> Flywheel motors have often been compared to capacitors, capable of quickly storing and releasing energy. Proponents cite weight, cost, and environmental advantages over traditional chemical-battery hybrids.
Indeed. Kinetic energy in the form of a flywheel has both the advantage and the drawback that it’s relatively easy to release all of the energy in a fraction of a second.
Of course you can do the same with chemical energy, but we have good intuition around things that will detonate.
With everyday substances, both a stick of butter and an SUV traveling at 100 mph contain about a kilowatt-hour of energy, but the latter would generally be more destructive if something goes wrong.
If anything, it's astounding just how much energy is in food like butter, and in such a stable configuration.
Alternatively, it's humbling how bad and inefficient we are at manipulating the physical materials in our environment, because we can't so much as replicate a stick of butter without making it noisy and dangerous.
Explosives, much as other forms of energy storage, tend to have lower specific energy (energy per unit mass) than slower-burning fuels.
1 kg of TNT has roughly 4.6 MG.
1 kg of petroleum has roughly 45 MG, ten times as much.
The difference is that the trinitrotoluene combusts rather more rapidly than the oil. But if, say, you simply want thermal energy and have a mass budget, you'll go with the oil.
C4 has an energy density of about 5.1 MG/kg, slightly greater than TNT, but only 1/8th that of petroleum.
This is because for the butter, you're not counting the weight of all the oxygen that is needed for combustion. 1kg of butter requires several kg of oxygen (or if you're using air, even more weight if you count all the inert nitrogen)
That's a valid point and does shift the balance somewhat, but not fully in favour of nitrogen-based explosives.
From a practical perspective, at least within Earth's atmosphere, that oxygen is "free". There's not need to carry it with you, though some combustion applications may require feeding it at high rates (turbochargers, compressors, gas turbines, and the like). Most automobiles functions strictly under natural aspiration, and even industrial combustion typically utilises relatively low-power bellows or air introduction.
Rockets, of course, carry their own oxidizer, so we can compare energy densities based on net yield.
The Saturn S-IC (first stage of the Saturn V stack) is powered by the combustion of RP-1 (a highly-refined grade of kerosene) and LOX, liquid oxygen.
The tank sizes were 730,000 L fuel, for which I'm assuming an average density of 0.9g/mL (0.81--1.02 given by Wikipedia), and 1,204,000 L oxidizer, at 1.41 g/mL.
That works out to 657 tonnes of RP-1, 1,374 tonnes of LOX, or a total of 1,511 tonnes.
The fuel alone is 43.5% of the total mass, which means total specific energy is reduced to 17.63 MJ/L, which is still about 3.5--4 times greater than TNT or C4, by comparison.
There's one further question as to whether or not the rocket is tuned to run rich, stoichometric (balanced fuel and oxidizer), or lean. Most LOX rockets burn somewhat rich, based on recollection and Wikipedia, so the LOX quantity is slightly short of what might support complete combustion, though not enough to affect the overall outcome here.
ooooh now I understand, I thought GP meant that both a stick of butter and an suv, both travelling at 100mph, had the same kinetic energy and I was scratching my head as to how big that stick of butter was.
My dad told me a story about some British pilots who were pitching in to help the Soviet air force during WW2. The day the temperature dropped below freezing, the women brought out big kettles full of fat to feed the pilots. The RAF pilots were disgusted. But after flying at altitude in the cold, they relished eating the fat. It was loaded with calories that kept them warm.
I noticed something similar when I'd go skiing. Loading up with eggs and butter for breakfast meant I'd never feel cold.
Probably no accident that old-school ski cafes on the slopes serve up calorically dense food like cheesy omelettes and fondues...
Talking of old-school, I am reminded of Robert Wood's The 2oz Backpacker from 1980-something that advised hikers to take a swig of cooking oil before going to bed, for calorie-burning warmth.
Idk if I could stomach the cooking oil, but extra virgin olive oil or preferably some salted butter would be more palatable. Cheese sounds even better still but not as calorie dense.
Really it's not just the butter though, it's the combination of butter + oxygen - there is quite a lot of oxygen needed. Start putting it in tanks like rocket fuel, and it looks more 'energetic'.
