Polycarbonate vs. Aluminum

[WSiggs]

My team is currently in a debate over whether we should use polycarbonate or aluminum for an intake which extends past the bumpers of our robot. The concern that some members of the team have is that 1/4" polycarbonate would crack or shatter if impacted by other robots. Teams that have had experience with using polycarbonate (or other plastics) structures that extend past the bumper and would be susceptible to impacts, how has it gone? Did your drivers need to exert caution when driving or could the plastic hold up to sudden impacts?

[Knufire]

See the original intake from the 111 2012 robot or the 1678 2014 robot for examples of polycarbonate intakes.

The idea here is that you want a material that is tough, not strong. While strength is a measure of the amount of stress a material can take before failure, toughness is a measure of how much energy the material can absorb before failure, and is a mixture of the strength and ductility of the material. You want the polycarbonate intake to deflect when it is hit instead of breaking, but this naturally means that your intake won't be stiff. On the other hand, you can have a very stiff aluminum intake that will always be in the same spot relative to your robot frame, but you will then have to build it to be strong/tough enough to withstand impact.

[BrendanB]

Originally Posted by Knufire

See the original intake from the 111 2012 robot or the 1678 2014 robot for examples of polycarbonate intakes.

The idea here is that you want a material that is tough, not strong. While strength is a measure of the amount of stress a material can take before failure, toughness is a measure of how much energy the material can absorb before failure, and is a mixture of the strength and ductility of the material. You want the polycarbonate intake to deflect when it is hit instead of breaking, but this naturally means that your intake won't be stiff. On the other hand, you can have a very stiff aluminum intake that will always be in the same spot relative to your robot frame, but you will then have to build it to be strong/tough enough to withstand impact.

+1

Definitely take a look at their After the Game video on their 2012 intake. Stacking up lexan plates will give you strength and flexibility and are incredibly easy to make replacements of if you break them and you can stack up more plates to add strength and rigidity.

[RoboChair]

Originally Posted by WSiggs

My team is currently in a debate over whether we should use polycarbonate or aluminum for an intake which extends past the bumpers of our robot. The concern that some members of the team have is that 1/4" polycarbonate would crack or shatter if impacted by other robots. Teams that have had experience with using polycarbonate (or other plastics) structures that extend past the bumper and would be susceptible to impacts, how has it gone? Did your drivers need to exert caution when driving or could the plastic hold up to sudden impacts?

Polycarb, 100%.

Our intake arms for our 2014 robot were 1/4 poly carb and in more than 1 match ended up getting caught on another bot and dragged them around the field by our intake. Those arms lasted for at least 150 matches, they finally died after being exposed to UV and then the heat of our crate during our last offseason event. Polycarb does not shatter like acrylic. Aluminum will just die this year.

11/10 Highly recommended.



Ask me your questions.

[JamesCH95]

In contrast to RoboChair...



Our 2014 robot used 1/4in aluminum plate for its intake and we had zero issues with it all season. We too got our intake caught up in other robots that year and never broke it. The air cylinders used to deploy them were about the smallest air cylinders we could use, so they collapsed neatly when hit. And we got hit a lot when opening up to catch a ball.

It doesn't matter what material you pick, it matters how you execute your design.

[bEdhEd]

I recommend polycarbonate. Any mechanism extending out of our bumpers (within volume constraints of course) are polycarbonate. Polycarbonate rarely cracks or breaks on impact in standard conditions. The only time we have ever cracked polycarbonate was in 2015 with our polycarbonate tote intake. The only reason we thought it shattered on impact was because loctite was used on the screws that attached our motors to the polycarb plate. We think that the loctite reacted with the polycarbonate during its assembly and made it brittle instead of flexible, and that's just our hypothesis. The only areas where there was evidence of shattering was in our motor mount holes. After avoiding loctite on the rebuilt intake, it never experienced any cracking after several more regional and championship matches.

Our lesson was to be careful of what possible reactive chemicals come in contact with polycarbonate. And DON'T use acrylic. That's guaranteed to shatter.

[PayneTrain]

Originally Posted by RoboChair

Polycarb, 100%.

Our intake arms for our 2014 robot were 1/4 poly carb and in more than 1 match ended up getting caught on another bot and dragged them around the field by our intake. Those arms lasted for at least 150 matches, they finally died after being exposed to UV and then the heat of our crate during our last offseason event. Polycarb does not shatter like acrylic. Aluminum will just die this year.

