worlds fastest human powered boat. It was a unique design paddle wheel that folded it's paddles flat for best wind resistance, and moved the padles through the water horizontally so they maintained the same speed regardless of depth. That boat flew over the water with just the two of us pedeling.
The designers had chosen an 8 foot tall paddle over a prop for a reason, and had suceeded.

But that can't be the only or best answer. Surely there has to be a better way

In order for a propultion device to be efficent, it needs to transfer energy to the water without moving the water. You want the boat to move in the water, you don't want the water to move. The more you shove, move, or force water in a new direction, the less you move the boat.


So my theory is that Large, slow moving blades are needed, moving in a linear a fassion along the axis of movement. I think that there is a relationship between the blade's surface area and the thrust needed to move the boat that affects efficancy, though I have no idea how to calculate that.


So any thoughts on this?

[Jeremy Harris]

There was a debate on this in the long running pedal boat thread on the Boat Design forum. A very large diameter paddle wheel, turning slowly (around 5ft or so in diameter) can be as efficient as a large diameter, slow turning, propeller, particularly if the paddle wheel has articulated blades.

The paddle wheel has some distinct advantages, like the ability to operate in shallow water, relative freedom from weed entanglement, ability to deal reasonably well with mud etc. A large diameter, slow turning, prop needs long, narrow blades to work well, and probably has a slight edge on a big paddle wheel in terms of efficiency (although both could be over 80% efficient if carefully designed).

Big props have problems, though, as they increase draught and tend to get fouled up with weed. Some have suggested that a snomobile type track, with attached paddles, could be as efficient as a big paddle wheel, but the frictional losses in the track drive create a problem.

As far as the boat design goes, then the lighter the better, as resistance (and hence power) is directly proportional to wetted area, and wetted area is largely a function of total displacement. The length/breadth ratio is also important, canoes being more easily driven than fat row boats, for example. Generally you want the length to be at least 5 times the breadth for an easily driven hull. Hull cross section makes a small difference, but as the length/breadth ratio increases this becomes less critical, and for a long, narrow, boat a flat bottom and vertical sides doesn't give much more resistance than a rounded hull.

[rolf_w]

In order for a propultion device to be efficent, it needs to transfer energy to the water without moving the water. You want the boat to move in the water, you don't want the water to move

The physical principal to accelerate and move a body / mass is Newton's law of the conservation of momentum: you have to move an external mass with the oposite momentum to move your boat. If the external mass is the water surrounding the boat you have to move the water backwards. If you move very much water you can do it very slowly, if you move little water you have to do it fast. You can't avoid moving water! The trick is to do it efficiently, with little losses (=turbulances) and of course to require as little momentum as possible i.e. reduce the friction of the moving boat to a minimum. This principal is very fundamental: if your boat is a jet plane, you have to move air backwards, if it is a spaceship you have to shoot fuel out of the ship and if it is a car? well then the oppisite momentum is from the entire planet earth - by accelerating the car your tires accelerate the road and the attached planet in the opposite direction (of course unmeasurably little since earth's mass is so large). If you have too much slip (burning tires) you do it inefficiently, same as too much turbulances in the water.

The best human built machines to transfer motion of a fluid to a mechanical motion e.g. a rotating shaft are hydro power turbines. A larger turbine with less than 93% efficiency is not state of the art. There are turbines which can operated in both directions: as pumps or turbines (though at lower efficiency). Well designed pumps are similarly efficient to turbines. I think it should be possible to use a pump (=ducted propellor) to move water, and thus the boat, quite efficiently, r

[Jeremy Harris]

Unfortunately pump jets (and I have experience of working with them, as I worked on our Sting Ray torpedo programme for a few years in the late 70's/early 80's, it has a propulsor similar to that on the US Mk50) are pretty inefficient propulsion systems for surface vessels. The problems are twofold:

- The low inlet dynamic and static pressure (essentially just the pressure arising from the vessel forward motion and the depth of immersion) limits the acceptable blade loading before the onset of cavitation, which restricts the pressure ratio across the blades. This in turn restricts the outlet velocity and hence the mass flow for a given diameter. The restricted mass flow restricts the thrust. It's not as big a problem on submarines and torpedoes, as they operate in deeper water and higher speeds, so both the dynamic inlet pressure and the static pressure from depth are significantly greater. Similarly it's not a problem for turbines and turbopumps operating with closed pipe inlets, as they will have a significantly higher inlet dynamic pressure (from the higher inlet flow velocity).

