Eco Marathon

Hello everyone,

Our team is preparing to participate in the Shell Eco-marathon in the Urban hydrogen category, and we’re currently working on selecting the best hub motor for our vehicle. As a mostly mechanical team, we don’t have extensive knowledge about hub motors, so any guidance would be invaluable.

Project Details:​

• Fuel Cell: 450W output, with peak power capabilities supported by superconductors. Full fuel cell specs are available in the attachment.
• Race Requirements: The track is 1600 meters per lap, with one start-stop per lap and a consistent cruising speed of 10 m/s to meet time constraints.
• Desired Motor Specs:
  • Maximum efficiency
  • Peak power of around 1000W
  • Suitable for 14-18 inch wheels
  • Light-weight design
  • Compatible with the hydrogen fuel cell system
• Vehicle Specs: 4-wheel design, approximately 200 kg total weight
• Average Power Consumption: Estimated between 100-200W during consistent operation

Questions:​

1. Motor Quantity: Would a single hub motor suffice for efficiency, or would using two be more advantageous?
2. Optimal Power Rating: Based on our setup and race specs, what motor power rating would provide the best efficiency?
3. Cooling Considerations: Would implementing a cooling system noticeably improve efficiency, given our operating conditions?
4. Additional Tips: Are there other key aspects we should consider to maximize performance and efficiency in a hydrogen fuel-cell-based setup?
5. Tools & Simulation: Can anyone recommend simulation tools or methods for testing and optimizing hub motor performance for our application?

We’re aiming for a lightweight, highly efficient solution that aligns well with the hydrogen fuel cell's power output and helps us meet the race’s speed and endurance requirements.

Thank you for any help, advice, or resources you can share!
View attachment Eco-Marathon versie 6 body.JPGView attachment Eco-Marathon versie 6 frame.JPG

10 meters/second = 22.36 mph

Would it be against the rules to contact Justin at Grin technologies (ebikes.ca) and ask him about the motors he sells? He's got them all on his simulator. And he has built vehicles with them to compete in (France-China) cross continent endurance runs.

Thank you for your reply, I will look into it.

Wow, you have extremely little power ( 400w continuous ) and a 200kg vehicle to move.

With an extremely aerodynamic chassis and the highest efficiency geared hub motor you can buy, your performance looks something like this with a continuous ~400w output:


Edit/play the simulation with this link:
Motor Simulator - Tools

This particular motor is rated ~500w and in this scenario, it's running close to peak efficiency. There's another 1% efficiency waiting for you if the output power of the hydrogen cell was increased.

You could make a better motor choice ( which will involve ditching hubs and building a two stage reduction ) and your efficiency could rise to 90-94%. Cost of the motor would go way up though. How are budget factors?

This ultracapacitor is going to give you maybe a few seconds of power to accelerate, which isn't enough. All capacitors have awful energy density. You're gonna need 2-3x that 400w to get the vehicle moving. What will happen if you can't supply that power is that the vehicle will accelerate at a crawl.... can you pedal this vehicle?

The only problem with using a hub motor is that the axles, shell, etc assume you are putting a normal bike and rider amount of weight on them, but you're not. You might need to chain drive the wheels.

Why did you pick this powertrain and what are the metrics for 'winning' this event?
The hydrogen system seems to be approx 50% efficient and you're going to lose another 10-15% in the electric energy to mechanical energy conversion.

A battery on the other hand will lose ~1% of it's power under such a light load.

If one of the metrics is vehicle efficiency, you lost no matter how good your vehicle design or motor choice is. Why did you chose a fuel cell to power this vehicle?

Thank you for your very interesting insights. Yes, you’re absolutely right; there are some significant challenges. Here is some additional information in response to your questions.

We aim to create a much lighter vehicle but for now we use a extreme total weight of 200 kg, including a 70 kg driver. For the Eco-marathon, we are using hydrogen fuel because our project manager prefers it, and unfortunately, the rules do not allow battery power.

Our knowledge of superconductors is quite limited, but I’ve attached a picture below of the superconductor. Do you think this would be unsuitable for accelerating the car to 10 m/s within 10 to 20 seconds?

We consider a chain drive a serious option, but it would be a bit more challenging to implement in our vehicle. That’s one of the main reasons we prefer a hub motor.

You're referring to a two-stage reduction. At the moment, budget isn’t a major concern, of course we want to keep costs as low as possible. However, we prioritize efficiency over cost. Based on your response, it seems you’re suggesting a small electric motor with a large gear reduction, possibly achieved through a chain or gear system. Is this something like what we see in the picture below?