15 pounds? Not gonna store much energy. Many cars already have several pound turbo impellers spinning at 50k+ rpm. (And remember force will be proportional to rpm^2)
Turbos are small and lightweight. The goal is for them to respond quickly to changes in requested power, with minimal lag (already a problem). So small and light, with effective (air) pumping power.
If your turboimpeller is doubling as kinetic energy storage, you're doing it wrong.
It's not accurate. Turbo impellers should not be heavy enough to store significant energy, except maybe in large diesel engines, and are also naturally housed in a large, often cast iron, housing that will easily contain any shrapnel.
Turbos also run periodically and reasonably infrequently, at least for street vehicles. They're invoked generally under high-acceleration or load, not when cruising at constant speed.
A flywheel would typically be spinning not only far more often but at great speed and load.
A bearing failure is probably the most common failure mode, would probably result in an unignorable insanely loud noise, and render the unit inoperable once they actually seize.
I replaced the cast aluminum bell housing on my dodge with a forged steel one, probably about 9x stronger. I've seen pictures of flywheels sawing through the body of the car. I like my feet.
All biologists are shown photos of exploded ultracentrifuges. It makes them care for centrifuge rotors, the consequences when limits are exceeded are spectacular.
Also this is a problem enough at <10,000 RPM with just normal metal flywheels and clutches that many drag racing and Motorsport rules require extra “scatter shields” to protect both drivers and spectators from them exploding or exiting from the normal bell housing.
As of today the characteristics of flywheels are definitely better for racing and not very useful to consumers. But I believe that they will find a usage if we find a way to miniaturize them enough.
Regenerative braking converts mechanical energy to electrical energy to send into a battery. This energy is then extracted, reconverted into mechanical energy, and then transmitted to the wheels.
A compact and efficient flywheel should remove the energy conversion and could improve the efficiency of EVs. We'll obviously hit a density limit at one point, but we're currently far off.
Williams proved it works, but because F1 cars have to use an ICE in a tight package, they had to pick between batteries or flywheels. We should hopefully start seeing EV race cars implement this soon.
Flywheels have a ton of drawbacks compared to batteries for normal passenger cars. Not least of which is that it won't just spin indefinitely and passenger cars constantly park and sit for hours and hours.
I think you misunderstood me: I meant flywheels could be used for temporary storage (regen breaking), as a complement to permanent storage (battery), because it is more efficient for regen and handles energy bursts better.
Flywheels should be more efficient. With the current tech for mechanical <=> electrical conversion, a super-capacitor with 100% efficiency would be as efficient as a flywheel with ~65% efficiency for regen braking. The flywheels by GKN are ~98% efficient.
The downside is size and weight, which makes me believe miniaturization is the key.
Miniaturization doesn't make sense for flywheels; to increase energy stored by a flywheel you need to do one of:
1. Increasing RPM
2. Increasing the mean radius of the mass
3. Increasing the weight
Two of these directly impact either size or weight (#2 and #3 respectively) and the first one is limited by the tensile strength of the materials in use. I'm going to go out on a limb and suggest that the F1 teams were already using exotic materials, so there may not be a lot of ground to gain here.
Agreed with you on the physics, but I think there is much more progress to be made on the rotor to handle higher RPMs, time will tell. For the F1 experiment, they used magnetic composite materials.
It sure is, and there are even more funds and more people working on this today than there were 30 years ago.
Breakthroughs have been found recently. People use busses and watch cars win 24h endurance races using this technology, partly thanks to the research on composite materials.
This is done, as an alternative to "spinning reserve" with renewables-loaded grid supplies.
The benefits are mixed. Amongst other problems, the Earth's own rotation actually puts a lot of stress on the rotational axis. free-floating flywheels aren't generally highly practical, so compensations (I believe magnetic bearings are common) are used.
The apparently obvious storage mechanism ... turns out to be rather difficult to operationalise.
Couldn't the gyros just be oriented so the axis align with earths rotation? At the equator thats easy, just mount it vertically. In the contiguous US I'd guess a what 50deg angle?
