I’ve been thinking quite a bit about FoS in FRC with respect to both the design of COTS FRC products as well as team implementation of said products.
Most COTS FRC products can be pushed well beyond their marketed limitations, and they often are, usually by less experienced teams. Given that the majority of the market for these products consists of these inexperienced users, I would imagine that the producers take this into consideration when determining the listed limitations for their products, and the resulting factor of safety. I have no idea if I’m on the right track or not, but I would love to be corrected by someone who knows what they’re talking about.
I’m also curious about how teams apply their own factor of safety for some specific FRC products. The Versaplanetary gearbox in particular is an interesting example. As many/most of you know there is a very well defined load rating guide for the Versaplanetary that says whether or not a certain gear ratio is acceptable for a certain motor. Now I’m guessing VEX has some kind of FoS built into that guide (maybe not?), but I’m wondering how teams use that guide to influence their designs. I would guess that most teams follow the guide directly, if it’s green it’s good and if it’s red it’s no good. Do any teams out there leave themselves some extra margin for error? How have your decisions regarding Versaplanetary ratios affected your VP performance/durability throughout the season? I’m going to go ahead and guess that in most cases the VP has outperformed expectations. There’s a reason there are like 4 or more of them on every competitive robot. How do teams design with a factor of safety in mind in general? What are some FRC products that you think are overbuilt? Underbuilt?
I think this is a great thread topic. I bolded the phrase above to highlight what I consider a common error.
Confidence that a product will continue to function as expected under loading conditions calculated based on markings or marketing claims is generally not a good practice, and has led to many disappointments, unless the markings or claims are based on reproducible test results, using methods that have been standardized by an authority other than the product’s manufacturer.
In most cases, the better practice is to base confidence on your own engineering analysis, using the best knowledge available about the product’s materials and construction. In other words, do the math yourself if you need to rely on the results. And get the best data you can, both on the product’s design, and on how it performs under the loading conditions that matter to you.
In general, for products designed for or marketed to FRC teams or hobbyists, we usually apply the numbers conservatively (that is, we’ll go straight up or try to buy something with excess capacity). For products created for the general market, we’ll run a bit more aggressively (50% or occasionally even 100% over manufacturer specs), knowing that we’re planning to run our robot for dozens vs hundreds or thousands of hours. Don’t forget to spec parts based on peak or shock loads, not steady-state!
All of the mechanisms in FIRST that I have worked with that had very small factor of safety built in, we battled it the entire season to keep it functional. Most of it stemmed from incorrect loading calculations, or “that should work” without checking the numbers, or lack of prototyping. Show me or list things that are under built, and I will show you a season long project.
My factor of safety is “how quickly can I replace it if it breaks?”. If the answer is “under 6 min” I don’t worry too much about it just try and make it durable enough to last a match. If the answer is “more than 6 min” I beef it up to a point I feel it won’t be broken in a full season of matches.
FRC FOS is kind of a lesson of time, the longer you’ve been involved the more you feel comfortable sacrificing in the name of weight savings.
The VP gearbox is an interesting case study. It is entirely possible to follow all of the load guides and mounting instructions perfectly and still break them. They can break when the mechanism they are driving is slammed into something, feeding more torque back through the gearbox than the motor by itself could supply at stall, but that the friction in the gearbox AND the motor can resist.
This leads to another point about safety factors in FRC: the loading speed. It is not usually practical to calculate all of the dynamic loading cases a mechanism could be exposed to.
In highly dynamic loading cases, such as drive chain, I aim for a FoS of 4 or greater using some of the many calculators available.
In more controlled uses of chain and belt I aim for a FoS of 2 or better.
For VP gearboxes and similar I rely on the manufacturers loading tables/ratings with a keen eye towards preventing over-driving from outside sources.
I aim to keep motors well under ‘dangerous’ current levels under stall conditions. Locked rotor stall test data is great for this.