With the relatively low power available and low speed target required, finding an "off the shelf" hub motor will be a a bit of a search.
For getting up to speed it may be more efficient to consider running a higher RPM motor and a shift-able transmission.
Top quality bicycle transmission components typically run in the 98 ~ 99% efficiency range.
Reducing frontal area with a reclined seating position and no body shape could actually reduce the aero drag more than running an upright seat position inside the streamlined body.

Guessing . . . Delft University team ?

Link to

695086 said:

For the Eco-marathon, we are using hydrogen fuel because our project manager prefers it, and unfortunately, the rules do not allow battery power.

Click to expand...

What is the Primary objective of this challenge ?....maximum distance traveled on a fixed fuel quantity, or shortest time over a set distance ?
Strange rules for an “Eco” event, not to allow batteries ?
And a shame that you are having to base your efforts on a energy source with a limited future in the mass market !
...but I guess the options were limited to Solar or wind !
PS.... “Superconductors” are very different to Supercapacitors ,..which are also a much over hyped technology for energy storage !

My napkin math says
9 watt hours * 60 minutes * 60 seconds = 32400 watt-seconds ( to invent a term )

If you have enough funding and the rules are such that you have to use the worst form of energy, i'd build a dual reduction drive system using an astro motor ( can hit 95% efficiency ) driven by a VESC. This would be the most efficient powertrain that you could get in a small size.

few hub motors will even approach 90% efficiency, the problem with the ones that can is that they're huge and their peak efficiency is in the 1000's of watts, so running them at 20mph means you will get poor efficiency..

With the big disadvantage in your power source, you'll want all the aerodynamic and electrical efficiency you can get, so you should basically go nuts.

Thank you all for your support, we truly appreciate it!

Answers:

We are not from Delft University, but we take that as a compliment! We come from a much smaller institution with far fewer tools and resources compared to Delft.

The primary objective of this challenge is to achieve the highest km/m³, measured using a flow meter and joulemeter. Here's an example: if we cover 10 km using 50 liters of hydrogen, and the total electrical energy recorded by the joulemeter is 10,000 J, we can calculate the total hydrogen consumption as follows:

Example: 10 km is covered using 50 litres of hydrogen and the total electrical energy measured by the joulemeter is 10,000 J. Using the equation: litres of hydrogen = (joules / fuel cell %) / (hydrogen NCV x hydrogen density), the corresponding total calculated consumed litres of hydrogen are:

• Hydrogen: 50 litres hydrogen
• Electricity: 10,000 / 0.5 / (119,960 x 0.08988) = 1.855 litres hydrogen
• Net = 50 + 1.855 = 51.855 litres hydrogen
• Where: 0.5 = efficiency of the fuel cell and 0.08988 = the density of hydrogen (g/l) at STP 28
• The final fuel consumption result is thus: Fuel consumption (km/m3) = (km / (l / 1,000)) or 10 / (51.855 / 1,000) = 192.85 km/m3 hydrogen

Apologies for the earlier confusion, we are indeed using an ultracapacitor. My mistake! ☹

New questions:
From what I’ve gathered from all the feedback, it seems that a hub motor might not be the best choice for our car. We are now focusing on an electric motor with single or dual reduction. Do you guys have any simulation software recommendations? Something like MATLAB, perhaps?

What do you guys think about the efficiency differences between 1, 2, or 4-wheel drive setups, excluding considerations like handling and installation complexity (e.g., 1x1000W, 2x500W, or 4x250W configurations).

Also, do you believe we should invest in cooling and regenerative braking, or would the potential efficiency gains not justify the effort?

Thanks for the clarification!

There exist lots of simulators for this, but i only use one, because i build electric bicycles primarily:
Motor Simulator - Tools

The data is based on real world, good tests. I find the simulator quite accurate in real life scenarios.
Finding a motor analogous to your choice and using this simulator would be a good idea.

If you are using dual motors in a very aerodynamic vehicle, you may lose or gain efficiency due to the choice.

Example scenario that's approximate to your application.. take the 2nd motor out of the picture and you gain 4% efficiency. Opposite case on a typical upright bike going higher speeds.
Motor Simulator - Tools

Your other issue is that you need to optimize for a point of load that's very difficult to achieve peak efficiency on. Two motors makes that harder. Slow speeds makes that even harder. That's why the single motor with a customizable gear reduction is your go to.