It'd be fun in a scifi movie to see large disc shaped buildings sticking half way out at an angle all over.
At the equator a gyroscope on its side would have the same orientation for its axis of rotation as the earths. Think of holding a gyroscope on a merry go round.
That would require a free-floating (gimbal) mount, so far as I understand.
Most grid-support gyros are in fact cylindrical rather than disk-shaped, as this optimises for mass at the outer radius of rotation. To gimbal-mount a large cylinder you'd have to have a proportionately large spherical enclosure.
I'm only somewhat generally and vaguely aware of the issues here and I don't know what R&D or theory are specifically ongoing here.
As noted, polar and equatorial locations would require the least adjustment, as polar flywheels could be aligned perpendicular to the ground, equatorial ones parallel (along the north-south axis), which would minimise any movements, as others have noted.
(I don't think that Earth's axial inclination would matter, and am presuming that precession of Earth's own axis is Too Small to Matter.)
A similar, common and safer mechanism is to just move water. Instead of spinning up a flywheel, water is pumped to an upper reservoir. When the power is needed, it's drained through a generator and the water can be reused when there is excess energy again.
That’s right, but it wouldn’t be a very large torque, I don’t think. Especially since the roll is resisted by the spring damped suspension and the sway bar. The resulting force would pitch the car up/down depending on turn direction and flywheel direction.
Well, no, since you have two equal masses rotating in opposite directions...
EDIT: Oh, I see what you're saying. I thought you were talking about the oppsite direction: hit brakes, and car turns. You're instead talking about putting a caliper on each disk, and snapping the car left/right. Oh man, that would be fun ^^
To better understand the intuition here, imagine you spin a really heavy bike wheel on an axle held horizontally between your hands. It would be easy to spin the axle, which is the equivalent of turning the car here.
Another option is a gimbal mount, though that makes any mechanical coupling more difficult, and requires more space. For electrically driven / driving flywheels, it's an option however.
I was talking to one of their engineers a long time ago, so the details may be fuzzy, but I think they're able to start supplying power in a single 60hz cycle. The idea is that the flywheels cover the start time of your diesel generators.
They were not considered a battery replacement though, what they did was for areas of the world that take frequent and quick power loss events they allow the UPS system to ride through a small power loss without hitting the batteries (prolonging life) or generators (saving wear/fuel). The flywheel only lasts a couple minutes under full load. Any power loss longer than that transfers to battery and triggers a generator run to prepare to take the load long term.
In that case why have batteries at all? Because generators and transfer switches aren't perfect. On typical data centers the batteries are sized for a fifteen minute full load capacity. That's fifteen minutes where the on site people can potentially fix something. That's one reason, the other is with flywheel you must start the genset earlier so it can take power before the flywheel dies, leading to more wear and fuel usage. Starting motors creates a lot of wear, drastically increase that and your gensets don't last long at all.
Why not have a lot more batteries than fifteen minutes? The batteries are sized for the computer load only, not the full facility HVAC load. In general in a server room that's at capacity you have fifteen minutes run time without HVAC before you start shutting servers down on high temperature.
So as you can see flywheels are an expensive addition that only make sense in certain isolated areas with bad power.
they always had regen braking. this is different: " Volvo’s KERS setup is based around a carbon fibre flywheel that spins at up to 60,000 rpm. Braking spins the flywheel up, and accelerating takes advantage of its momentum by supplying drive to the rear wheels."
Why would they make the flywheel out of carbon fiber? Isn't carbon fiber supposed to be very lightweight? In a flywheel you would want a massive wheel.
my guess that it's easier to spin-up during braking and it's less weight to carry during driving. I guess they went through a bunch of materials before they figure out price/performance/efficiency sweet point. Flywheel even enclosed in vacuum to minimize friction...
It's immensely heavy and expensive, and not very effective.
Edit: okay, that's probably going to need a little more explanation.
You've got a car travelling (frictionlessly in a vacuum, as per all good physics experiments) with a flywheel that can be connected to the driveshaft with a clutch. When you engage the flywheel it disengages the driveshaft from the engine.