11/10 Highly recommended.



Ask me your questions.

How did those VP mounts hold up?

[Michael Corsetto]

Originally Posted by PayneTrain

How did those VP mounts hold up?

We didn't have issues with it, besides being difficult to assemble.

The key was 2 bearings in the VP mount tube (one on each side of the tube), and then a third bearing on the other polycarb arm. This meant that, even as the arms twisted, the C-C for the chain was held constant. But since the polycarb arm is flexy, the intake shaft isn't over-constrained, even with three bearings (as it would be if all three bearings were held constant relative to each other).

Hope this makes sense.

Best,

-Mike

[Chris is me]

Originally Posted by Michael Corsetto

The key was 2 bearings in the VP mount tube (one on each side of the tube), and then a third bearing on the other polycarb arm. This meant that, even as the arms twisted, the C-C for the chain was held constant. But since the polycarb arm is flexy, the intake shaft isn't over-constrained, even with three bearings (as it would be if all three bearings were held constant relative to each other).

Just to go on a tangent here, but any time you have a long shaft like this (>2 feet between supports), you can sometimes get away with more than two bearings per shaft since the span is so long. It's not something that can be only done with polycarbonate or anything like that, especially if 2 of your 3 bearings are in pockets milled into the same tube. Just another fun fact to add to the conversation here.

[JamesCH95]

Originally Posted by Chris is me

Just to go on a tangent here, but any time you have a long shaft like this (>2 feet between supports), you can sometimes get away with more than two bearings per shaft since the span is so long. It's not something that can be only done with polycarbonate or anything like that, especially if 2 of your 3 bearings are in pockets milled into the same tube. Just another fun fact to add to the conversation here.

Confirmed. Here are 4x bearings on the same shaft!

[Michael Corsetto]

Originally Posted by Chris is me

Just to go on a tangent here, but any time you have a long shaft like this (>2 feet between supports), you can sometimes get away with more than two bearings per shaft since the span is so long. It's not something that can be only done with polycarbonate or anything like that, especially if 2 of your 3 bearings are in pockets milled into the same tube. Just another fun fact to add to the conversation here.

Very true, key word being "sometimes"

[Cothron Theiss]

This is a pretty significant tangent, but it's not worth making a new thread over. What actually happens when you overconstrain a shaft? Does it automatically bind? Is it somewhat more prone to binding? Is the resistance (load) substantially increased? I know not to do it, but I don't know that actual effects of doing it.

[Michael Corsetto]

Originally Posted by Cothron Theiss

This is a pretty significant tangent, but it's not worth making a new thread over. What actually happens when you overconstrain a shaft? Does it automatically bind? Is it somewhat more prone to binding? Is the resistance (load) substantially increased? I know not to do it, but I don't know that actual effects of doing it.

If you have a significant gap between bearings and good manufacturing/assembly tolerances (as in James' example above), you won't see much negative impact to performance.

Some potential effects:

1) You won't be able to install the shaft. This is an easy problem to spot. You have three bearings, try to slide the shaft in, and no dice.

2) You figure out a way to install the shaft, with a axially mis-aligned set of three bearings (maybe you constrained the bearings to a rigid member AFTER installing the shaft). In this scenario:

2a) The shaft rotates, but with additional load due to the shaft constantly bending/additional radial load on bearing. This inefficiency can sometimes be hard to spot right away, but can suck power from your system

2b) The whole thing seizes up and won't move. I've seen this happen, especially when the torque applied to the shaft is fairly weak (obviously, since the system is more susceptible to inefficiencies).

Like Chris said, do it right, and you won't have issues. But people don't always do it right, so it is important to understand where things can go wrong.

Best,

-Mike

[s_forbes]

Polycarbonate is tough stuff, and will likely survive robot-to-robot interaction. It usually bends out of the way and springs back to shape. It's more likely to get bent or tear if you overload it, rather than cracking into pieces like other hard plastics.

Don't let loctite get near it though!

We are designing an intake that goes out past the bumper this year, but are planning on using aluminum with pneumatics that take the impact load, similar to 95's robot above. Both methods are feasible.

[bEdhEd]

Originally Posted by s_forbes

Don't let loctite get near it though!

Ok, hypothesis confirmed haha.