So we’ve been designing our shooter for this year, and it includes a relatively large flywheel on an overdrive from the shooter wheel axle. Right now our flywheel is around a 1.4:1 reduction from 775pros, putting it at about 13k rpm. With a MoI of ~0.0023 kg-m^2 that gives it a rotational kinetic energy of 1500 J, or about the same kinetic energy of a bullet fired from an AR-15.* That scares me.
We are probably going to send outsource the manufacturing of ours to our sponsor who makes airplane parts. Our original plan was to make it in-house, but I assume that they can probably hold better cocentricity tolerances than our school lathes which should help minimize vibrations. I also want the whole flywheel assembly to be added to the pre- and post-match inspection list to verify nothing is broken or coming loose.
I’m wondering what teams are doing to make sure their flywheels are attached and used safely.
* This is lowered from our original design, which had a 2:1 overdrive from the 775pros, putting the flywheel at 30k rpm. That would have given the flywheel a kinetic energy of 13kJ, or about the same as an anti-tank rifle. We changed that after being pretty scared and also realizing that we were exceeding the bearings’ speed rating by almost double.
Some of the speeds that people are quoting are getting pretty scary. Also, there was a technique quoted about wrapping a flywheel with safety wire to keep it intact. While I’m sure this works reasonably well, it does seem to be pushing boundaries of safety when you are relying on a secondary method like this to keep things intact.
The basic analysis for hoop stress is easy too look up and relatively straight-forward to calculate for simple rings and simple cylindrical disks. You will probably find that the material, in theory, has a decent safety margin. The problem is really how well do you trust the manufacturing of the parts. If you are dealing with a metal wheel, you can probably be reasonably satisfied that the material is decent quality without defects, but other types of wheels like plastic molded wheels may not be as high of quality. Small defects can significantly reduce the strength of the material relative to it’s theoretical strength.
There are ways to assure yourself of the safety of the system. A test to something like 120% or more of the maximum speed would provide adequate assurance of the strength of the wheel. Obviously, that test would be done in a safe enclosure of some kind that would protect personnel from any failures. 120% overspeed testing is the standard that is used for jet engine rotors.
As far as balancing the rotor. This is obviously important. Most of the wheels that are commercially available would have a decent balance. But if you are joining things together in assembly, then this can be a problem. It is also relatively easy to test and doesn’t require high speeds or dangerous setups. Mounting the wheel on a reasonably lightweight structure, any imbalance, even at lower speeds, will result in tactile vibrations. Correcting it is another matter. It is possible to balance wheels, but you need some fairly sophisticated balancing equipment to do so. The best bet would be to build your wheel assemblies carefully making sure that you keep things concentric and co-axial as you go.
I agree with you, it is a little scary. So please be careful.
My 2009 Subaru Outback’s curb weight is about 3500 lb [1590 kg]. When I drive at 80 MPH [36 m/sec] its kinetic energy is about 1 MJ [one million Joule]. That should probably scare me more than a bullet from an AR-15, but it doesn’t. The bullet is scarier.
Back to your flywheel: I think you are wise to have it machined by professionals, especially aviation industry professionals. I bet it comes out well-balanced and purrs like a kitten at speed. I am concerned that some teams will not have the resources to do that, and may show up at FRC events with poorly balanced high speed flywheels that are potentially dangerous [see S1].
Robot inspection might catch that, but right now I am not quite sure how. Way back in 2006, we tried to test shooters; test procedures and their sometimes arbitrary application were the subject of some heated discussions, here on CD and in the real world.
The massive increase in available power on FRC robots*** combined with the incredible competitive push is worrisome from a safety standpoint. A lot of people don’t know what they don’t know, and even pros get it wrong sometimes. At some point someone is going to get crippled or killed by a catastrophic robot failure, either in the pits, on the field, or in their home shop.
***yeah, yeah, technically battery output has remained pretty static for a long time, so technically maximum power output has been stable…but now people with little to no experience doing so can rather trivially put two horsepower on a ten pound flywheel if they really want to, and that’s both relatively new and downright terrifying.
Backing up to the OP’s ask about what other teams are doing for flywheel safety, here’s a summary of the safety aspects of The Spring Konstant’s 2020 launcher design.
One of the first things to demonstrate is that the flywheel won’t fly apart. Flywheel material selection is therefore quite important. Our flywheel is brass, 4 inch diameter, and will top out at around 12k RPM. We used a 10x safety margin on the surface stress due to the centrepital force (or centrifugal inertial effect, if you prefer) when selecting this.
There’s of course the mount, which is bearing based with reinforcement plates so the bearing outside race is fully supported. We’re tapping the ends of the shafts to secure it in the bearing assembly with loc-tite-ed (and maybe even nord-locked) cap screws. We plan to have spacers only with no shaft collars between the bearings. That’s all surrounded by several aluminum standoffs and some aluminum and polycarbonate guards to contain anything that may fly loose.
Of course, we’ll be thoroughly testing this under many conditions.
We had a flywheel (separate from the shooter wheel) in 2012. Nothing as extreme as what some teams are using these days, but we did take safety precautions all the same. Supported the shaft properly, secured everything in place, and put relatively thick polycarbonate shields around it to contain anything that might come loose (and to keep fingers away from it while running).
I can’t help but compare this to the high powered catapults in 2014. Those were scary in a similar way, with teams taking all sorts of safety precautions around them. Do everything you can to ensure things will stay together under every circumstance, and if it does come apart to contain the pieces so they don’t hurt someone.
We did our prototyping with a flywheel we had from an off-season project that’s about 2x the MoI but spinning at the same speed as the shooter wheels (as opposed to here where it’s geared up a bit). We want to be able to shoot all 5 balls in rapid succession without having to wait for the wheel speed to recover, so we need a significant amount of inertia in the system. Unfortunately we didn’t get a chance to prototype with different size flywheels, so we decided to err on the side of more inertia and slower spin up time than having to wait for recovery time. At first we took this a bit too far, but we realized that yesterday and changed some ratios to be a bit more reasonable.
To be fair, cars really should scare us a lot more than they actually do; the only reason we don’t notice how dangerous they really are is because we’re desensitized to them. Cars kill more than a million people per year. And I’d assume most of those deaths are accidental (due to operator error or poor maintenance) as opposed to guns who generally kill because someone wants them to. 1MJ of energy is nothing to scoff at.
To give a more likely scenario, let’s say a piece of brass the size of a gold dollar delaminates from the flywheel and shoots off at tangential velocity. That piece would be traveling at ~100 m/s and have an energy of 160 J, about equivalent to a civil war rifle. That could do some serious damage if it ends up hitting someone
I feel like I want to point out the vast difference in the amount of safety equipment and technology in the average modern car vs. the amount of safety equipment in the average gun. I feel a greater sense of trepidation driving my 1972 Land Cruiser around than I do my 2015 F150.
The false sense of security, “feel a greater sense of trepidation” is different from the actual physics. Hence the public drives faster, takes more risk and has less skill in driving than people 50 years ago. Cars park themselves, back themselves up, have 50 airbags, ABS brakes, automatic braking etc, but the physics have stayed the same. The gun does not need any more safety added to it than a kitchen carving knife, or a drill does. You pick up the tool for a reason. What safety does a drill have? That could drill a hole in a skull and easily kill. Plenty of power, energy and cutting ability there, but we don’t ever put drills near our heads. That’s the only safety it needs.
We simply plan to not spin ours that fast. Our range is 3200-4300 RPM right now.
If you are so concerned about safety while spinning something so fast, it might be wise to ask if you need to. Slowing down may change your gameplan, but if it makes your design safer it may be worth it.