My teams robot this year uses pneumatics for the Intake and climb. We keep running into the issue of our storage pressure slowly dropping pre match as air leaks from the system. How do you seal everything to prevent all the tiny leaks from adding up and becoming an issue. To give you n idea of the leaking I’m talking about, we have 10 air tank, and storage pressure goes from 120 to 100psi in about 10 min.
By far the biggest problem I see every year is improperly cut tube. Push-in fittings hate crooked or jagged cuts and I’m talking barely crooked.
Get a tubing cutter, no scissors or wire cutters. They are cheep and will prevent many problems
Also these fitting hate it if the hose is too tight or have side force applied.
You can use your Automation Direct coupon or buy one from AndyMark
18 Likes
Minimize your pipe fittings. Use sealant on the ones you have to have. Make sure your O-ring sealed fittings are tightened with a TOOL, not your fingers.
+1 on clean, square cuts. Make sure to form your tubing back to round before stuffing it in the hole.
Listen to your bot in a VERY quiet area.
5 Likes
A little bit of soapy water can help find leaks too (bubble formation).
6 Likes
I have to ask, because no one else has yet…10 tanks? Just how much are you using this pneumatic system? Since you must have an onboard compressor in any case, it seems like serious overkill to use that many tanks and, in doing so, make a lot more in the way of connections than is probably needed. We make serious use of our system (both for intake and for moving our climb arms) and we only have four tanks. I’d check the between-tank connections, especially if you have the tanks ranked together in a bank as many do. Those tend to be tight turns between one tank end and the other and those kind of connections are always more prone to leaks because of the sideways stress on the air tubes at the tank fittings. We specifically plumbed our tanks so that this was kept to a minimum. Simply reducing the number of tanks might help.
2 Likes
We use a series of fitting from SMC that specifically tolerates angled cuts. This really helped out because they don’t require students to be very accurate. They still suffer from other issues that plague fittings such as the tube not being fully inserted, but they’re totally worth it IMHO.
1 Like
Can you provide part numbers for those/links?
I’ll have to call my SMC rep, he provides them to us so don’t have it off hand.
2 Likes
2nd on the square tubing cuts. Also make sure you have smooth tubing being inserted into push-in fittings. If the tubing is ridged or gouged from previous insertion/removal cycles, cut off the used part and insert a new, smooth section.
It also seems much more foolproof to seal metal-to-metal joints with pneumatic sealant like Loctite 545 (purple) rather than using teflon tape. My team suffered with hard-to-find slow leaks for years until we switched from tape to liquid sealant.
Our team has struggled with leaky pneumatics almost every year I’ve been on the team. We finally made some progress this season with our robot holding air for hours by the end at our second competition. My advice is two-fold:
- Reducing number of connections, cutting tubing well like other posters mentioned, and in general being careful and precise and giving due attention to pneumatics.
- But honestly more realistically for a team finishing a robot just before going to competition is to trouble shoot and check for leaks after every match at your first competition to optimize it and work out the kinks (pun very much intended).
- We use a stethoscope with the listening end cut off and listen over major connections (good for catching small leaks)
1 Like
A big thing as mentioned is the joints. Air doesn’t leak from the center of a hose (usually), so making clean cuts (square, and using a proper tool) as well as minimizing joints is critical. You said you are using 10 tanks? Have you considered switching to one large tank instead? In 2019, the team I was involved with used an aluminum, 1 gallon ARB tank which I think is equivalent to about 7 of the normal clippard tanks. Using one tank eliminates a ton of joints. Just note, if you switch to a single aluminum tank, you’ll hear a lot of comments of people claiming your tank is illegal, prompting inspectors to stop by to confirm that the tank is indeed legal.
2 Likes
We dropped our leaks by using a solenoid manifold. This got rid of a bunch of tee joints in the air main.
10 tanks is probably your issue though. It is most likely leaking where the fitting is screwed into the tanks. Some tank thread can get damaged fairly easy. We had to replace a bad tank after initial robot construction this year.
We only use 2 tanks and have enough pressure to start a match several hours after charging.
David
2 Likes
I spoke with my rep today and he said all SMC push-in fittings tolerate poor cuts. Our teams experience supports this, you should see some of the cuts on our bot.
1 Like
been using SMC products for a very long time. He is not correct.
Wait, a salesman isn’t accurately describing their products? No way!
On a serious note, I would be curious to compare which fittings you use to ours and see if there are any differences. We have had excellent luck so far and our cuts are typically done with my multi-tool and are jagged.
3 Likes
We don’t have an issue running 4 or 5 tanks. Here are some tips.
If you use Teflon tape make sure you wrap it counter clockwise to the thread.
Don’t put too much / too little Teflon tape on the fitting. We normally do 2 or 3 wraps.
Make sure you have sufficient thread in your solenoids, we bought some airtac solenoids and it wasn’t threaded far enough.
Don’t reuse tubing ends, always cut new ends. ok to reuse for testing.
Allow sufficient length of tubing so it doesn’t pull on the fitting
Push pull test on all fittings
Use flexible tubing, I recommend the tubing from automation direct.
To Clarify … EVERYBODY gets a FREE tubing cutter when they place an order using their AutomationDirect voucher (to get more free stuff).
We should probably advertise that better.
Use your 2022 Voucher - get a FREE tubing cutter!
Go here: https://library.automationdirect.com/free-first-voucher/ for details!
11 Likes
It sounds like you have a relatively slow leak that you might not be able to hear (10 minutes is a bit of time). hitting all of your connection points with some Windex or a spray bottle with soapy water is a good way to locate where your leaks are due to creating large bubbles at the leaking point.
Most of our issues with pneumatics come down to bad cuts on the tubing or bad connections on the tubing due to not re-cutting the tube when reseating a connection (or somebody just didn’t push the tube in far enough). In our experience once we have gotten everything sealed up we have little to no issues, and any problems can be solved with a bit of Windex and re-making a connection.
This year our team had a fully pneumatic climb that needed 23 air tanks, and we were able to hold 120 psi for over 24 hours. So you will probably be able to get your 10 tanks to work if you can find the leaks and address them with some of the pointers Running_nightmares listed.
Another thing to note is that since it probably takes a significant amount of time for your tanks to fill up, your compressor may be running long enough to heat up the tube at the connection point causing it to expand. you will want to watch that since eventually it may bloat to the point that you have a blow out.
A certain level of quality of the construction is needed for it not to leak. It isn’t difficult to achieve. Lots of good suggestions have already been made. Perhaps, it would be good to slow down and re-run all your tubing, making fresh cuts.
It is also possible there may be a defective component that is leaking. The soapy water would help find it.
With the push on fittings, if the tube has been pushed in a few times, it might be good to cut around 1/2" off the tube so a “fresh” part of it is now being pushed in. The tubing may be getting scratched up when it is pulled out of the fittings.
Lots of great suggestions here guys, thanks.
After reading through this I’d guess our problems stem from having many many push to connect connectors in our loop, and most probably have imprecisely cut and in a few places strained hoses. I will address those issues and hopefully that will solve it.