Author Options:

Carbon Fibre Flywheel (for energy storage) Answered

I'd like to build a flywheel but I'm not sure what I need to know before i start, Is it difficult to do? Is it possible to accomplish with off the shelf parts? Any help at all would be appreciated greatly.


So, a question. Most of what I've seen on flywheels concentrates to a significant degree on energy density and issues related to mobility. How about if I want to go the other direction? If I have a wind/solar energy source and am looking for a storage sink - how about a big, fat, slow - but massive flywheel? I kind of like the idea of 100 tons of concrete cylinder moving at about 30 RPM to provide temporary storage. I've got a big yard. I know I'm not going to like the answer - but could someone point me to the right equations for determining the energy storage capacity of a rotating cylinder of mass 'x'? Thanks

Let me tell you about one of my concrete flywheel experiments.

500# of concrete and 3/16 steel rim to hold it together.

4'x6' hole in the back cement block wall of the garage.

totaled out a 3 month old Mercedes parked on the other side of the wall.

3 weeks in the hospital

6 broken bones

. That ought to work. It's the ?gyroscopic? effects that make moving it around so cumbersome. Not a problem when it sits in one place. And a huge, slow rotating flywheel is much safer. . Not sure about the calcs. I think angular momentum is involved. Try searching for that.

As others have pointed out energy storage goes up with the square
of the speed, so it is better to have a faster flywheel that weighs less.
To put numbers on it. If you have a 400kg flywheel ( yes I'm ignoring second moments defined by flywheel shape for now) doing 500rpm(51 rads/sec) you get about 1067kJ of stored energy if you take a 4kg flywheel and spin it at 5000rpm (517 rads/sec)you get the same amount of energy. at 10,000 4271kJ4 4x more, at 20 000 rpm 16x more. The problems of drag at high rpm and electrical losses of getting the enegry in and out aren't trivial, but it's (almost) always betters to sacrifice mass for rotational speed.
BTW Audi used a flywheel sucessfully at lemans this year to recoup braking enegry.

As far as I know, no vehicle has ever been successful with a flywheel as its primary energy source. Certainly no car has, as a large-enough flywheel plus associated framework and gearing will not fit inside a car.


I have heard of trials with buses that use flywheels as intermittent energy stores - the breaking system transfers energy to the flywheel as the bus slows, then that energy is used to assist the bus pulling away from stops.

The wheel needs to have as much mass as possible, moving as fast as possible, as far from the axis as possible. Carbon fibre alone is just not massive enough (maybe you could bind a steel wheel in CF to restrain possible shrapnel?).

Although this car (www.time.com/time/magazine/article/0,9171,985185-2,00.html and money.cnn.com/magazines/fortune/fortune_archive/1996/09/30/217438/index.htm) doesn't use the flywheel as the "primary" power source, it takes care of the power demand for acceleration.  I remember seeing a television documentary about the Rosen Motor Company that both articles treat in depth.  In contrast to the failed test that the article mentions, the television program highlighted a successful test of the Rosen Motor power train, installed in a modified Saturn sedan, which used the flywheel storage and a turbo generator for the electricity.

Again the flywheel is never going to be the "primary" power source in anything, but its ability to "buffer" power and even out the demand curve will make flywheel energy storage systems (FESS: my acronym?) almost "primary" in their ability to reduce or "even out" fuel consumption.

I've also read that mobile applications of FESS can use a gimbal arrangment to "unlock" or free the flywheel's angular momentum/velocity from being "tied" to the vehicle's angular momentum/inertia when negotiating all those pesky curves in the road.


8 years ago

The rotational kinetic energy of a hollow cylinder =                                           Energy rotational =1/2 M ( r_1^2 +r_2^2 ) * V_a^2

Where M= mass of cylinder   r_1= inner radius r_2= outer radius and V_a= Angular velocity in radians/sec ........... This equation states that rotational energy increases with the square of the angular velocity. Which is why commercial high energy density flywheels have insane rpm rates, more energy for a given system mass. However, as has already been stated, when these units fail, they dont just stop working, they explode!!! I have heard of carbon fiber flywheel failures that resulted in a pile of red hot ash at the bottom of the chamber. That does not sound like something that I want in my car...

He's been gone for >2 years, but the maths is a good addition.
You were thinking of those flywheel-toy-cars like I was? (vrrrrrroooommmmm....)


I don't know how you could do it right with off the shelf parts. The flywheel would have to be perfectly balanced with bearings that produce near zero friction. Transferring the power to and from the flywheel would be another mater altogether.

I'm thinking about using rare-earth magnets, or some kind of electromagnetic set-up (preferably the former). The only thing I'm not sure about is how to make a flawless carbon-fibre flywheel, or if I should just resort to buying one directly from a manufacturer.

I seriously doubt you will be able to find a ready made carbon fiber flywheel that would be appropriate for energy storage. It should not be terribly hard to make one though. I would use a very high tensile strength carbon or spectra prepreg. T-1000 carbon fiber is pretty much the strongest you can get. Spectra has similar tensile strength but it less dense so it should have a higher specific energy storage capacity. The shape of the highest possible specific energy storage flywheel is similar to a discus (possibly not intuitive but...). The greatest stress is towards the center of the flywheel since the center carries all the stress of everything outside of the center. For this reason the flywheel is actually thicker in the center than it is at the outer edge. If you don't plan on storing the energy very long then you may be able to do away with the fancy magnetic bearings and vacuum housing. The flywheel will eventually slow down, but if you have already used most of the available energy then there won't be much lost. All the fibers in the flywheel are laid across the axle perpendicular to the axle and parallel to the plan of rotation. The potential for an explosion is directly proportional to the energy stored in the flywheel. If you have the energy equivelent of 20 gallons of gasoline stored in the flywheel, then the explosion will release the energy stored in 20 gallons of gasoline. I would think that one of the hardest things to work out on a prototype will be a motor generator that can run over the entire speed range of the flywheel.

