Re: Slowing down the action of a cylinder with flow control, mechanical or via softwa
 

All good points, many thanks. That's a very good point about shortening the working length of the cylinder, and that a hard stop on a cylinder is at least as hard (and damaging) as a hard stop on the frame.

I am inclined to have us look carefully at your suggestions in your earlier post about reworking the design of the arm to use springs, counterweights, and/or shock absorbers to more organically use the pneumatics for what it is intended. We will also experiment with flow control fittings (we just ordered some this afternoon) to see if that might get us most of the way there, since it's such an easy solution.

Thank you again!

Re: Slowing down the action of a cylinder with flow control, mechanical or via softwa
 

Tristan: That makes a lot of sense to me. Thanks for pointing out that distinction between flow control fittings and flow control valves. This afternoon, we ordered some flow control fittings, with 1/8" NPT ports that screw right into the exhaust ports on the Festo valves. As fittings, those would seem to be within the spirit of the rules.

Interestingly, there was actually a question on the official Q&A about the legality of using flow controls. The questioner pointed out that in the Pneumatics Manual Rev B for 2011, the sample pneumatics diagram on Page 8, and the photos on Pages 9-10, show flow control fittings screwed into ALL of the cylinder ports. These are the NPT-to-tubing fittings with the needle-type flow control screws that you referred to above. But somewhat confusingly, <R66> seems to say that only the items it specifically lists are allowed in a pneumatics circuit on the robot, and flow controls are not specifically listed in <R66>. The questioner asked for a clarification on this apparent contradiction from the GDC. Unfortunately, the GDC's response was simply that "All pneumatic valves must comply with Rule <R66>," which still (to my mind) would seem to leave the answer ambiguous.

The wording of <R66> seems to add to the confusion, since it simply provides a list of "pneumatic system items" that are "specifically permitted," and <R67> lists pneumatic items that are prohibited. But what about the pneumatic items that are in the "no-man's land"; i.e., not shown on either the <R66> list or the <R67> list? Strictly speaking, it would seem that both flow control fittings and flow control valves are in this no-man's land.

However, <R66> Paragraph E does allow for unaltered COTS "connecting fittings" with a working pressure of at least 125psi. Does a flow control fitting that also provides connection to a tube count as a connecting fitting (which is legal), or is it a pneumatics part that provides extra functionality and thus it not permitted because it is not specifically listed as permitted in <R66>?

The point is mostly moot, because the official competitions are over for the year, and as you point out, most off-season competitions don't even do inspections. So, we'll feel free to use flow control fittings and not worry about it. But I kind of wish the GDC had made an explicit ruling to allow them. Maybe this ambiguity will be cleared up in next year's competition.

Thanks again for your thoughtful (and helpful) response, and for clarifying the difference between the two types of flow controls.

Re: Slowing down the action of a cylinder with flow control, mechanical or via softwa
 

Randy:

If you take a close look at the 2011 pnuematics manual, you will see flow control devices in both the schematic and photographs. The SMC NAS2201F-N01-07S flow control device was included in the 2008 KoP, and previous years, if I recall. We have some of these in our shop if your shipment doesn't arrive in time. Give me a call if you need them.

Also, you can indeed use separate Festos to switch between a high flow and low flow path, as long as you use a separate control signal for each one. You are correct that you are not allowed to control two valves from one control signal.

I have also read CD postings of people who rotate the valves with servos to provide software control of the flow rate.

With regard to rapid on/off pulses to the Festos, 1073 did this in 2010 to provide software-control preload pressures in our kicker. I was amazed that the Festos were so responsive, but we got very repeatable results, with no observed reliability problems.

Good luck at Beantown!
-Jeff

Re: Slowing down the action of a cylinder with flow control, mechanical or via softwa
 

Hey, Jeff, great to hear from you!