- The second problem is duct drag (if the pumpjet is external) or duct flow resistance (if the pumpjet is internal). This pretty much cancels out the efficiency gain that comes from shrouding the blades to reduce or eliminate tip vortex shedding at anything over a few knots.

Pumpjets are a good choice where you want shallow water capabilty and are willing to sacrifice efficieny, or where you want high performance (and hence high propulsor power absorption capability) in a limited space. They are also a good choice if you want acoustically quiet propulsion and wish to mask the tell-tale blade passing frequency of a conventional propeller (something only of merit in warfare, really).

For a slow speed boat, then the best overall efficiency would be achieved by a large diameter, high blade aspect ratio prop, turning at low speed, and there is a large body of evidence from the HPB enthusiasts to prove this (they've tried just about every drive system imaginable). With careful optimisation of the propeller design you might get this to work at around 85% efficiency. I believe that you might be able to get a large diameter, articulated blade, paddle wheel up to around 80% efficiency. You can also get up to around 80% efficiency (maybe more) with a large diameter, slow turning air propeller. Mark Drela's Decavitator showed this back in the 90's. A smaller diameter ducted in-water propeller, with square tips, operating inside a close-fitting Rice-type accelerating duct could also get you up to around 80% efficiency or so, but, based on my own practical experience of using a small electric boat with such a system I have to say it isn't practical, as it sucks up weed, plastic bags, bits of fishing line etc like a vacuum cleaner, and so gets fouled up very easily.

For comparison, the small propellers fitted to small outboards, trolling motors and the like, are generally around 45 to 55% efficient. It's pretty rare to find a conventional prop on a small boat that does better than about 60%. The reasons for this are largely practical, as large diameter propellers are a nuisance (they increase draught and are more prone to damage from hitting things) and long narrow blades tend to get tangled up with weed and debris. Making the propeller smaller in diameter, with swept back blades with a wide chord, reduces efficiency a fair bit but makes the propeller stronger and less likely to get fouled.

[Xrain]

You can get a significant efficiency increase by using a forward facing propeller. This system is used on the azipods that are frequently used on large transport ships and the like. I've heard of 20% efficiency gains on a forward facing prop vs. a similarly sized rearward facing. The reason for the increased efficiency is due to the the prop now being able to act on non-turbulent water, using the full surface area of the propeller, vs. the rear facing props.

This would allow you to use a smaller prop for a similar efficiency level, with reduced risk of prop fouling. One negative is however the prop is more exposed and can more easily get damaged. However since most forward facing props assemblies are capable of rotating 360 degrees, you could simply reverse the prop when you feel you are at risk of damaging it.


The paddle wheel idea is interesting, however I would be concerned with weight, especially if you have a articulated paddle wheel. So the overall efficiency gain might not be as high as it might appear.

The snowmobile track absolutely would not be as efficient, I have worked with snowmobile tracks a fair amount, you can expect upwards of a 30% power loss in the drive-train simply from the frictional losses in turning the track. and I imagine only the leading edge of the track entering the water would be working at optimal efficiency, with all of the rest of the track suffering losses due to the turbulent water. You could certainty go fast with a snowmobile track, but I wouldn't use one to save on gas.


I would be interested if you could develop a low speed hydrofoil to partially lift the boat out of the water. I imagine that large reduction in surface area might well give you a notable increase in efficiency.

[Jeremy Harris]

The efficiency gain from the Azipod system isn't primarily from improved prop efficiency, it's specific to the geometry of these large pods and the severe flow disturbance they would create if they operated the other way around.