Your car is travelling at 50mph, so it has some amount of energy that you can calculate from its mass. The flywheel is at rest. When you engage the flywheel clutch, the car will slow down and the flywheel will speed up until both have (neglecting losses, remember) an equal amount of energy, and the car will be doing 25mph.
Bring the car to a stop, with the clutch disengaged again. The car has no energy but the flywheel is still spinning.
Now engage the clutch to move off. The car will accelerate to 12.5MPH, because it gets half the energy in the flywheel and once they have an equal amount no more energy can be transferred.
You can get around this with clever gearboxes, a bit, and if you're building a multimillion pound racing car that's definitely the way forward.
So you've got a heavy and complex motor-generator in the drivetrain, a heavy and complex motor-generator in the flywheel, and a lot of heavy and complicated power electronics to join the two?
I'm not advocating for anything here and not in it for a discussion - it's just that your criticism gave the impression you might have missed part of the article.
Huh, the flywheels discussed here are charged/discharged electrically. I would have assumed that part of the point of a flywheel would have been to make it mechanical rather than electrical! Is electrical more efficient? Or just easier to build?
To slow down a car while speeding up a flywheel requires an infinitely variable gearbox. Ie. When you first start deceleration and the flywheel is stationary, you'd ideally have infinity:1 ratio. When the car is pretty much stopped, and the flywheel going fast, you'd use 1: infinity. And in between you'd use every other ratio.
Such gearboxes aren't really a thing (at least not without massive caveats).
The best you can do is fixed gear ratios and a clutch, or a CVT and a clutch to handle the ends of the range.
Both are mechanically complex and not very efficient (a typical shift under load with a clutch wastes half the energy into heat in the clutch material).
Compare to electric, where a 'gearbox' consists of software/inductors/capacitors/switches, and you can effectively have an infinite ratio (software) gearbox.
Turbochargers are pretty light weight by comparison. The energy they store is minimal - they're just there to shunt more air into the engine, not provide 20kW on demand.
Turbocharger fans/impellers are not offset weights though but perfectly balanced and they are also lightweight.
A flywheel on the other hand is intentionally unbalanced, an offset weight, kind of like a clock pendulum spinning really fast, making it incredibly dangerous when something does go wrong. It basically becomes a centrifugal slingshot.
Mazda has regenerative braking that goes to a supercap, but AFAIK it's only sporadically available in the U.S. because the American fuel economy testing standards don't benefit much from it.
118 comments
[ 4.1 ms ] story [ 179 ms ] thread---
See also:
https://hackaday.com/2021/05/30/bicycle-flywheel-stores-a-bi...
https://en.wikipedia.org/wiki/Dual-mass_flywheel
Having a fifteen pound weight spinning at 30,000 RPM being maintained as well as your average vehicle sounds utterly terrifying.
Obviously there’s energy from batteries or gasoline or whatever in regular cars
Of course you can do the same with chemical energy, but we have good intuition around things that will detonate.
With everyday substances, both a stick of butter and an SUV traveling at 100 mph contain about a kilowatt-hour of energy, but the latter would generally be more destructive if something goes wrong.
Alternatively, it's humbling how bad and inefficient we are at manipulating the physical materials in our environment, because we can't so much as replicate a stick of butter without making it noisy and dangerous.
A butter stick sized piece of C4 is similarly stable but contains more stored energy.
1 kg of TNT has roughly 4.6 MG.
1 kg of petroleum has roughly 45 MG, ten times as much.
The difference is that the trinitrotoluene combusts rather more rapidly than the oil. But if, say, you simply want thermal energy and have a mass budget, you'll go with the oil.
C4 has an energy density of about 5.1 MG/kg, slightly greater than TNT, but only 1/8th that of petroleum.
<https://en.wikipedia.org/wiki/C-4_(explosive)>
<https://en.wikipedia.org/wiki/Energy_density>
From a practical perspective, at least within Earth's atmosphere, that oxygen is "free". There's not need to carry it with you, though some combustion applications may require feeding it at high rates (turbochargers, compressors, gas turbines, and the like). Most automobiles functions strictly under natural aspiration, and even industrial combustion typically utilises relatively low-power bellows or air introduction.
