Pressure in pneumatics system quickly diminishing

My team this year is using two cylinders to actuate our climber. We can only actuate and de-actuate the cylinders once. There isn’t enough air in the system to go a second time,

So, in the scenario that we miss the rungs once, we can’t attempt to climb a second time. I’m not sure why there isn’t enough air in the system to allow us to do so- we have four air tanks (@ 120 psi / 60 psi). I also tested with 5 air tanks and there was no difference.

Does anyone have advice on how to get our climber to actuate and de-actuate more than once/ maximize the pressure in the system?

Thank you!!

We are using two of these cylinders: Stroke length of 22", 1.5" bore size, etc.

With two of these double acting solenoids: Air solenoid valve VUVG-LK10-B52-T-M5-1R8L-S | Festo USA

Would using single acting solenoids help? I am also looking into purchasing faster acting solenoids as the current ones have 0.31 Cv. I’m not sure if this would significantly help, though?

We were helping a team with climbing with pneumatics and they happened to have cylinders they wanted to use. They had 2 cylinders that were 3 inch diameter. I think the calculation would let that lift 800 lbs. So the first thing to consider is how big and how much lift your cylinders can manage. With 2 cylinders I wouldn’t consider anything larger than 1 1/2". I think you need at least 1 1/4" typically. If you have larger cylinders you might want to go down to using just one. We lifted a robot with a 2" which used a lot of air. And then a ~1 3/4 cylinder which used less. You also have to consider the stroke length, with the bigger cylinder using more air. Above we had a 24" stroke, which was a lot of air.

I think you should reasonable be able to lift the robot with 6 or fewer air tanks. That is if you don’t have a lot of loss through usage or leaks. If you use a lot of air, you can end up needing to refill the tanks. You want them to be above 60 psig to have maximum lifting power. When they are lower the air tank reservoir volume makes it tougher to get to higher pressure since you have to first fill the tanks. In some cases being able to skip tanks and directly pressurize the cylinder may be better, but we always increased tank volume. In our case we could get all the air we needed with an air lift aluminum tank. But I don’t think that is necessary in this years game, we were dealing with the lifting mechanism also being an elevator in the 2018 game, so we had to use the cylinder up to 15 cycles during a match.

There are potential mechanical work arounds, but I would look 1) at cylinder size and 2) to verify you aren’t losing pressure and seek out any leaks.

edit: I forgot to say that our 1 3/4" cylinder lift was not lifting the entire robot.

1 Like

If you are lifting the entire robot you should have ALL your air tanks at 120 psi and accept the speed hit that running through the regulator causes! You may also want to have separate valves for each lifting cylinder to double your available flow rate. Plus, that will help keep them sorta synchronized…
This might be stupid, but you should verify that your compressor starts as soon as you hit the climb cylinders…
Another thought: If you have to extend and then retract under load you might want to regulate the “extend” function pressure MUCH much lower. Maybe 5-10 psi? This will significantly reduce the volume you spend on that low-force motion.

On to the exotic stuff that may or may not be allowed under the rules:

I don’t know if its possible or allowable to do a regenerative cylinder connection, but that’s another thing that could help. You would extend at low pressure and then flip a valve to connect BOTH sides of the cylinder together under the low pressure. This would take up the slack in your system while spending relatively little air. Then, you apply full pressure to retract. There are three position valves that can do this sort of trick in the middle position.
A quick exhaust valve on the extend side might speed up your cycle time a little bit, as the venting process could happen faster than the valve would normally do it. Again, I -think- they are OK to use…
One final trick would be to have a way to isolate one or more air tanks after they are used up, so you don’t end up re-filling them while you are desperately trying to get a lift. Again, I’m not sure that you are -allowed- to do this trick.

1 Like

It would help to know what the piston dimensions are. Also if all the piston force is applied in the lift direction. Also check if your high pressure side is filling up to 120 psi, the shut off switches are not always set to the max. Also are you sure there are no leaks, i.e. once full at 120 it stays that way for the whole match until you need the pistons. Also check for no leaks in the extended piston position that aren’t charged until activated.

I highly recomend using the ReCalc simulator if you are new to pneumatics:

Make sure to get your tank volume and pneumatic cylinder sizes correct (should be provided by the MFGR or reseller).

BTW, the accumulators (tanks) that are commonly used are 574mL (1 mL = 1 cm^3)

a trick you can use is to deploy the pneumatic cylinder with less pressure by using a regulator and pressure gauge down-stream of the valve. That way the pressure that goes into the cylinder is much lower and uses less air. If you are not using the cylinder to do work in both directions this can be an easy adjustment

Thanks for your great tips. I along with my team are new to pneumatics so this is super helpful.

We need the stored pressure to be at 120psi, but the working pressure at 60psi as the FRC manual states. We do have two solenoid valves, one for each cylinder. The compressor does immediately start, but that it just isn’t enough to make up for what’s lost.

Follow up questions: How exactly does one specifically regulate the “extend” function pressure, but not the “retract” function pressure? Do you know how to isolate an air tank and then use this air later on (if allowed)?

Use a secondary pressure regulator. You can place a regulator between the solenoid and one of the ports on your piston to reduce the working pressure that goes into this port. Add on a gauge if you feel the need, but it’s not strictly necessary, as this secondary regulator is downstream of the primary regulator. I recommend using the small Automation Direct regulators for this!