Forward facing props are less efficient that aft facing ones in clean flow, because the accelerated flow velocity behind the prop increases skin friction (viscous drag) by flowing at a rate faster than boat speed over the pod and hull aft of the prop. Viscous drag makes up the vast majority of total boat resistance at speed below the onset of significant wavemaking, so it is important to reduce this to a minimum.

In the specific case of big ship electric propulsion pods, like the Azipod system, that use large diameter housings containing big motors, the efficiency loss from this higher viscous drag over the pod and supporting fin from the prop outflow is less than the loss they would get from disturbed inlet flow if they faced the other way around. It's a specific case applicable only to pods where the diameter of the prop is small compared to the diameter of the pod.

If you have an aft facing, "pusher" prop, operating with a relatively undisturbed inlet flow, then it will always be more efficient than a forward facing arrangement where the accelerated flow from the prop passes over part of the hull.

I agree about the losses in a snomobile type drive, they'd be far too high to make it practical.

Watch this space with regard to a semi-hydrofoil, as I have a part built foil that I intend testing on the aft end of my boat in the next couple of weeks, the intention being to lift it up and reduce viscous drag!

[Honk]

Hi again.

Iv'e been thinking some about power input vs force created.
In Javaprop I designed a 74% efficient prop at 0.3m dia 2-blade for 1100 RPM and 270Watts shaft power.
It was calculated to generate a force of 99N (approx 22 lbf).

The thing is....I have tested the force of a 2-blade plastic Motor Guide prop at 270W and it
generated 40 static lbf, almost twice the force of the calculated prop at 22 lbf and this is very strange to
me as a good and efficient prop should convert energy to a lot of thrust to push the hull efficiently forward.

Efficiency for me is maximum thrust being generated, but 22lbf is a lot less than actually measured 40lbf
from a real prop. Perhaps the thrust calculation in Javaprop is way of from real world results....I don't know?

Do you have any input on this Jeremy?

[Jeremy Harris]

Hi again.

Iv'e been thinking some about power input vs force created.
In Javaprop I designed a 74% efficient prop at 0.3m dia 2-blade for 1100 RPM and 270Watts shaft power.
It was calculated to generate a force of 99N (approx 22 lbf).

The thing is....I have tested the force of a 2-blade plastic Motor Guide prop at 270W and it
generated 40 static lbf, almost twice the force of the calculated prop at 22 lbf and this is very strange to
me as a good and efficient prop should convert energy to a lot of thrust to push the hull efficiently forward.

Efficiency for me is maximum thrust being generated, but 22lbf is a lot less than actually measured 40lbf
from a real prop. Perhaps the thrust calculation in Javaprop is way of from real world results....I don't know?

Do you have any input on this Jeremy?

Propeller thrust depends on speed through the water, and will decrease as speed increases (assuming a fixed pitch propeller at a constant rpm). You can get a very high static thrust with a fine pitch prop (the useless selling point that all the trolling motor companies seem to advertise) but this won't translate into anything like the same thrust at speed.

When a prop is delivering a high static thrust it's efficiency is zero, because it isn't doing any useful work (it's not moving the boat forward). This means that maximum thrust, on its own, doesn't indicate efficiency. Efficiency is the maximum thrust at a given boat speed for a given power input. Vary any of these and you can't make a valid comparison.

Javaprop gives results that are generally within about 5% at worst from test-validated results. In my experience the error is usually less than this, 2% being fairly typical.

As an illustration, the prop on my Cordless Canoe Challenge boat was (optimistically!) designed to run at best efficiency at 4.5 m/S (about 8.7 kts), and to absorb 300 W of power at 900 rpm at this speed, delivering about 57 N of thrust. When run at the same 900 rpm but at a boat speed of just 1 m/S (about 1.9kts) it would give about 1860 N of thrust, but would absorb a massive 6350 W of power to do so and would only be running at around 29% efficiency. It turns out that my optimism with regard to the speed potential of the boat caused the motor to absorb far more power than I planned, as I reckon it used 54 Wh of battery in less than 4 minutes, so around 800 W, rather than the 300 W I had hoped for! The reason is that the boat really ran at around 3.5 m/S, so the prop was absorbing a lot more power than planned at the design rpm.