That is great info, thank you. I noticed those elbow-shaped, flow-control, 1/8" NPT to 1/4" O.D. quick-release fittings in the 2011 pneumatics manual on Page 8 (diagram) and the photos (Pages 9-10), but until your message, I wasn't sure where to get those flow control fittings. I'm still a little confused why <R66> doesn't specifically list flow control fittings as an allowed pneumatics part. But the universal opinion here on this thread seems to be that flow control fittings are allowed (falling under the category of "connecting fittings" in <R66> paragraph E), whereas flow control valves (separate inline flow-control components) are not.

Anyway thanks so much for the SMC NAS2201F-N01-07S part number for those flow-control elbow fittings from the 2008 KOP and earlier. I just now ordered some, which should arrive in a few days. If we run into trouble receiving them, though, we might take you up on your offer to borrow a couple of them from your team, much appreciated!

One odd thing about using the flow control fittings on the cylinders themselves is that it seems like it will reduce the flow of air in both directions -- going into the cylinder in one direction, and exhausting from the cylinder in the other direction. We don't really want to slow down the flow of air into the cylinder, we're just trying to slow down the exhaust from the cylinder, to control the speed of the action. It seems like the ideal place to install a flow control fitting would be on the exhaust ports of the Festos, so that the flow control would only be used on the exhaust. In fact, I ordered and received some very nice $3 flow control fittings specifically designed for the exhaust ports of solenoid valves. But when I received them, I remembered that the Festo valves have these odd port threads (they are 7mm threaded ports, if I remember correctly), rather than the 1/8" NPT threading that I assumed they had, so we will need to use some kind of fitting adapters if we want to use them.

Cool idea! We could start with a high flow path to raise our arm, and as it reaches the top, we could switch to a low flow path to slow down its motion so it won't slam so hard into the top.

That's great to know. We'll hold that idea in our bag of tricks as well.

Many thanks! And if 1073 is going to the Beantown Blitz, good luck to you folks as well!

-- Randy

Re: Slowing down the action of a cylinder with flow control, mechanical or via softwa
 

Sorry with using your terminology, but we did use the fittings. Just as you say above, we placed it on the exhaust port of the valve so it only effected the flow in one direction. Good luck and I hope you figure all of this out.

Re: Slowing down the action of a cylinder with flow control, mechanical or via softwa
 

If you're using a double solenoid, then you can just turn it off (rather than reversing it) to stop the additional flow of compressed air before the cylinder reaches the end of it's travel. Probably using an adjustable reed switch feedback (fairly exact final pressure), or less reliably-time, to signal when to stop adding air. That would result in a lower final pressure in the cylinder than the 60-30psi you might be operating from, but it'd start with full pressure to make it move, then have lower than normal pressure by the time it hit the stop.


The flow control fittings typically only restrict the air flow in one direction if that's what you are worried about.
For instance, NAS2201F-N01-07S translates to: 1/4" elbow, meter-out, one-touch fitting, NPT, 1/8"port, 1/4" tubing, with seal.
The fittings are marked on the side with a large arrow (no restriction) and a small arrow (restricted flow) showing the direction.
(Added a photo showing the flow arrows)

A side note on the high flow/low flow switching idea.
Just be careful in your design to use only one solenoid value per<R74>.
Restricting the flow from the exhaust port of a single valve would affect both extension and retraction of the same cylinder if the valve is connected to both ends.

Re: Slowing down the action of a cylinder with flow control, mechanical or via softwa
 

Mark -- Ah, that's very useful info. I didn't understand what "meter out" meant, but now I get it. That's a clever fitting. I also found a description of the NAS2201F-N01-07S which says the same thing you said, that flow is only restricted from the threaded port (i.e., the cylinder) to the tube, but is free-flowing from the tube to the threaded port (cylinder). So, we can restrict the exhaust, without impeding the inflow of air when moving the cylinder in the opposite direction. Sounds perfect, and it appears the NAS2201F-N01-07S fittings I ordered today will do the trick. Many thanks!

