Team 100’s 2020 FRC shooter uses 60A durometer 4" Fairlane neoprene (chloroprene) rubber wheels, McMaster-Carr part number 2476K37. There is a question whether the wheels can be spun at the speeds necessary to launch power cells (foam balls) in the Infinite Recharge game without over-stressing the wheels. Accordingly, we have performed a axisymmetric stress analysis to calculate the radial and hoop stresses and the radial displacements in the rubber and steel.
The analysis shows that radial stresses in the rubber are highest at the interface with the steel core, and are maximum 300 psi at 9,500 RPM as compared to a tensile strength at break of >1723 psi for the chloroprene referenced in the paper. At 9,500 RPM, radial strains in the rubber at the interface are 27% as compared to an elongation at break of >287%. Hoop stresses in the steel are maximum 4,000 psi at 9,500 RPM as compared to 36,000 psi for garden-variety ASTM A36 steel. The wheel is expected to expand <0.25" at 9,500 RPM.
This analysis does not at all support the wrapping the specific 60A-Durometer Fairlane rollers with stainless steel “safety” wire. The analysis indicates that stresses, strains, and displacements would not exceed values that the rubber would be expected to withstand, with considerable margin, provided the mechanism providing the rotation itself was robust and maximum speed was controlled.
This analysis carries certain limitations. It assumes that Seals Eastern neoprene parameters are applicable to Fairlance 60A rollers. We would not expect significant differences if we had the actual Fairlane neoprene properties. This analysis also assumes that the rollers are free of defects, including manufacturing defects or holes drilled for “safety” wires. This analysis is only applicable to the specific geometry, materials, and use conditions modeled. Any use of this analysis is at the user’s sole risk.
I may be misunderstanding, but it doesn’t look like you are doing an analysis as to whether the neoprene will delaminate from the steel core. I believe this is one of the primary reasons that teams should/did wrap their shooter wheels in safety wire.
Michael, we did do that analysis. Delamination is driven by the radial stress at the rubber/core interface. That comes in at approximately 300 psi. Delamination, i.e., development of a circumferential separation, is a lot less concerning to me than development of a radial separation due to hoop (circumferential stresses). A radial separation could lead to flying rubber pieces.
I didn’t mean to suggest that you didn’t analyse it (you did a great job). I meant to say that we don’t know what the delamination stress would be, and thus can’t make a conclusion as to whether a wheel at these speeds would delaminate.
I completely agree that radial separation is the more dangerous failure method. But your conclusion is that “This analysis does not at all support the wrapping the specific 60A-Durometer Fairlane rollers”. I’d argue that this analysis proves that the Fairlane rollers will not radically separate, but that doesn’t necessarily mean that they are safe or safer to be run without safety wire.
Michael,
There is an unsupported premise that “safety wire” is in fact safe. Find me a licensed PE who will sign off on the strength of a twisted stainless steel wire. What is the safe working stress for that? What is the effect of drilling a hole in the Fairlane to hide the twist? What is the rubber/core interface stress when a wire is installed? Until someone presents those analyses, there is no engineering basis for using “safety wire” on 60A 4" Fairlane spun at up to 9,500 RPM.
RKlopp
This is changing the goal posts. For one thing, licensed PEs are uncommon in most engineering fields other than Civil Engineering. In particular for mechanical engineering, licenses are incredibly uncommon even for medical and aerospace applications.
All of the questions you brought up are important questions. But until someone does an analysis, we also can’t say that these methods are unsafe. When it comes down to it, engineering is also an experimental field and there has been plenty of success shown in correctly installed applications (i.e. 254 and 1678 in 2016).
Finally, it still is going to be a subjective decision, based on the subjective, relative consequences of delamination and radial separation. I don’t think there is going to be an objectively “correct” decision.
Will use of Fairlane wheels be limited to teams that have mentors capable of boundary value problems? This is cool analysis (I’ll have to dig more into this in the offseason), but I don’t expect many people to follow it or for it to work as proof that fairlanes or safety wire are safe or unsafe or basically anything.
FIRST designed a game that encourages teams to use a ton of energy to shoot balls and didn’t mention safety of shooter design once in any video I’ve seen. Teams will highly stress components and high school students may or may not know how dangerous this can be. It seems FIRST has gotten really lax with the rules / safety aspect of this and has taken the approach that the it’s up to the LRI (who has basically no training or expertise on the matter and is not legally representing FIRST).
Don’t want to derail this thread, but If a flywheel explodes this year and someone gets hurt - what happens when the lawyers get involved? Is FIRST on the hook? Is the LRI? A PE wouldn’t sign off on any of this.