[JennyB]

Watch this space with regard to a semi-hydrofoil, as I have a part built foil that I intend testing on the aft end of my boat in the next couple of weeks, the intention being to lift it up and reduce viscous drag!

How about this?

[The fingers]

One idea to reduce drag, air bubbles!

[Honk]

Thanks Jeremy.
You helped me understand the numbers in Javaprop a lot better now and that I shouldn't test any
prop in static mode as this reveals nothing of the actual performance at nominal speed.

Do you think I have over-estimated my low speed prop design in JavaProp.
I figured it would push an easily propelled aluminumboat (approx 100kg heavy) at 2m/s consuming 270W input being 73.35% efficient at 1100 rpm.

Please see the attached Javaprop saved file and screen dump.

[Jeremy Harris]

The best starting point is to try and determine the speed vs resistance characteristic for your boat, then match the prop thrust to your desired cruising speed. For example, the plot below shows the total hull resistance (or drag) and the power requirement to overcome it, for my Cordless Canoe Challenge boat.

Hull resistance.jpg

(148.55 KiB) Downloaded 3 times

I modelled this using a free hull resistance modelling tool called Michlet, which isn't that easy to use, unfortunately, and requires knowledge of the hull lines and dimensions. You can get a rough approximation of the resistance at low speed (below the speed given by the S/L ratio x square root of the waterline length (in feet), with the speed in knots) by determining the wetted area of the hull and calculating the viscous drag from the water. The S/L ratio is given by the boat speed (in knots) divided by the square root of the waterline length (in feet) and is also known as the Froude number (sorry for the imperial units - naval architecture is still wedded to using them!).

The formula for calculating viscous drag (in N) is:

Hull total wetted area (m²) x drag coefficient x water density (1000 kg/m³) / 2 x V² (m/S)

where V is the boat speed

The drag coefficient can be approximated using this formula:

drag coefficient = (1.1 + V/100) x 4.27 / ((Logn(Re) - 0.407)^2.64)

where Re is the Reynolds Number, given by:

waterline length (guessed at 4.5m) x speed (m/S) x kinematic viscosity of water (1,000,000 s/m²)

So, if you have a hull with a wetted area of, say, 5 m² (this is just a guess - you'd need to try and get some sort of measurement of it) with the maximum load you want to carry, then at 2m/S the resistance that the motor thrust needs to overcome would be about 5 x 0.0034 x 1000 / 2 x 2² = 34 N

Sorry for all the math, there's no easy way around it if you want to try and get a good match between the boat characteristics and the propeller.

[Honk]

Thankyou very much again, Jeremy.
Now I have some thinking to do.

I'll be in touch.....

[Dee Jay]

how 'bout the way golf balls are designed with dimples, except instead of air turbulance it's water

One idea to reduce drag, air bubbles!

imagine if you can pump air bubbles from pores on the hull, bubbles could lift the boat slightly off the surface of the water like a hovercraft!

well, maybe...
http://access.aasd.k12.wi.us/wp/baslerd ... s-mileage/

[Honk]

Hehehe, Mythbusters did a test on a car with "golf ball" cavities and it did indeed get somewhat less drag.
Perhaps I should go and knock a lot of smooth circled cavities into my brand new aluminium boat as soon as I go get it....

Joking apart ...I have actually been thinking a bit about lubricating the hull by a dense air cushion.
It's being tested at this moment on big vessels with great sucess, and perhaps it could be helpful on smaller boats as well if properly designed.
Practically it could be made from some kind of "strap on" tube with tiny holes going below the bow and then it should be run by a
efficient and silent low power air pump, no more than 12W usage.... Well, as I sad...Perhaps...