The piston has a 1.5" diameter and 22" stroke. More details here:

The high pressure side is filling to about 110 psi, and I’m not sure why it’s stopping there… I’ve tried adjusting the regulator and resetting the pressure switch but it still won’t go above 110 psi. Once it reaches 110 psi, it doesn’t drop so there should be no leaks.

Assuming you don’t have other pneumatic functions that you use during the game (like an intake), you can plumb all of your air tanks to hold 120 psi. You regulator steps that down to 60 psi before the solenoid valves. This will maximize the energy storage.
There’s three ways to get a lower extend pressure than retract.
1 Put a regulator between the solenoid valve and the extend port on the cylinder with a check valve to allow the reverse flow to get back to the solenoid valve. You might want to hook the extend side of the two cylinders together to avoid having to buy and set two regulators… Its possible you could find a pressure regulator with an integral reverse flow check valve. I didn’t find one in the first couple of searches. YMMV.
2 Use separate non-manifold solenoid valves for the extend and retract ports. This will allow you to put a regulator on the inlet of the extend solenoid valve. This requires more complicated controls and more valves but gives you direct control of the pressure you want! You might get away with using one valve for both extend ports and living with one cylinder going up first.
3 Put a flow control valve with reverse flow check on the extend side and throttle it back a LOT. Play with the timing and minimize the time that you are in extend mode. Fiddly adjustment… But its simple. The idea is to prevent the cylinder from ever reaching full pressure.

I would go with option 2 if I wanted to make my climb happen as quickly as possible.

I don’t think isolating tanks is going to come out FRC legal, so I’d take that off your list :wink:

Keep in mind that most pilot operated valves may not work or become unreliable at under 20psi.

1 Like

The other option is to use a spring or surgical tubing to retract the cylinder instead of air.

Others have already mentioned ReCalc, but I haven’t seen you mention using it. Here’s a link matching your setup (as I understand it at least).

As you can see, it recommends 9 tanks, which would explain why you didn’t see any benefit increasing from 4 to 5 tanks. If you instead reduce the extension pressure to 5psi as suggested above, it changes the recommended tank count to 4. Of course, all this assumes you have no other pneumatics on the bot. Adjust as necessary if you do.

Is air escaping from your pressure relief valve at 110 psi? It’s often not obvious with all the racket of the compressor. If you have an adjustable relief valve, it may need to be adjusted to relieve at a higher pressure. If you have a non-adjustable type, you might need a different valve.

If the filling is stopping at 110psi, you may have a faulty pressure switch.

Check valves are not legal as it prevents the system from emptying when the dump valve is opened.

Springs or surgical tubing are a great way to quickly extend without using air. Many teams use a latch to hold against the spring force, then pull the latch to begin the climb.

This weekend I saw a robot with 16 air tanks. They designed the bot to hold 12, then added 4 more. I am told that is not the record number of tanks.

Just want to gently push back on this point: check valves are explicitly permitted by R804F, and can definitely be employed in ways that dump pressure when the main shutoff is vented.

OP, if you’re looking for another way to optimize air usage then I would recommend putting those valves you have as close to your actuator as possible. This minimizes the volume of air in the working lines between the valve and actuator, which in turn reduces the amount of air you’re venting on every actuation. It can also speed up actuation as you now have to let a smaller volume of air through the valve before the actuator and working lines are fully pressurized. I will say that more intentional pressure regulation and actuator size choice will likely have a larger impact overall than a few feet of tubing, but it all adds up.

Festo also offers a valve similar to what you currently have with M7 ports instead of M5, which has a nominal flow rate of 340 l/min as opposed to the 195 l/min: Air solenoid valve VUVG-LK10-B52-T-M7-1R8L-S | Festo USA

Thanks @Josh_Fox , I meant “IF” not “AS”

Keeping in mind that R804(F) is pursuant to R813(A) “…to relieve all stored pressure…”

Perplexed at everyone telling him how to make the pneumatics faster…his climb even with a tiny valve actuating a 2- inch dia cylinder will still be much faster than motorized winches.

The math you need can be in those programs, but you can also get Pascal’s law and Boyle’s law out of any science book.

A long time ago I made an easy spread sheet that shows what standard sized (Bimba) cylinders can push and pull at 60 psi. It’s an easy reference…I call the “Air Force chart”
air force cylinder chart - Google Sheets.

In 2018, as @ngreen already mentioned, we built a robot that used a 1.75 dia air cylinder both for it’s elevator and to climb. originally we designed a complicated circuit that kept three tanks in ‘reserve’ at high pressure for the climb, and we were able to get that circuit okayed by LRI. The thought was the elevator could operate on lower pressure through the match and only needed 60 psi for the climb. Still was iffy.

Turned out to be much easier to buy a 2-gallon tank designed for car suspension–equal to almost 16 clippard tanks…and air was limitless. important to not ever let it go to zero because more than ten minutes to fill from zero between matches. Big tank weighed same as four clippards.

Approved in Wisconsin by Big Al himself. He insisted on bolts holding it in, not zip ties.

I don’t think that’s likely - FRC robots just don’t have the flow rate to fill a large cylinder that quickly. For example, here’s a video of an endlessly repeatable climb my team had last week (red alliance) - 3 seconds from entering the hub to being a foot off the ground on the mid bar. I’d love to see a video of that sort of performance from a pneumatic mid-bar climber!