Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Carbon fibres for flywheel

  1. Sep 30, 2005 #1
    hi all!
    Are composites like carbon fibres (CFRP) the best material for flywheel?
    There is no doubt on its high strength to density ratio.
    Regarding the safety consideration, is flywheel made of composites safer than that made of other materials? I've read through some articles claiming that it should be safer in the sense that composite materials would break into many small pieces once broken. Is this a well-proven fact?
    Also, regarding the manufacturing of flywheel using composites, is it sill difficult? Where are the difficulties?
    All in all, what material would you consider to be the best material for flywheel?

    By the way, can anyone tell me what "toughness" actually means? Is titanium alloy tough? are composites tough?
     
    Last edited: Sep 30, 2005
  2. jcsd
  3. Sep 30, 2005 #2

    Cliff_J

    User Avatar
    Science Advisor

    Search this forum for a more in-depth discussion on toughness, as I recall its strength across a wide variety of loads (as in the area under the strength curve, not just a point along curve).

    What is safe? Having a small air bubble because of a small construction error that leads to a crack that causes a catastrophic explosion of the flywheel at speed sending high velocity mass at people sounds like a recipe for a problem. A simple oversight could cause a delamination or incorrect fiber orientation would result in a much weaker structure than initial calculations would indicate.

    An open wheel race car CAN be constructed with the compositie fibers arranged in a way to absorb impact energy from a collision with a wall. But this is not necessarily a good thing by itself, they have a "Zinardi bar" in the cars because a driver by that name was injured in a crash because the protection offered by the breaking away carbon fiber was inadequate.

    In short, its a science (backed with empirical testing) to construct a proper failure mode into the composite.

    I would suggest starting with a steel flywheel, it has the potential to be much more durable. If you dropped it, it might be out of balance and destroy your bearings and cause a seizure, but it wouldn't be anywhere near as likely to shatter at RPM and send out shrapnel.
     
  4. Sep 30, 2005 #3
    It was said that the composite flywheel is safer because when it brokes, it would disintergrate quickly and end up with very tiny pieces instead of big chunks of fragments of high speed.

    So what are the drawbacks of using composite for flywheel?
     
  5. Sep 30, 2005 #4

    FredGarvin

    User Avatar
    Science Advisor

    The whole point of a flywheel is to store energy an/or regulate speed via the largest mass moment of inertia required. It has to have mass. Going to a carbon fiber type of flywheel will reduce the weight which would mean less energy in a wheel of the same dimensions. Working with a metal would definitely be my first consideration. A metallic flywheel will exhibit failure symptoms before it would fail catastrophically. Cracks, losseness of the hub, vibrations are all signs.

    What's the reasoning for using a carbon fiber flywheel? What is the application?
     
  6. Sep 30, 2005 #5
    The flywheel is going to maximize the k.e. stored per unit mass. So a high strength to density ratio would be most desired.
     
  7. Sep 30, 2005 #6
    "When the tensile strength of a flywheel is exceeded the flywheel will shatter, releasing all of its stored energy at once; this is commonly referred to as "flywheel explosion" since wheel fragments can reach kinetic energy comparable to that of a cannon shell. Consequently, traditional flywheel systems require strong containment vessels as a safety precaution, which increases the total mass of the device. Fortunately, composite materials tend to disintegrate quickly once broken, and so instead of large chunks of high-velocity shrapnel one simply gets a containment vessel filled with red-hot sand (still, many customers of modern flywheel power storage systems prefer to have them embedded in the ground to halt any material that might escape the containment vessel). Gulia's tape flywheels did not require a heavy container and reportedly could be rewound and reused after a tape fracture."

    From http://en.wikipedia.org/wiki/Flywheel_energy_storage

    Is it true and why?
     
  8. Sep 30, 2005 #7
    For making your own flywheel steel should do fine. You're probably not aiming for an 80,000+RPM flywheel that rides on magnetic bearings. So composites shouldn't be needed.

    The best material, eventually, for flywheels should be carbon nanotubes. They have huge tensional strength, but lousy compression strength. Tensile loads is what a flywheel has, so they'd work great. I'm not totally sure about this, so don't bet the farm, but I've read that nanotube composite flywheels should be able to store energy in equal or greater densities than gasoline.
     
  9. Sep 30, 2005 #8

    FredGarvin

    User Avatar
    Science Advisor

    When metals pass the yield point, they don't immediately shatter. They yield plastically. I guess flywheel explosion is the result of letting one go past the point that point of yield. I have seen turbine disks let go but they had a huge amount of energy in them at those times. Most flywheels don't have that much in them.

    I see where you're coming from. The fact remains though that you would still have to increase the size or angular speed of a lighter flywheel to accomodate the same energy as a more dense rimmed flywheel. That tradeoff would have to be compared to the actual strengths involved (which I am not up on as far as composites). That is why I would still contend that all of the design constraints of the situation will dictate if the use of material will maximize anything.
     