Rockets, of course, carry their own oxidizer, so we can compare energy densities based on net yield.
The Saturn S-IC (first stage of the Saturn V stack) is powered by the combustion of RP-1 (a highly-refined grade of kerosene) and LOX, liquid oxygen.
The tank sizes were 730,000 L fuel, for which I'm assuming an average density of 0.9g/mL (0.81--1.02 given by Wikipedia), and 1,204,000 L oxidizer, at 1.41 g/mL.
That works out to 657 tonnes of RP-1, 1,374 tonnes of LOX, or a total of 1,511 tonnes.
The fuel alone is 43.5% of the total mass, which means total specific energy is reduced to 17.63 MJ/L, which is still about 3.5--4 times greater than TNT or C4, by comparison.
There's one further question as to whether or not the rocket is tuned to run rich, stoichometric (balanced fuel and oxidizer), or lean. Most LOX rockets burn somewhat rich, based on recollection and Wikipedia, so the LOX quantity is slightly short of what might support complete combustion, though not enough to affect the overall outcome here.
<https://en.wikipedia.org/wiki/S-IC>
<https://en.wikipedia.org/wiki/RP-1>
<https://en.wikipedia.org/wiki/Liquid_oxygen>
<https://en.wikipedia.org/wiki/Rocket_propellant#Mixture_rati...>
I noticed something similar when I'd go skiing. Loading up with eggs and butter for breakfast meant I'd never feel cold.
Talking of old-school, I am reminded of Robert Wood's The 2oz Backpacker from 1980-something that advised hikers to take a swig of cooking oil before going to bed, for calorie-burning warmth.
How's that?
This can be converted to 0.94 kWh: (calories)(joules per calorie)(watts = joules/second)(kw = 1000w)
Kinetic energy is 1/2(v*v^2)
The average SUV weighs about 2000 Kg. That yields about 0.55 KWh. Perhaps the parent used a larger SUV.
If your turboimpeller is doubling as kinetic energy storage, you're doing it wrong.
Not that I think this is a good counter argument...
A flywheel would typically be spinning not only far more often but at great speed and load.
Also this is a problem enough at <10,000 RPM with just normal metal flywheels and clutches that many drag racing and Motorsport rules require extra “scatter shields” to protect both drivers and spectators from them exploding or exiting from the normal bell housing.
Regenerative braking converts mechanical energy to electrical energy to send into a battery. This energy is then extracted, reconverted into mechanical energy, and then transmitted to the wheels.
A compact and efficient flywheel should remove the energy conversion and could improve the efficiency of EVs. We'll obviously hit a density limit at one point, but we're currently far off.
Williams proved it works, but because F1 cars have to use an ICE in a tight package, they had to pick between batteries or flywheels. We should hopefully start seeing EV race cars implement this soon.
The downside is size and weight, which makes me believe miniaturization is the key.
1. Increasing RPM
2. Increasing the mean radius of the mass
3. Increasing the weight
Two of these directly impact either size or weight (#2 and #3 respectively) and the first one is limited by the tensile strength of the materials in use. I'm going to go out on a limb and suggest that the F1 teams were already using exotic materials, so there may not be a lot of ground to gain here.
What makes you think there is unexplored potential at spinning even faster? What material or process do you think is unexplored in specific?
1: e.g. https://www.researchgate.net/publication/261388468_Optimal_d...
Breakthroughs have been found recently. People use busses and watch cars win 24h endurance races using this technology, partly thanks to the research on composite materials.
The benefits are mixed. Amongst other problems, the Earth's own rotation actually puts a lot of stress on the rotational axis. free-floating flywheels aren't generally highly practical, so compensations (I believe magnetic bearings are common) are used.
The apparently obvious storage mechanism ... turns out to be rather difficult to operationalise.
It'd be fun in a scifi movie to see large disc shaped buildings sticking half way out at an angle all over.
Most grid-support gyros are in fact cylindrical rather than disk-shaped, as this optimises for mass at the outer radius of rotation. To gimbal-mount a large cylinder you'd have to have a proportionately large spherical enclosure.