. Flywheels are great for applications that don't involve moving the flywheel around (other than spinning it), but don't work well in mobile applications. . Gyroscopic effects mean the bearings, &c have to be MUCH heavier than a static installation. . Safety is a BIG concern. Flywheels have a distinct tendency towards catastrophic failure. The amount of energy released, in a very short period of time, is not conducive to human health (at least for one that would do any good propelling a car). Scattershields have to be very heavy to be effective. . To store enough energy and still be able to keep the RPMs to reasonable levels, the flywheel will have to be quite heavy. I've never run the calcs, but will guess it will be heavier than a V-8 engine. . Heavy, heavy, heavy.

Have you folks read the article about flywheels on Damn Interesting? It has a lot of information on this subject. Based on the article, I draw the following conclusions:

  • Using a heavy material allows the flywheel to store more energy, as Kiteman pointed out. However, even steel is not strong enough for speeds greater than about 3000 RPM. The advantage of carbon fiber is that it is strong enough to spin at up to 100,000 RPM, which yields a pretty massive energy storage density.
  • It is really unlikely that you're going to be able to get into the tens of thousands of RPMs with a DIY project. That kind of setup requires magnetic bearings, and the flywheel needs to be suspended in a vacuum. Plus, anything spinning that fast is extremely dangerous. Unless it is _very_ carefully controlled and contained, it is likely to cause violent deaths.

So I agree with trebuchet03. A heavy, (relatively) slow spinning flywheel is the way to go.

And it seems that the major barriers to using flywheels to power cars are:

  • The aforementioned tendency for explosions. Even carbon fiber is very dangerous when it is spinning at 100,000 RPM.
  • Cars move around and accelerate a lot, which causes the flywheel to wobble. And wobbling can cause the wheel to bump into its housing, which causes damage and can produce one of the aforementioned explosions. Magnetic bearings just aren't sturdy enough for use in cars yet.
  • A very high-speed flywheel is very expensive. And less expensive variations, like the flywheel used in the gyrobus, do not store enough energy to be worthwhile and are really heavy.

You can build a slow turning flywheel :p So balance isn't as big of an issue ;) Last year there was a human powered water purification senior design team at my school -- they used a urg rowing machine, but replaced the fan with a 25 pound weight (as found in a gym).... What are you trying to do? That usually helps determine a design :p

I'm actually trying to build a system small enough to fit in to a car, so weight is of importance. I've read that carbon fibre is great for this, since it is pretty light but had the added advantage of being stronger than steel (which would give me a much higher rpm potential)

The amount of energy stored in a flywheel is proportional to its mass and its angular velocity. Consequently, a lightweight flywheel (or a slowly turning) flywheel would be very inefficient for storing energy. Cheers, Pat. Pending

Not so much inefficient - it just stores less energy ;)

Luckily the relationship for angular velocity is omega2 and mass is linear. So if you double the speed - you can reduce the mass by a factor of 4 and get the same amount of stored energy :)

Thing thing that baffles me is why this tech hasn't been implemented earlier. The advantages are obviously there, and the shortcomings could be easily overcome with enough effort..Is it the danger of explosion that keeps this tech from entering the auto market?

Is it the danger of explosion that keeps this tech from entering the auto market?

That, plus reliability.... The materials we have now can handle the rpm.... But the issue comes from the sudden angular acceleration. Over the lifetime of the thing, there will be many many cyclical loads - so fatigue is a problem.

So I did some reading, and found a UPS system that used a flywheel (instead of a battery) to keep equipment powered in the event of a power outage - which has been working for many of years. The spin up a FW and keep it spinning (very little power necessary to do this) - when the power goes out, the motor becomes a generator and power is fed back into the building... Keep in mind that the fatigue cycles on something like that are much lower than a car's :/

Alright, so the angular acceleration is a large problem to overcome. What if a bank of capacitors is used in instances where the car needs to accelerate? if the car is going at a constant speed (let's say 60km/hr), would the fatigue be any less? Would a split system of capacitors and flywheels be executable?

Not so much inefficient - it just stores less energy ;)

No I was right the first time - it will be less efficient due to increased bearing friction and air drag losses (for a given amount of energy stored) ;)

Pat. Pending

Fair enough :p I guess I'm just stuck in "ideal" mechanisms mode :p Although s/he did mention magnetic bearings -- but viscous losses wouldn't really change.

You'd def. want to seek out a manufacturer - especially if it's going into a car.... FW's with a decent angular velocity can be very catastrophic when failure occurs - I've seen photos of a clutch assembly 'explode' - it was violent enough to tear a hole in the cast aluminum bell housing o.0

Yea I've read that one of the drawbacks of having a flywheel powered car would be the risk of a spectacular explosion. That is part of the reason why I want to use several smaller flywheels, working in conjunction. Carbon fibre would be best for the job. Apparently when a carbon fibre flywheel fails, it disintegrates in a much less explosive fashion and doesn't cause as much damage as an identical flywheel of steel or iron. Do you recommend any manufacturers? resorting to a manufacturer would be my last option..