To pneumatic or not to pneumatic

My team has never run pneumatics before. We have always found ways around them. It just seems like adding another layer of things that could go wrong, but I don’t really know pneumatics well. Our darts this year also caused many issues at our first event, which is the main reason I am asking this now.

My questions are, how significantly does running pneumatics help teams? And how difficult is it to do pneumatics for the first time?

1 Like

You guys should really be running them next year. There are so many times that I see teams engineer their way around pneumatics, only to find out their workaround is much more costly in terms of weight, and engineering resources, for what ultimately proves to be a less robust solution.

Intake deployments are almost always best done with pneumatics, and are the most common recurring example I can think of.

15 Likes

Pneumatics are intimidating. I’ve had plenty of students shy away from them in the past because they don’t truly understand them. There’s definitely a learning curve involved into both how to create a functional pneumatics system as well as designing systems (particularly linkages) to use them effectively. There’s also a lot of initial investment into getting pneumatics onto your robot. You need air storage, a compressor, pressure regulators, and gauges before you can even worry about putting a solenoid valve and pneumatic cylinder on your robot.

But pneumatics open up a massive design space to teams, and once you’ve mastered the basics you’ll never look back. Spend some time this off-season reading the FRC manual to get an idea for what is involved in a pneumatic system, read up on some of some pneumatics manuals (note - some rules changed since this was written), and build a pneumatic system to practice with. That system doesn’t have to be on a robot, it can be on a piece of corkboard or plywood, it’s just to teach you what each component does and how they interact with one another. As a bonus, it will let you do all sort of pneumatic prototyping and testing in the future.

As a design principle, my $0.02 is - if you’re only doing “one thing” with pneumatics you can probably avoid a pneumatic system on your robot. But if you need to do 2 or more “things” with linear motion or pneumatics, then it’s worth installing a pneumatic system. The biggest design cost of pneumatics is the initial system installation, but once that’s already factored in, adding additional solenoids and cylinders is relatively minor and allows you to add plenty more features and movements into your robot for minimal cost. It also allows for motions that don’t occupy additional slots on your PDP/PDH.

7 Likes

We didn’t run them for years because we weren’t familiar with them and so avoided them. Under FRC restrictions, this was a mistake.

Play with pneumatics in the off-season. You’ll find they’re pretty useful.

4 Likes

During my 4 years in FRC, we only used them once, and it was my senior year. If I was still in FRC, I’d probably still be using them.

We always shyed away from pneumatics since weight was a concern, as well as potential leaks being a big issue. We messed around with them in the off season prior to 2019, and decided to run them during season. We had a PTO gearbox, intake actuation, as well as a claw for cargo and hatches that also used pneumatics to extend. We later added in the season some mechanical pneumatic brakes for our climber, as well as some extra force to extend another part of our climber; so across many aspects. Out of 2 districts, state and world’s, we never lost a match due to pneumatics, and I truly believe the design of robot we made could not have been done without pneumatics.

A big thing we considered when laying out our pneumatics was that leaks only really occur at joints/ connections. Less connections, less leaks. Rather than having a lot of small tanks for air, we used a 1 gallon aluminum tank from ARB so we minimized a lot of connections there. We also used manifolds for solenoids rather than individual solenoids.

The only two issues we had with our pneumatics that season was our pressure relief valve, and our pressure switch for the PCM. We tried used a pre-set pressure relief valve from McMaster Carr, but learned that for some reason after blowing once (which is required during inspection), it would continue to blow at a pressure lower than the cutoff. Our pressure switch for the PCM also broke a wire at world’s, and took us a second to realize that was our issue.

1 Like

Being able to work with pneumatic components is a good tool to have in your toolbelt when it comes to your robot design. Some years you may have a lot of little parts that need to move, and being able to just stick another cylinder somewhere is convenient. Other years you might have a simpler design with few moving parts, in which case you might not need pneumatics at all.

That said, over the past few years I’ve been preferring to omit pneumatics whenever possible. I’ve become a fan of the light-fast-simple robot; less parts and less weight and less stress and more fun.

Another thing to keep in mind is that all of the work that your pneumatic system does comes from the compressor, which is not a super powerful motor or efficient method of storing energy. I would avoid using pneumatics for major components that need to do a lot of work. I’ve worked with three robots that used lots of air for major components (1726 in '08, 842 in '14, 842 in '16. All catapult robots). Each one has had a point where we were starving for more air in at least one match.

Try adding a pneumatic system to your robot this summer to get experience!

1 Like

Last robot (2020/21) we had a motor/VP/chain-n-sprocket deployed intake. During the season virtually every part of that broke, some bits multiple times. It also required a bunch of fiddling from the controls team. This robot we have a pneumatically deployed intake. It has been soooo much better. And having pneumatics already on board meant adding a locking pawl to our climber was dead easy.

Give them a try — having another tool in your “design catalog” is a good thing.

1 Like

Pneumatics can be intimidating. There is a whole new set of things to understand (from basic plumbing techniques to programming) that do take a while to figure out.