This is known as a recumpence drive.. basically astro motor -> belt -> belt -> chain.. the most efficient motor paired to one of the most efficient transmissions. You might want the second stage to be chain so that you can easily vary the ratios during experimentation to find where peak efficiency is for your vehicle and scenario.



Regen is just a function of a controller - buy a programmable controller & you now have the function.
Invest in cooling only if you need it - a highly efficient drive system would probably be very happy to be driven on air.

Also if you want to go this route then i'd call astro for an ideal motor selection for your scenario. You want to be peaking at around 450 watts and the choice of motor is crucial.

Any reduction, geared, chain, or belt, will introduce frictional losses and reduce the overall drive efficiency. In addition to adding weight, complexity and potential failure points.
Likewise, multiple motors, controllers, etc will each have their own associated efficiency losses and additional weight which are best avoided.
I have not checked all the options but, with such a low power drive I cannot imagine being able to better a simple, single, direct drive, hub motor set up to minimise the losses and get the most out of the limited energy available.

I can't get a hub motor setup to reach 85% efficiency in the load conditions of OP's vehicle. The problem is that we can't get it spun up fast enough to get into it's efficiency band at 450w. The best small DD is going to be rated 1000w and achieve peak efficiency at around 800w continuous at the low voltages he'll be using.

With losses, the RC drive suggested should be 5-7.5% more efficient. That's because we can adjust the gearing to be just right to hit the perfect load point. A custom hub motor would need to be built for it to work.

CSIRO did the latter a long time ago, demonstrating an astounding 97% peak efficiency in 2011. The motor cost around 10k to build as a one off. Hub motors for bicycles and motorcycles won't resemble this at all.

Delft University has created some high efficiency streamlined bicycles that ran at the Human Powered Speed Championships, Battle Mountain Nevada USA which is why I guessed . . .
One wheel drive is enough to drive a light weight vehicle and reduces the engineering and parts count.
Schools have an inherent resource advantage in that product donations are much easier to achieve.
To lower mechanical friction change all bearings to top quality "shielded" not "sealed".
On a circuit course sometimes a freewheel ratchet on the wheel can be a bigger energy saving strategy then trying to recoup power via regen.
Once over the peak let the vehicle coast, gravity may actually increase vehicle speed where regen will hinder speed.
Tires . . . as a general rule of thumb, look for high thread count supple casings made of nylon.
Pressure is set by percentage of sag with rider on board, ready to run. 10 ~ 15% sag by profile measured from axle center to ground.
A 20mm tire will sag 2 ~ 3mm.
Over pressure or under pressure can be an unintended source of rolling drag.

I've previously competed in Electrathon America, World Solar Rally, and Human Powered Racing.

Hillhater said:

Any reduction, geared, chain, or belt, will introduce frictional losses and reduce the overall drive efficiency. In addition to adding weight, complexity and potential failure points.

Click to expand...

Not always a true statement.
Here's a couple examples :
A twin cam, 4 valve engine has more mechanical friction that a single cam 2 valve engine yet the 4 valve will be overall more efficient.
All record setting Streamlined Human Powered vehicles have a two stage gearing system.

PaPaSteve said:

A twin cam, 4 valve engine has more mechanical friction that a single cam 2 valve engine yet the 4 valve will be overall more efficient.

Click to expand...

Yes but,... that is THERMAL efficiency, as opposed to the mechanical efficiency of the drive train we are discussing.

All record setting Streamlined Human Powered vehicles have a two stage gearing system.

Click to expand...

Again, different application....human power,..and looking for max speed from a limited rpm input , rather than efficiency.

Hillhater said:

Yes but,... that is THERMAL efficiency, as opposed to the mechanical efficiency of the drive train we are discussing.

Again, different application....human power,..and looking for speed rather than efficiency.

Click to expand...

In both examples the transfer of energy potential to mechanical results is improved by using more parts.

The best example of THERMAL efficiency is a modern Formula 1 engine.
A highly complex system that transfers slightly over 50% of the energy potential into mechanical results.
Most road going vehicles using less parts are under 40% THERMAL efficiency.

Although, I'm an advocate of designing with the least amount of parts.
Ideally if one part can do two jobs all the better.
I opine, based on winning experience in 3 different race series, the best system efficiency for this project can be achieved with a small, light, high efficiency, high RPM motor combined with one or two stage gearing.

Ultimately it is up to the students to math this problem for themselves.

Attached is a design example of one part doing four jobs.
1) main frame
2) roll cage
3) half the seat frame
4) streamlined body attachment

PaPaSteve said:

In both examples the transfer of energy potential to mechanical results is improved by using more parts.