  10. Sep 30, 2005 #9

    PerennialII

    User Avatar
    Science Advisor
    Gold Member

    In relation to the other thread and failure of the flywheel, you can get exact results by using the fracture toughness data and an appropriate stress intensity factor solution (most handbooks have them for flywheels with the appropriate initial cracks)(the fracture mechanical analysis is relatively simple to do).
     
  11. Sep 30, 2005 #10
    um....I don't really understand since I have not really learn much about fracture.

    But can you tell me whether the composite flywheel or the metal alloy flywheel is safer when both are them are broken at high rotational speed? and why?
     
  12. Sep 30, 2005 #11

    Cliff_J

    User Avatar
    Science Advisor

    Please post in one section at a time, the duplicate in another forum is unecessary.

    Composite construction can be very challenging to ensure that the different materials are bonded properly and that the fibers are oriented properly.

    If you are storing a lot of energy, it needs to be treated with a level of respect. The space shuttle was damaged by a peice of falling foam that happened to be traveling fast enough to cause damage. In storms with straight line winds (near hurricane force) in the midwest there will sometimes be wheat stalks embedded in the sides of trees or homes. You may have seen experiments with a soda straw pushed through a grapefruit. So what if the composite chunk of that destroyed flywheel only weighs a few ounces, would you want a baseball or golf ball in flight from a pro athele headed your way, especially if it was sharp and jagged?

    Unless you're trying to spin this incredibly fast, a ductile metal is going to offer a much larger safety margin.

    Even there, research flight 232 that landed in Sioux City SD a few years back, a manufacturing defect the size of a BB resulting in a stress crack after years of service.

    With a composite, one air bubble near a vibratory mode could easily lead to a similar crack and cause the wheel to split into two pieces. Once those pieces hit a solid object they may indeed shatter more easily than a metal flywheel, but you would certainly not want to be the object they hit. Composite golf clubs certainly handle that small impact quite well repeatedly.
     
  13. Sep 30, 2005 #12

    PerennialII

    User Avatar
    Science Advisor
    Gold Member

    Like many wise statements above imply don't believe there to be a single answer, don't think we can get it in much greater accuracy than Cliff gave above when talking about general principles. We'd need to do a bit more thorough analysis to weigh what is truly optimal in the end. Unless your high rotational speed means "extremely high" doing it from a metal is much simpler what comes to design, manufacturing, safety overall etc. The composite one is "interesting" though :biggrin: .
     
  14. Sep 30, 2005 #13
    'Toughness' is generally a term used to imply, or infer, durability, meaning it will last.

    What is the flywheel for? as application tells of the needs of Compressive V Tension, strain and or stress, loading.
     
  15. Sep 30, 2005 #14

    Danger

    User Avatar
    Gold Member

    A thought crossed my mind while I was reading this thread, but I don't know if it would be practical or not. If non-fragmentation upon failure is the primary reason for not wanting to use metal, how about a thick-walled, but hollow, wheel moulded from a structural plastic such as Torlon and filled with lead or tungsten dust?
     
  16. Sep 30, 2005 #15
    It seems that most of you think a composite flywheel is more dangerous than a metal flywheel. But I keep on finding articles in the internet saying the opposite. I don't really know what to do.
    " Furthermore, fiber reinforced composite rotors have been shown to fail in a less destructive manner than metallic rotors -- an important factor for safety reasons."
    from http://www.esm.psu.edu/labs_facilities/cmtc/flywheel.html [Broken]


    by the way, sorry for dulibrating the post in the material section
     
    Last edited by a moderator: May 2, 2017
  17. Oct 1, 2005 #16
    You need to take a materials class IMO. You don't have understanding of fracture modes or fracture mechanics. Understanding fracture means you have to look at more than UTS or yield strength. You have to look at the are under the stress-strain curve. You have to dig up stress intensity factors(K_IC, K_IIC, K_IIC). You have to know how a material is going to fail.

    Another point(mentioned here already) is "flywheels" are stored energy devices. The energy is stored by spinning---as you know---the wheel and retrieving the energy at a later time. It is very difficult to spin a light usable mass(something larger than a model airplane brushless motor) at 50KRPM but it is easy to spin a heavy mass at 1000 rpm.

    Another thing to consider are the costs associated with producing CF perfectly. One small flaw in CF means failure if the part is subjected to significant load. CF race car panels are used because they are light; however, they are not subjected to significant load. A disk spinning at 50K with one tiny flaw will fail while a metal disk with a similar flaw spinning at 1000 rpm is less likely to fail.

    CF ages while metals are stable. The polymers(epoxies) that bond CF do age and weaken. They are not very temperature friendly either so your service environment for CF is limited.

    You CANNOT look at a stress-strain curve and determine if a material will meet your needs. You need to look at the big picture. You need to look at modes of failure before saying a substance is right for an application. Spinning a CF disk in a vacuum chamber is fine and dandy, but putting that disk to work in a real life situation is something completely different.

    Would I use a CF flywheel? No, I've seen CF rims fail catastrophically while a cast aluminum rim would bend and crack it would not fly to pieces under the exact same circumstances.