Re: Slowing down the action of a cylinder with flow control, mechanical or via softwa
 

Regardless of its legality (IMHO I think it would be legal, and very cool) I think your ideal solution would be to use something like this electronic flow control unit. You could reduce the flow of air into the piston as it reaches the end of its travel.

Alternatively, if you relied on gravity to bring the arm down rather than power it (more efficient on air usage) you you program a duty cycle into the "lift" solenoid valve, i.e. when the arm is initially lifting the solenoid valve is open 100% of the time, but as your arm nears the top of it's travel the solenoid would proportionally reduce its duty cycle as appropriate.

Using a fixed flow control valve will penalize you in overall ground-to-lifted time (if you will) but would work just fine. My second idea doesn't cost you any weight either, which is always good :D Just throwing a few more ideas at you!

Re: Slowing down the action of a cylinder with flow control, mechanical or via softwa
 

Adding to my rules list for next year.

Re: Slowing down the action of a cylinder with flow control, mechanical or via softwa
 

Sounds like flow control fittings + some counterforce (surgical tubing or gas spring) should address the issues you are encountering.

Slightly off topic, but does anyone know if Bimba spring return pistons have a "constant force" spring or a Hooke's law linear spring on their return? If it's the latter, using a spring return piston could also address this problem (increasing resistance force as the piston nears full extension).

Re: Slowing down the action of a cylinder with flow control, mechanical or via softwa
 

I think they are just constant force springs.
If I get a chance I'll check one of ours.

Re: Slowing down the action of a cylinder with flow control, mechanical or via softwa
 

I like the idea. However, I don't know if we could talk the GDC into allowing an electronic flow control valve next year. The pneumatic rules seem to be written so that on any robot, if you bleed the stored air at the manual pressure vent plug valve, there will be no stored air left in the system. If an electronic flow control valve were permitted, then in theory, it seems like the software could completely close the electronic flow control valve and leave stored air in the system. Manually-set flow-control fittings (and regular manual inline flow-control valves) presumably wouldn't have this problem, because even though they constrain air flow, they probably wouldn't be manually set to completely block air flow, because that would make them fairly useless on the robot.

It would be really cool if the GDC relaxed the pneumatic rules a bit, though. It would be nice to use tricks so that pneumatic control could be almost as flexible as motor control.

Re: Slowing down the action of a cylinder with flow control, mechanical or via softwa
 

We used the SMC flow control fittings (both elbow and inline) on our arm and minibot deployment this year. They make all the difference when trying to smooth out the motion of an air cylinder. As was mentioned, they only work in one direction, so you can have independent settings for extension and retraction. This was especially helpful with our arm, which would slam down when we retracted it in a near vertical orientation.
Another thing that may be worth looking into is air cylinders with cushioning built in. Most air cylinders are available with internal air cushions at the end of stroke to prevent damage due to excessive impact.

Re: Slowing down the action of a cylinder with flow control, mechanical or via softwa
 

Another strictly mechanical option (though possibly not the most elegant) would be to add an accumulator tank to your exhaust and then put one of the flow control fittings on the other end of the tank.
What this would do is allow for quick initial exhaust (nearly the same speed as if you vented directly to atmosphere) as the tank is filled but as pressure builds up in the accumulator (since the outlet is restricted) the exhaust will be slowed, meaning your end of stroke should move slower, essentially the same as if you just put the flow control fitting directly on the exhaust port.

The effectiveness would vary depending on the size of your cylinder and accumulator. If your accumulator is too large, it will have almost no effect, if it's too small, you'll slow down very quickly.

Re: Slowing down the action of a cylinder with flow control, mechanical or via softwa
 

Such a device would not be legal. Not because of pneumatics rules, but because of control system rules. No electromechanical devices are allowed other than the KOP motors, servos, and permitted solenoid valves. Several teams last year inquired about using electronic regulators as a way of varying the air pressure for kicking, but they were disallowed because they involve a electromechanical actuator of some kind. Output signals may only control KOP motors, servos, and solenoid valves.