There is a layer of “overhead” associated with adding a pneumatic system to the robot such as the compressor, storage tanks, regulators and manifolds that carry a certain cost and weight. However, once you have invested in that overhead, each new actuator that you add to the robot comes with minimal cost and weight such that you quickly “pay for” that overhead in terms of the overall robot cost and weight. If you are only going to actuate a single action with the pneumatic system, it may not be worth it (depending on how easily you can perform that same actuation using motors, gearboxes, etc. But by the time you have 2 or 3 or more actions being done using the pneumatic system, it is pretty much going to be worth it.

Pneumatics have limitations in terms of the types of actions that they are good at. But for linear actuation between two fixed points, pneumatics tend to be a much better choice than motors. by utilizing bell cranks or linkages, it is easy to turn that single linear actuation degree of freedom into a lot of useful types of motion.

Leaks are a reality that you have to be prepared for. But, just about any system has “gotcha” aspects that you need to be prepared to address, so there is nothing more or less tricky here. It is just a new set of “gotchas” that you need to address. There are several “best practice” ways to manage leaks in the system. First off, build up your core on a bench before mounting it to the robot to make sure everything is leak free. This includes all the components from the compressor to the manifold (regulators, tanks, pressure relief valve, pressure switch, etc.). By pressurizing the core system on the bench (even to a relatively low pressure of 10 or 20 psi, you can see if it is holding pressure and hunt and fix the leaks before you bury that core into the bowels of the robot. Using soapy water and a rag or brush, it is easy to wipe down each of the joints and find the leaks (leaks cause bubbles in the soapy water mixture). Once you have a leak-free core system, then you want to add each actuator to the system one at a time and test it to make sure that new system is leak free. You also want to make sure you are using a proper tube cutter to cut your tubing. Tubing that is not cut properly is generally the biggest cause of leaks.

I agree with others that you want to try to build up this new expertise on your team during the offseason. Building a pneumatic demonstrator is a great way to learn about the system. If you incorporate a bunch of linkage type mechanisms and bell crank actuated mechanisms, you can learn about how to design with pneumatics. This project can be combined with a CAD camp and a camp where you learn about the machine tools and a programming camp to have an end-to-end learning experience for the team. The older students who are helping to teach the skills they already know (CAD and machining) will then become the students as you learn about pneumatics. You can use some of the types of mechanisms out there in other teams’ CAD to understand how the mechanisms work and how to select the actuators that you are going to use to perform a given task. In the end you will have added this technology to your bag of tricks so that you are ready to use it if it makes sense to do so in an upcoming seasons

6 Likes

Pneumatics is wonderful. Failure rate is fairly low. If we can replace a motor with air we will as it’s easier for the programmer no need for encoders and gear boxes. Also it’s much faster to replace a piston than a motor. Only down side is if you’re constantly using air you have to have a sufficient size tank as we found that the compressor can overheat and drains the battery quite fast if running the whole match.

Wpilib has some great resources:

Tutorial on how to put it together

And how to program it

2 Likes

Resource management with the compressor is probably the hardest thing to do well.

In terms of battery, we always charge our pneumatics in queue with an off-board battery and tether cable, then disconnect and plug in the on-board battery. Old, non-competition grade batteries that are nonetheless still functional are good for this. The bigger issue is the compressor itself.

If you have huge air tanks, then the poor little thing has to run and run and run to fill them up every time you drop below the “fill” threshold, and compressors generally aren’t designed for that kind of duty cycle so it gets really, really hot (and really, really inefficient)…but if you don’t have enough tanks then the robot doesn’t function when you need it to. [I do wish FIRST would let teams charge pneumatics with a second off-board system, but alas, they don’t and haven’t for a long, long time, so we live with overtaxing the poor little thing.] Hitting that sweet spot of “we always have enough air but the compressor isn’t on for two minutes straight” can get a little tricky.

It’s certainly not enough of a concern to avoid pneumatics altogether. A heat sink on the compressor can help, as can fans, though not as much as using the “right” amount of storage rather than aiming for overkill…and to be honest, we’ve never done either of the former two, though I’ve seen teams that have.

Perhaps the best argument I can give for using pneumatics is that it lets you start the game with additional stored energy on the robot. You’ve already got chemical energy in the battery and elastic energy on any stretchy things, why not also have compressed air? 120 psi in, say, two one-liter tanks is over 1600 joules of additional stored energy that can be piped to mechanisms all over your bot.

Once you start using pneumatics, you’ll wonder why you didn’t start earlier.

A few years ago we build a robot without pneumatics to try to simplify the design. This was one of our worst performing robots. It had a good elevator, pretty good drive train, ok climber, but we struggled mightily with intaking game pieces consistently using only motors/springs.

From that robot on we have always budgeted for pneumatics from kickoff and have not been sad that we made that decision. There are almost always a few functions that are better suited for pneumatics. Intakes, brake deployment, shifters…

That being said as an inspector I love it when teams do not use pneumatics. It takes about half the time to do a complete inspection.

