Trying to track down the fieldside video of this. I was about 8ft from it. Was quite the time.
Will follow up if I find the video.
12 Likes
I did field reset during this match. Here’s what we know:
Another robots intake hit 1023’s air tank. Afterwards, the air tank exploded and there were multiple pieces across the field. Most of the pieces have been returned to 1023, the rest of which we have been looking for. Here’s another picture of some pieces.
I’ll see if I can get more information
3 Likes
Supposedly, 1023 also hit the wall and there was aluminum on the plastic.
1 Like
Frankly, we don’t need rules about air tank placement - we need rules about air tanks. Only plastic air tanks have this explosive failure mode. The stainless steel air tanks that were standard in FRC for years would not fail explosively.
Rather than trying to make an unsafe product safer with subjective rules about robot construction, let’s just go back to only allowing metal air tanks. Sure it’s a weight penalty, but plastic isn’t worth the risk.
29 Likes
Is this (taken from their TBA page) a good representation of the setup?
Reminds me of this: FRC Blogged - Plastic Air Tanks -- Important Safety Notice
1 Like
Yes it is! The one on the right is the one that blew up
8 Likes
I have to disagree. The rules do warn teams not to place tanks in areas where they will be vulnerable and that should be sufficient. Looking at the pictures and the response from @percussionette, it’s pretty apparent that this was not good tank placement. It put the tank in question low on the robot, close to the frame perimeter, exposed to the outside, and backed by a solid frame element of the robot. In other words, it was in a perfect position to be sandwiched between the support it was attached to and extended parts of other robots in a collision with nowhere to flex or move out of the way. Looking at the video, you can see where the extended intake frame on 5712 (an aluminum 1x2 tube) spears straight into it with both robots moving at speed. That might well pierce an aluminum tank too. Robot inspectors need to be more careful of things like tank placement for this reason. Tanks need to either be behind protective elements of the robot or mounted in positions where they can move and/or flex away from the force of impacts. The fact that this kind of rupture is a rare occurrence should tell you that forcing everyone to go back to only metal tanks is overkill as a solution.
That said, this should have been a red card for 5712 for damaging contact inside frame perimeter. But that sort of thing has been called only inconsistently this season. Our own team was once disabled by another team who did receive a red card, even though we were actually not damaged (they hit out main breaker through our electrical panel, a completely freak shot) but also had a pneumatic actuator bent so badly it wouldn’t operate in another match by an impact and no yellow or red card was given. Unfortunately, though the rules try to discourage this sort of contact, they have not been having quite the effect intended.
4 Likes
There’s a clan now? Am i in?
1 Like
Which rules would that be? It’s always best in these discussions to quote the applicable rules so things are clear…
While I don’t disagree with this in terms of general robot design, I don’t find it supported in the rules anywhere currently, and frankly I find it hard to imagine how such requirements could be consistently enforced across events.
7 Likes
This is what is known as a leak-before-burst or leak-before-break pressure vessel design. I’ve done design and testing of pressure vessels for this criteria in the past and, in general, you need a fairly thick wall vessel and high fracture toughness material to meet this criteria. In crack growth mechanics, there is a quantity known as a critical crack size. Before a crack reaches this size, it will grow using fatigue mechanisms (each stress cycle causes the crack to grow a little bit and it takes many cyclic loadings to lead to failure). However, once the crack grows to the critical crack size, then the crack will grow from that point due to overload stresses and will generally propagate to failure in a single stress cycle. The key to designing a vessel that can leak before burst is that you want the crack to be able to grow completely through the wall thickness before it reaches its critical crack size. Thus, the vessel will leak through the crack relieving the pressure (or allowing detection of the leak) before the crack grows to the size necessary for catastrophic overload failure or burst.
I’m not saying that the stainless steel tanks do not meet this criteria, but I would be a little surprised if they did. The pressure in our applications are quite low, so the wall stresses are a lot lower than what we typically think of for high pressure vessels. But even at that low pressure if the wall is thin, a crack in the vessel can still propagate catastrophically once it has been initiated.
In this case, it appears that the tank was punctured during a collision. There is another criteria for pressure vessels which is referred to as non-fragmenting design. This is harder to design for, but there are ways to test a vessel for this. The military has a test for pressurized vessels used in various aircraft where they shoot the vessel with a relevant projectile (typically a 50 cal high powered rifle). This is a very dangerous test and is done in a very fortified test facility with a high level of safety precautions by trained individuals. But if the vessel can be punctured by a bullet without fragmenting explosively, then it is considered a non-fragmenting vessel.
Do you know whether those older steel air tanks were ever tested or certified as leak-before-burst or non-fragmenting? I see that Clippard still sells these Stainless Steel tanks, but there is nothing in their catalog indicating whether these tanks are, in fact, non-fragmenting. Unfortunately, there is also not enough information (such as wall thickness) to perform an analysis for the leak-before-burst criteria.
18 Likes
The tank in question is held on by being ziptied to another ziptie around the other tank, it’s not against the aluminum. There is 3/16ths lexan behind it but if i recall correctly there’s a decent amount of space between them.
It would be helpful for this discussion if you cited the specific rules you’re thinking of here.
You may well be right that there was something about this setup that contributed to the event, but in all fairness this looks like a fairly typical setup. There are probably hundreds of teams with something similar.
I can’t see anything in the Inspection Checklist that would have flagged this setup.
12 Likes
Yup. Our setup was pretty similar, until we had tanks get ripped off mid-match (with fitting damage, we replaced them all and added a shield).
This mid-match impact failure seems fundamentally different from the failures of previous generation tanks that led to the “white Clippard ban”, because it occurred in the context of match play as opposed to outside of it.
I don’t care about robots disassembling on the field, I care about robots disassembling while students faces are near or inside them. My understanding is the ban was related to spontaneous disassembly off-field.
1 Like
We had mentors and students hit with shrapnel (fortunately no serious cuts or scrapes) up 5-7 rows from the bottom of the stands.
I absolutely care if robots disassemble themselves on the field.
25 Likes
Thank you, I shouldn’t have spoken so confidently before knowing exactly how well- or not-well-contained the incident was.
9 Likes
Would you expect a sleeve like 9284K618 to contain fragments of a tank like the one we are discussing, in a similar episode?
4 Likes
“Should be sufficient” is fine for things that don’t become literal bombs when teams make mistakes. There’s simply no good reason to leave in this potential for disaster in FRC robots. “If they did it right, it would be fine” isn’t a great attitude toward having high schoolers building machines handling potentially explosive objects. It’s not as if there aren’t air tanks that don’t explode in this manner - the failure mode of a stainless steel tank is much, much safer - far less shrapnel.
While everyone was fine this time, when someone is seriously injured in a completely preventable way, when a safer alternative exists and is readily available, do you want to be the person to tell that injured individual “hey, this doesn’t happen very often, that’s why we agreed to expose people to this risk”. Are the pounds of robot weight worth it?
I am not saying stainless steel tanks will all leak before burst. When stainless steel tanks burst, they don’t shatter into dozens of tiny pieces that launch in all directions. A crack generally forms which ruptures, tearing some of the metal apart, but generally not launching fractured pieces of that metal away. This failure mode is basically inevitable with plastic tanks, and it is not inevitable with steel tanks. I’m not saying stainless tanks will never fragment, but plastic tanks will more or less always fragment on failure.
15 Likes