Click to expand...

1)..we are not discussing improving thermal efficiency, but electrical to mechanical efficiency.
Basically, is a high rpm motor with a reduction drive train more or less efficient than a low rpm, direct drive, motor needing no drive reduction ?
And I say again,.. any reduction drive system will introduce losses, so the ultimate question is ...
....can a low rpm DD (<1000rpm) motor have a similar efficiency as a high rpm (10,000rpm ?) equivalent power motor ?

Turning stored energy into mechanical movement is the overriding discussion.
Heat loss (thermal efficiency) is always part of the equation including with anything electric powered.

By Neptronix calculations a hub motor can only provide 85% efficiency (see post #12)
Quality bicycle chain and sprockets run 1 ~ 2% loss
Say a double reduction is used that could add up to 4% loss . . . and a small, light weight motor runs 95% efficient that adds up to 91% system loss so then there is a 6% gain in system efficiency over the hub motor.

While simplicity is a good rule.
Sometimes turning over every rock finds the best results.

Thank you al for thinking along, I really appreciate that!!

I will gather all your information and develop it to achieve the best results for my project. I’ll reach out if I need assistance or when I’m ready to present my plan/project.

Wishing everyone a great day!

PaPaSteve said:

Quality bicycle chain and sprockets run 1 ~ 2% loss
Say a double reduction is used that could add up to 4% loss . .

Click to expand...

An interesting discussion...
And whilst your figures for bicycle chain efficiency are well documented, I do not believe they are applicable to the application here .
Bicycle chain sets operate at low (<100 rpm ) drive speeds with a less than 2 m/s chain speed on a 30cm+- drive sprocket dia.
An RC motor reduction drive would be operating at 3-5000 rpm on a much smaller (5cm ?) drive sprocket , consequently with a much higher chain speed (8+ m/s)
Transmission efficiency is known to reduce with increased chain speed and reducing sprocket sizes
Where do we get reliable efficiency data for small RC motors ?.

Astro has published some efficiency data for their motors in the past, has some peak numbers on their website.
95% peak efficiency not unheard of for exceptional RC motors.
Recumpence drive used two belts, could be more efficient than chain.

All losses included, you still need a custom built hub motor to beat a very high efficiency 2 stage setup. The best peak efficiency i could find on ebikes.ca simulator is 83% out of a Grin All Axle. Geared hubs weren't more impressive.

One would have to lose 12% of their energy to get as poor of efficiency as the most ideal hub we can buy, i don't think a dual reduction drive, even the worst one, is going to lose that much energy.

Bicycle hub motors are far too large to be efficient at 400W. So much drag from all that iron and all those poles. Not to mention their immense weight.

I would recommend a single stage reduction from a small, lightweight drone motor like a 9225 (also known as 8318). Read this thread for more info on choosing a motor for a similar application.

I came across this video on the internet about a drone motor on a bicycle: []. In the video, it performed decently, but the motor still overheated significantly. Since our car will need more torque, it’s likely this motor would overheat in our setup as well, even if we use two drone motors.

He also has another video that seems more relevant for our car project: []. In that video, he mentions that the hub motor seems to perform better, though this may be due to the specific setup and engineering choices. I imagine that the difference between a single and dual-stage reduction depends on the gear ratio, but I need to refine my calculations in Excel before buying parts and testing.

For now, the best option seems to be an Astro motor or something similar with a one or two-stage reduction using gears, chains, or belts. This approach not only improves efficiency but also allows us to adjust the gear ratio and motor mounting, which is ideal at this stage since our school’s budget limits frequent motor replacements.

If anyone has tips or experience with a powered dual-reduction drive system, please let us know, it would be a huge help!

Thank you all!

Keep in mind RC model airplane motors are designed to be cooled by the airflow of the propeller.
Do a search for "400 watt, high efficiency, brushless electric motor" see what comes up.
A low-ish RPM motor might be easier to engineer for.
> This ? <
> Custom ? <
Remember the purpose of the project is learning
Some of the best lessons come from mistakes

Searching various other posts on “Eco Challenge” etc, such as the link posted by Pronghorn above, it is apparent that this issue has been well worked through many times with similar outcomes, .
Single reduction belt drive from small “Drone” pancake motor seems to be the go to set up.
Typically..
However, I have not been able to find any hard data for the % efficiency of these systems,.. input power vs output power to the road wheel .
Most teams seem to focus (understandably) on overall Wh/ km results