1 Like

We try to make it easier for the ri. This year we even color coded our lines for the different pressures. Our pnumatic inspection was like 2 mins cause I already knew what they were looking for/at.

Having pneumatics already on your robot can enable so many mid-season upgrades. There are so many situations where adding a cylinder and solenoid is much easier than a motor, gearbox, and often an encoder. This is especially true if your team is a fan of programming position control for motor driven mechanisms.

Clever pneumatic cylinder uses over the years:

  1. Two position shooter hoods
  2. Release of spring loaded climbers
  3. Control panel mechanism deployment
  4. Friction brakes
  5. Ratchet engagers
  6. Dozens of different 2019 hatch panel grabbing and placement mechanisms
  7. 254’s 2014 robot
1 Like

Calculating the air that you would use in a match is not that difficult. It is fairly easy to build an excel spreadsheet that determines the volume of air that is consumed for each actuation of a cylinder. You can either run the calculation at the working pressure of the robot (which is fine as long as you never drop below that regulated pressure) or you can run the calculation in terms of standard cubic feet which I recommend since then you can include the compressor performance in the spreadsheet (all the FRC compressors will list their performance in terms of SCFM [standard cubic feet per minute] at different pressures, so you can quickly calculate the time it will take to replenish the tanks). Once you determine the volume of air used per actuation, you can calculate how much the pressure in your storage tank will drop (I’ve approximated this in my spreadsheet using a derivation of the ideal gas law PV=m(Rs)T. By holding T constant at room temperature, setting V to the volume of the clippard tank, you can calculate the mass of air in the tank, m(Rs) (mass of air times the specific gas constant for air) for the air in the tank before actuation using the initial storage pressure, P, and then calculate the P after actuation by subtracting the amount of air used per actuation from the original mass and solving the same equation for P. I’ve found that this gives a pretty accurate representation of what happens in most FRC pneumatic systems. Since it is in a spreadsheet, you can easily make the number of tanks a variable that you can play with to see what happens to P in your storage tank as you use the stored air.

If you “right size” your cylinders in terms of diameter and stroke to produce the forces you need (i.e. don’t over-build), you will find that you can cut down on the air volume used quite a bit and, unless you have some really large bore cylinders or you are actuating a large number of cycles per match a single storage tank is usually sufficient.

2 Likes

A pneumatics test panel can be pretty simple. You can use it to easily test prototypes and first iterations of your mechanisms before the full pneumatic system is mounted on the bot.

Here’s our test panel.

You can see it in the background of our intake prototyping video.

Having a panel like this is invaluable for rapid prototyping, and rapid prototyping is the key to getting a highly functioning robot built fast, which in turn is the key to getting practice time with your bot, which translates to success on the field.

5 Likes

Pneumatics are easier than they have ever been to implement.

Reca.lc has calculators for air usage and compressor performance. ReCalc - A collaboration focused mechanical design calculator.

Just use loctite 545 for threaded connections and real tube cutter for push-to-connect fittings and you’ll be 90% of the way there.

4 Likes

And in “right-sizing” those cylinders, don’t forget the value of mechanical assists like constant force springs and gas shocks, as well as linkages and over-centering! Teams often use cylinders that are way larger than they need to be.

Playing around with “how do I use only a pneumatic actuator to do this task, and what is the smallest one possible I can get to work?” is a fun engineering challenge that I’m admittedly not as good at as I’d like to be.

(Admittedly, “it’s mid-season and we’re out of time so let’s just use this chonky boi” is a great bit of design versatility to have when you need it.)

2 Likes

For something like a 2-position intake (inside frame perimeter & outside perimeter intaking), pneumatics are often the best choice despite the baggage of needing the various elements on the robot. Our team this year was quite anxious about using them with our intake this year, because in 2019 when current seniors were freshmen we had lots of pneumatics and lots of problems tied to them (mainly an L3 climber powered by 4 huge cylinders).

Now that they’ve learned how to use pneumatics again this year with a more modest application, they’re comfortable & appreciate the advantages. It’s a super useful tool in the FRC toolbox. Highly recommended to learn. Agree teams should also look at incorporating constant force springs, gas shocks, & even elastic tubing.

I’ve heard that some top levels teams are happy spamming Falcons everywhere, since the power and integrated encoders allow them to replicate the performance of a pnuematics in a 2 position use case. I totally get the logic, it’s less complexity, fewer failure points. However for the average team, I don’t think this calculus adds up. Once you get over the initial hurdle pnuematics are a breeze to use and relatively easy to design for.

Weighing in with our teams approach to pneumatics. We consider if we need them for more then one reason/subsystem. If we want one thing accomplished by them then we should find another way. This is because the payoff of weight to functionality. It is good bit of weight to add the compressor, PCM, solenoid bank, and air tank that generally doesn’t out weigh just adding a motor, controller, gearbox.

That said we use them this year for our intake and our climber. In 2020 though we didn’t use pneumatics at all.