[Jeremy Harris]

Bubble "lubrication" works, primarily by reducing viscous drag (it effectively dramatically reduces wetted area). The snag is the power need to get a layer of bubbles under the hull, as they need to be at a fairly high pressure to work. Low pressure bubbles have the opposite effect and cause the vessel to sit lower in the water. It's how some old torpedo and depth charge explosions used to sink ships, they'd great a large gas bubble under the centre of the hull, leaving it unsupported by the water, so the ship would then only be supported at the ends and would break in half. The same effect has caused big ships to sink when large undersea gas deposits have been suddenly released.

The golf ball dimpling effect does work if you can get it right. Some fast swimming fish (sharks, for example) use the surface roughness of their skin to reduce drag by creating a thin turbulent flow layer that effectively "lubricates" the area immediately around them, reducing the energy they need to expend on swimming. Dolphins go one better, by having semi-active control of surface flow over their bodies. They have flexible skin, with an elastic layer underneath it that deforms with local dynamic pressure. This acts to constantly move the skin surface in and out as they swim, optimising local flow conditions at the micro level and reducing the energy they use when swimming fast.

[edited for typo]

[The fingers]

"Whitewater" being mostly air, will not float you but you can't breathe it. Surfers usually find this out the hard way!

[Drunkskunk]

The textured surface idea works. the boat I mentioned in my first post was painted with a non glossy paint. I'm not sure what sort of micro surface it might have had, it was 80's tech, but that reportadly gave it a few seconds faster time. My friend promptly painted it glossy blue.

Making the paddlewheel efficent I can understand, as I've seen one in action. The propeller I can't. I can know that it is because people who know said so, but I don't understand it, something hasn't clicked in my head yet.

How can a prop reach that efficancy while it has to travel so far through the water? Since the tips of the blades have to move several feet through the water (and displace that water) for every inch forward the boat moves, how can they be efficent? I don't dispute it, I just don't understand it yet.

[The fingers]

Can a type of oil or grease make the hull slide through water with less drag without polluting the lake or violating the rules? I have used vegetable shortening to waterproof my work boots when I went out in wet conditions, and the traction soles were dangerously slippery upon returning to smooth surfaces. It needed to be reapplied often to remain effective; but it is edible, inexpensive, practically odorless (unless rancid), and readily available.

[spinningmagnets]

As outrageous as a kayak hydrofoil may sound, I believe an ultra-light carbon-fiber version might be well worth some consideration. Of course there would have to be a certain minimum speed to maintain elevation above the water.

An inverted V-wing seems as though it would be the easiest to experiment with, and I also think a V-hull would be optimum (Rinspeed Splash shown below in pic). As you begin accelerating, every inch of elevation you attain would remove more of a V-hull out of the water. As to what type of propeller (pedal-driven?), I have no idea what would be best.

Jeremy, I am intrigued by your efforts using a semi-hydrofoil...

[Jeremy Harris]

The textured surface idea works. the boat I mentioned in my first post was painted with a non glossy paint. I'm not sure what sort of micro surface it might have had, it was 80's tech, but that reportadly gave it a few seconds faster time. My friend promptly painted it glossy blue.

Making the paddlewheel efficent I can understand, as I've seen one in action. The propeller I can't. I can know that it is because people who know said so, but I don't understand it, something hasn't clicked in my head yet.

How can a prop reach that efficancy while it has to travel so far through the water? Since the tips of the blades have to move several feet through the water (and displace that water) for every inch forward the boat moves, how can they be efficent? I don't dispute it, I just don't understand it yet.

Propellers are just rotating wings, and just like wings they work most efficiently when they are long and thin (like a glider or sailplane wing). The math required to understand them gets complex pretty quickly though, as each bit of a prop is working with at a different speed and different inflow angle, making performance analysis complicated.

[lpbman]

Take a "mini submarine"
increase the size of the dive plane

tether to your boat (thoughtfully)

as it dives, dive planes produce forward thrust. blow ballast, same thrust,
repeat.

No idea how efficient it would be, it just seems low power. It wouldn't be fast, but there is no limit to how large you could make a concrete tube.