GreenPower F24+ Electric Race car tips, tricks, and guide: What I’ve learnt
I was once part of the Jaguar Land Rover Team Driven. Which is a JLR funded race team, that design, build, and race, an electric car.
The race is deceptively simple in concept: Given a motor and Batteries which cannot be modified. Travel the furthest distance in the given time.
This is a summary of what I’ve learned. My hopes with this, are that we can collectively learn more, by spending less time and money investigating things that we already know the answer to.
The freedom of information promotes progression.
The year before I joined the team, there were a plague of overheating issues. I took it upon myself to solve them. If only to stop me having to tear down the car to replace thermally damaged components as frequently.
I was limited with what was available on the car to start: a NACA duct on the underside of the car.
- The NACA duct had radiused corners, which is not how it is supposed to be. It was also in a low pressure region.
- These two factors meant that natural air flow was poor.
- Follow this up with a length of flexible ducting for a 30% flow rate reduction relative to rigid tubing for further losses.
- Then the end of the duct was just pointed at the motor, which was in free space. The air will take the path of least resistance, which in this case, is around the heat sink, not through it.
- The heat sink was made up of strips of aluminium, cut a third in on both sides, at 5mm intervals, then bent into a U section, then bent around the motor can.
- This provided poor surface contact between motor heat sink and motor can. imagine a hexagon around a circle.
- The thermal paste was ceramic type, and put on excessively. It is for filling the microscopic gaps between the surfaces. It isn’t as conductive as you’d think.
- The thermal conduction is a function of the interface pressure, the heat sink was held on with cable ties.
How I fixed it:
- I took off the flexible ducting, and fashioned a shroud over the duct that entrained the air through the motor heat sink. Made out of a for sale sign and duct tape. How fitting…
- I designed and had a new heat sink made with the right Internal Diameter and better fin density.
- I removed the paint, and polished the motor can, and the heat sink interfaces.
- I used Arctic silver thermal paste.
- I added 2x80mm fans , one between the NACA duct and the motor, and one on the exhaust side of the motor.
It didn’t overheat… Until someone thought they didn’t need the fans on anymore.
I built a test rig that loaded one motor up against another. One was covered in heat sinks and had excessive air flow over it to ensure stable temperatures. The other was insulated so that it would get hot.
The efficiency was characterised, then logged against the motor temperature.
The result was that it could be said with certainty, that we were losing at least 15w to thermal losses.
Taking this, I had no idea exactly how much air flow we needed. So I took the position that as long as I saved some energy overall, I was doing good. I assumed a 7w budget of power for cooling. 50% reduction in thermal losses if effective.
CFM (airflow) is no use without static pressure. Static pressure is what pushes the air through the heat sink blades. No pressure, no air movement. The other thing is that air is limited in it’s heat absorption capacity. The specific heat capacity of air is 1.006 kJ/kg.K. More air, means more heat removal. So what we are really interested in is the kilograms of air moved, per second, per watt.
I collated a spread sheet of the times most popular computer fans. Worked out their g/s/w values. Found that Scythe Gentle Typhoons are the winner by some margin.
Here is the spread sheet:https://docs.google.com/spreadsheets/d/1ODB5Pm65EUrnNHhY4swx7L9HkrxtLwlAzfIlfiir-As/edit?usp=sharing
From PC water cooling tests:https://martinsliquidlab.wordpress.com/2012/01/15/radiator-shroud-testing-v2/7/ Show that “pull” fans are more effective than “push”. and that shrouds to help the flow stabilise between the fan and the heat sink also help. They also show that two in parallel, are better than two in series push-pull configuration.
- Hard anodised aluminium has the highest emissivity of infra-red radiation. colour is irrelevant.
- It is more important to cool the motor shaft, than the motor can.
- Use disks in close proximity with holes near the centre like a Tesla Turbine. This has proved very effective.
- 120mm fans are most developed format available. Scythe Gentle Typhoons are the king of fan efficiency. 7W is more cooling than you need.
- Gentle Typhoons have been discontinued. In their absence, the new king is the Noctua NF-F12PWM
- Use rigid ducts.
- Duct the air so that it can only go through the blades of the heat sink.
- Use Arctic silver thermal paste if you can afford it. A little goes a long way. If you can’t afford it. Other thermal pastes are ok too. It isn’t a game changer.
- Use fans in suck rather than blow configuration.
- two fans in parallel suck are better than one suck, one blow in series.
- Have them as close to the heat sink as possible.
- The wheel wells are a good, relatively high pressure, region to take air from without messing up aero.
- Exhausts aren’t strictly necessary. Chimneys or reverse NACA’s can be used. Also tail tip holes.
Aerodynamics and bodywork:
- Enclose your wheels
- skin your wheel spokes
- A certain amount of negative camber allows you to attain the same width at full lock as straight ahead
- Negative camber also allows you to potentially drop the car bodywork down slightly, which can help you to achieve a lower frontal area
- The tail should start tapering in as close to the nose as possible
- The tail is much more important than the nose
- The top and side surfaces can curve in more aggressively than the bottom without flow detachment because they aren’t “sucking” against the ground
- Fair the helmet in to the head rest and down into the tail
- Get the helmet so that the drivers eye line is on the same line as the bodywork. This is the minimum helmet you can have outside of the cockpit
- You can lay the helmet back slightly to further reduce the frontal area. If done too much, this can be uncomfortable to look down out of the bottom of your eyes for 45 minutes.
Aim for sub 50kg.
Aluminium saves you 1/3 weight but you need to heat treat it after welding to re-strengthen the welds. You don’t have to heat treat, but if you don’t you’ll have to beef up the welds which will add weight and reduce your saving. The aluminium will also be larger in volume than the same strength steel chassis. May cause packaging limitations.
A steel chassis is fine.
Ply wood is fine
Composites have the potential to be lighter. But they often aren’t due to their implementation.
The aim is to keep the motor at it’s most efficient RPM. Too fast and you lose top speed. Too slow and you pull to many amps, which heats up the motor, which pulls more amps, and off the cliff you fall.
You can work out what gear ratio cars are using by their wheel diameter and lap speed. Gearing as aggressively as the fastest cars is likely to end in a melted motor, damaged batteries, and broken speed controller. It’s better to be 10% too cautious, than 1% too aggressive.
Winning cars often have a fixed gear ratio. The car spends most of it’s time at wide open throttle doing roughly the same speed.
Gearing allows you to maintain the motor in it’s efficient rpm band as you go up/down hills, into head winds, and as voltage drops off.
iWhoosh used a NuVinci CVT for a few seasons and did well. It does weigh more and is less efficient in steady state than a fixed gear ratio though.