I’ve had the same thoughts over here. Our initial design using a vexpro wheel spun the tread right off the wheel above 6k RPM or so. Switching to injection molded andy mark wheel that has the tread glued to the rim has better results, but we make sure no one is near the robot when we spin up those wheels. In point of fact, some people were asking what RPM the common hex bearings FRC uses are rated to as well. And people using flywheels have another potential failure point. Seeing a spinning flywheel get free would be a nightmare scenario.
The paper and the analysis thereafter imply that we can use stainless steel wire to wrap our Fairlane wheels for any reason, so long as it isn’t used to “increase the safety of the system”. In other words, safety of wheel delamination is not a determining factor when it comes to wrapping a wheel with untwisted stainless steel wire. So as long as the wire itself is secured to the wheel, and the wheel stays under 10k RPM via software, there aren’t any safety concerns.
The analysis doesn’t indicate that the wheels won’t delaminate. In fact, it indicates that the highest radial stress is placed at the steel-neoprene interface, of which we don’t know the strength of. Further, in practice teams have had problems with delamination running wheels at these speeds. Delamination is a safety concern that should/must be addressed.
In practice, safety wire (installed correctly) has been show to eliminate delaminations, while being safe according to your other metrics.
This is a bigger assumption than you’re giving it credit for.
Are you running your assembly on 1/2 hex slip fits on COTS 1/2 hex bearings? I haven’t tried to measure the runout on our 1/2-hex-based assemblies, but they sure as heck aren’t balanced.
(On an early wooden prototype, at around 9k-10k RPM we ended up expanding our bearing mounting holes during operation by almost 1/8" due to the unbalanced load “rattling” against the bearing bores… which led to unplanned disassembly)
Is it? There might be a “big bang”, but as long as the wheel doesn’t have unplanned radial disassembly in the process of delaminating I don’t see this as being particularly less safe than, say, lathe operation…
@RoboChair has delaminated at least one in testing - have you seen radial disassembly?
Part of the problem is that if the wheel delaminates, the analysis presented in the paper falls apart.
I think this is what you were getting at, but the “worst case” in my mind being that only a portion delaminates and the wheel continues spinning, but the internal stresses holding the wheel together disappear and the axisymmetric assumption becomes false. Under those conditions the wheel could very rapidly, radially fail.
Your statement is blatantly false.
The paper analyzes the unmodified product and does not address the modified condition.
This is clearly stated in the paper.
I am a ME. I let my PE license lapse several years ago for several reasons. One of them being in the 20 years or so that I had my license, I never had a need to stamp a drawing. Anyway. A lot of things done on an FRC robot that you would not get a PE to sign off on. (A PE cannot just “sign & stamp” a drawing. By doing so the PE is certifying they have fully reviewed the design and engineering calculations. In fact there is a certain legal liability for a PE to be in the same room as a FRC robot.
This is not a credible comparison to demonstrate that delamination will not occur.
The interface between rubber and steel might be glued/bonded, but even with high end 3M material it’s difficult to reach 300psi radial strength off the datasheets alone.
I have a note in to 3M asking about tensile values, will post back if they have any suggestions for me.
Has Team 100 spun a fairlane wheel to 9500RPM to verify the analytic results in this paper?
This is a credible comparison to demonstrate that radial disassembly will not occur, with almost 6x FoS. The analysis doesn’t fall apart by 6x.
Combined with Robochair’s datapoints where delamination does not result in radial disassembly, I do not expect Fairlane delaminations to result in radial wheel disassembly. I would not treat a Fairlane under 9500rpm to be an inherent safety risk above and beyond the normal realms of FIRST Robotics risks.
The conversation after the paper is what I was referring to in regards to safety about wire, or any, wrapping.
Some of us just want to increase the consistency of our shots at 4k RPM, but are concerned that we will get extra unjustified scrutiny because some teams do not know how and/or do not care to tune their flywheels for the lowest-possible RPM with repeatable results. For example, suppose I want to 3D print a TPU flanged capture disc that ensures the edges of a Fairlane wheel maintain a consistent circle as the center bulges. Neither this paper nor the other discussions provide any insight into whether that is or is not safe within the realm of FRC. So all we can say to the inspector is “it has shown no signs of issues so far”. The same exact question could come up with any design that has a custom-machined flywheel, like 610, fwiw.
Simply due to the conversation, lack of lay terms in explaining the results, and the outright bickering of “both sides”, I fully expect any inspector to have the right under the rules to rule it unsafe even if it is safer than every other mechanism on the same robot. There aren’t measurable boundary conditions for safety here, and even the paper above provides little to no guidance when put into perspective of the very large variety of high-RPM wheels inspectors will see this season.
I have never seen radial disassembly OR a full delamination of any hub and we have delaminated a lot of wheels in testing over the years. The failure mode is always things were operating smoothly and then it started vibrating a lot more than before. That is it, just very aggressive vibration caused by the wheel no longer being balanced at speed because one side of the wheel delaminated in a single location allowing that side to expand much more than normal.