    In fact, I've seen CF bike frames fail completely(a friend broke a brand-new Cannondale Raven in 2000 while all of the Al framed bikes survived that was $5000 out the window) while my Al frames only began to crack(we bought our bikes the same day and rode together all the time---anecdotal yes).
     
    Last edited by a moderator: May 2, 2017
  18. Oct 1, 2005 #17
    I Remember reading articles on the speeds, materials, fracture rates, etc. of the fan blades in jet turbine engines, similar in nature to the stresses and strains that a flywheel undergoes when the application isn't side loaded, as in a cars' Clutch system.

    As for a CF Bike frame, rigidity sometimes hampers more then it helps, AL has a Flex.
     
  19. Oct 2, 2005 #18
    So is composite material really have a safer failure mode than metallic materials?
     
  20. Oct 2, 2005 #19

    FredGarvin

    User Avatar
    Science Advisor

    There are aspects of the failure modes of fan discs vs. flywheels. Most fan discs fail due to a huge imbalance due to the loss of a blade. I have done many tests like this and will attest that a fan letting go is not a fun thing. It is very destructive. However, the flywheel does not suffer that failure mode. Plus, most flywheels I have come across do not experience the speeds we're talking with a fan disk. Does anyone know of any high speed flywheel applications and their respective energy/speed ranges?
     
  21. Oct 2, 2005 #20
    The articles I had been reading concerned themselves with the idea of the Fan Blades grain boundaries separating simply due to the velocity of the spinning blade, the centrifugal force upon it. They were very expensive metal alloys.

    It is also why, I too, had asked, what was the application?

    A clutch undergoes side loading as well as suffering from centrifugal forces acting upon it.

    Good question I don't know the Most correct answer, but I can tell you I had a cutting wheel on a grinder explode and lacerate my leg in two places, it wasn't a "Powdered" explosion, at 11,000 RPM it broke into several pieces and powdered as well. All that because someone used it as a Grinding Disc, using the face of it to grind a Piece of Metal, not something you do to a Disc that is only ⅛ of an Inch thick.

    I would suspect thought, that a Composite, because it's fibers can be 'arranged' might be manufactured to incorporate better strengths, then a Metal, which is usually cast and subject to whatever defects occur in that process.
     
  22. Oct 2, 2005 #21

    PerennialII

    User Avatar
    Science Advisor
    Gold Member

    Start doing some design calculations and/or thinking about overall specs for the system in question and it becomes possible to to investigate the failure events more in depth. If I'd be pressed for an answer on this I'd answer 'yes' on the basis what I know of composite fracture, but don't believe any general answer like this can be too accurate or useful when actually designing a flywheel.
     
  23. Oct 2, 2005 #22

    Try using Memory Metals for fly wheels, You can read about Memory Metals here: http://www.stanford.edu/~richlin1/sma/sma.html

    and

    http://www.fwmetals.com/resources_specsheets/nitinol.html [Broken]

    The Memory Metal will resist expansion when the fly wheel is under angular momentum.

    Give it a try.
     
    Last edited by a moderator: May 2, 2017
  24. Oct 2, 2005 #23

    Cliff_J

    User Avatar
    Science Advisor

    The carbon fibers and resins are not too expensive, yet finished parts designed to carry loads are incredibly expensive. It has only a small amount to do with the material costs and more to do with the development process to bring those parts to market.

    hanson - to be direct, if you have to ask how to create a flywheel the question changes dramatically to whether or not you can create a safe composite flywheel at all much less in comparison to a metal one. Composites is a science in itself and even when millions of dollars are invested mistakes happen.

    For the flywheel, lets say you started with a billet of T6061 that should be very consistent through the material and if machined well it should be pretty well balanced. With some calcs you could find the expected point of failure pretty easily, you know the density and strength of the material and can easily make assumptions that those density and strength specifications hold true. You might get caught with an oversight on some vibration mode or something else, but should be able to sneak past within the safety factor.

    With a composite, the strength is a combination of the fiber and resin. If you build the flywheel but do not allow the fibers to share the load and instead have the resin/glue supporting the load you may end up with less than 1/10th the strength you think you have. And should you not get the distribution correct, how would you correct a balance issue? Even with pre-impregnated carbon fiber, the correct peel-plys, bleeder plys, and an autoclave it would also require high-quality molds. And we still need to include some quality control to verify the construction was successful, a very expensive process to complete.

    If you have the correct equipment because you have access to an aircraft or race car facility to test the finished product, yes it would be possible to create a composite flywheel that could fail in a less destructive manner after it contacts the safety shields. But it is far more likely that you would create a very hazardous condition to haphazzardly construct one thinking its safer without being able to verify the construction methodology, especially when at the same time its very unlikely you'd experience a metal flywheel failure because you're likely not at a high enough RPM or power level to have that affect.
     
  25. Oct 3, 2005 #24
    Good Answer
     
  26. Oct 3, 2005 #25

    Astronuc

    User Avatar
    Staff Emeritus
    Science Advisor

Share this great discussion with others via Reddit, Google+, Twitter, or Facebook