I'm only somewhat generally and vaguely aware of the issues here and I don't know what R&D or theory are specifically ongoing here.
As noted, polar and equatorial locations would require the least adjustment, as polar flywheels could be aligned perpendicular to the ground, equatorial ones parallel (along the north-south axis), which would minimise any movements, as others have noted.
(I don't think that Earth's axial inclination would matter, and am presuming that precession of Earth's own axis is Too Small to Matter.)
EDIT: Oh, I see what you're saying. I thought you were talking about the oppsite direction: hit brakes, and car turns. You're instead talking about putting a caliper on each disk, and snapping the car left/right. Oh man, that would be fun ^^
To be clear I’m agreeing with you.
Added weight and complexity.
Another option is a gimbal mount, though that makes any mechanical coupling more difficult, and requires more space. For electrically driven / driving flywheels, it's an option however.
One of the gyros seized and the whole thing flipped and slammed into the ground with tremendous force.
Which are often tangential.
I was talking to one of their engineers a long time ago, so the details may be fuzzy, but I think they're able to start supplying power in a single 60hz cycle. The idea is that the flywheels cover the start time of your diesel generators.
They were not considered a battery replacement though, what they did was for areas of the world that take frequent and quick power loss events they allow the UPS system to ride through a small power loss without hitting the batteries (prolonging life) or generators (saving wear/fuel). The flywheel only lasts a couple minutes under full load. Any power loss longer than that transfers to battery and triggers a generator run to prepare to take the load long term.
In that case why have batteries at all? Because generators and transfer switches aren't perfect. On typical data centers the batteries are sized for a fifteen minute full load capacity. That's fifteen minutes where the on site people can potentially fix something. That's one reason, the other is with flywheel you must start the genset earlier so it can take power before the flywheel dies, leading to more wear and fuel usage. Starting motors creates a lot of wear, drastically increase that and your gensets don't last long at all.
Why not have a lot more batteries than fifteen minutes? The batteries are sized for the computer load only, not the full facility HVAC load. In general in a server room that's at capacity you have fifteen minutes run time without HVAC before you start shutting servers down on high temperature.
So as you can see flywheels are an expensive addition that only make sense in certain isolated areas with bad power.
https://www.lanemotormuseum.org/collection/cars/item/gyro-x-...
https://www.youtube.com/watch?v=tqGB1zOICfg
Edit: okay, that's probably going to need a little more explanation.
You've got a car travelling (frictionlessly in a vacuum, as per all good physics experiments) with a flywheel that can be connected to the driveshaft with a clutch. When you engage the flywheel it disengages the driveshaft from the engine.
Your car is travelling at 50mph, so it has some amount of energy that you can calculate from its mass. The flywheel is at rest. When you engage the flywheel clutch, the car will slow down and the flywheel will speed up until both have (neglecting losses, remember) an equal amount of energy, and the car will be doing 25mph.
Bring the car to a stop, with the clutch disengaged again. The car has no energy but the flywheel is still spinning.
Now engage the clutch to move off. The car will accelerate to 12.5MPH, because it gets half the energy in the flywheel and once they have an equal amount no more energy can be transferred.
You can get around this with clever gearboxes, a bit, and if you're building a multimillion pound racing car that's definitely the way forward.
Such gearboxes aren't really a thing (at least not without massive caveats).
The best you can do is fixed gear ratios and a clutch, or a CVT and a clutch to handle the ends of the range.
Both are mechanically complex and not very efficient (a typical shift under load with a clutch wastes half the energy into heat in the clutch material).
Compare to electric, where a 'gearbox' consists of software/inductors/capacitors/switches, and you can effectively have an infinite ratio (software) gearbox.
A flywheeel is basically a bomb in all but name.
A flywheel on the other hand is intentionally unbalanced, an offset weight, kind of like a clock pendulum spinning really fast, making it incredibly dangerous when something does go wrong. It basically becomes a centrifugal slingshot.
A few years ago supercaps were predicted to be in every car for storing shorterm energy from breaking, since then Tesla sold Maxwell Technology [1].
Have any car makers deployed a different technology than Lithium?
1: https://passive-components.eu/tesla-sells-maxwell-technology...