Synchronizing Pnematic Cylinders?

Is it possible to synchronize pneumatic cylinders without the use of electronics?

Yes. Pneumatic relay logic is applied. Look up supplier names like Parker-Hannifin or Norgreen for further reading.

Describe your problem in more detail, otherwise you can't get answers much better than "yes," or "no."

I'm trying to Raise a canopy that works on a parallelogram by pulling the two front legs of the canopy with pneumatic cylinders. The canopy is on mobile equipment and used in a dirty environment. The cylinders need to be in phase for both raising and lowering the canopy to keep from binding the joints of the canopy.

Do you mean like making two cylinders extend and retract at the same rate. If so, you can use needle valves to synchronize extend/retract speeds.

Mike

Sorry - forgot the "?".

Corrected:

Do you mean like making two cylinders extend and retract at the same rate? If so, you can use needle valves to synchronize extend/retract speeds.

Mike

They do need to be at the same rate, but more importantly in the same position at any given time.

Depends on how synchronized you want to be. It will be difficult to tune without electronics.

There are several factors to consider, one of which is the length of the tubing supplying air to your cylinders. The cylinder farthest from the air supply will be delayed a bit. Thus, your speed control device will be set differently from one cylinder to the other.

If I may suggest, why not just use a single, large cylinder and use wire rope to raise the four corners using a system of pulleys? That way, you eliminate three cylinders (less to maintain) and eliminate the problem of synchronization.

The cylinders need to be fairly close. Because the cylinders are attached on the lower part of the legs, there is much travel in the top of the leg compared to the travel of the cylinders.

The wire & pulley is a good idea but, I am trying to avoid this to minimize finger traps. I appreciate the suggestion.

Check out a hi lo camper trailer. They raise the roof totally by cables. To get hydraulic or pnumatic cylinders to operate together you will have to have some kind of framework to keep them together. Even with needle valves and other regulating devices,if one cylinder has a draggy spot anywhere along it's travel distance or more drag in the framework on one side or the other then one cyl will get ahead of the other and bend your roof frame.

Electronics helps, but it is not a practical idea. The very least amount of load difference will change speed. It is better to let air be the dumb power, and synchronize mechanically. I don't know if this mechanical device is practical in your application, but it solves the problem if you have the space. Here's a link: http://www.powercomp.com/

This is a fairly common problem. You can attempt to correct the flow rates into each cylinder with valves, equal hose lengths etc but it never works very well, eventually one cylinder moves faster, so you readjust the flow, then a cylinder gets a bit slow in one part of its range, etc, etc, etc.

You can control the speed more accurately if you use an air over oil system. The pneumatic system pressurises a low pressure oil system, since oil isn't compressible it is easier to control the flow and it all behaves more smoothly.

Better yet, use electric linear actuators. Jeff

Contact a local pneumatics distributor. They will be glad to assist you with diagrams, BOM's, costing and delivery. Call three of them to get a base price and maybe three ideas.One of the thing you have to keep in mind is air leaks over time. If you want this unit to stay were it is you must also include check valves.

Run them in series. That way they will always operate in synchrony. You will need a T in valve in the middle to ajust the trapped fluid in the isolated leg for leakage etc

By running them in series you mean running a hose from the capped end of one to the rod end of the other? Don't you need two different size cylinders to do this?

Here is a graphical representation of how I have balanced multiple cylinders here in our facility. Whether it is the "best" method or not, I do not know but what I do know is that it works well and we've never had any problems. Also, it is very similar to various suggestions noted above. For your stated application it should work fine for you.

Thanks,

Sam

you need a means to make sure the same volume goes into each cylinder. In the parallel setup the cylinder with the lightest load will hog the motion while the other stays put, unless there is a parallelogram lift frame.

Series is best as you are forced to have equal motion in each cylinder. Yue need to adjust for any leakage and the initial setup, as leakage can cause one to lag

Due to the compressability of air it is unlikey that you will accomplish a reliable solution using two pneumatic cylinders. You may consider reinforcing the canopy to make it rigid, then it would insure that the cylinders operate in unison. If possible it may be preferable to use a single cylinder.

I was speking only of using an incompressible fluid, oil, water etc. Air is also suitable, but the compressibility means you need some sort of balancing. The series method workd fine with oil

Use the series method Aurizon described, but put a light oil in the middle between the cylinders. Air is too compressive, and this makes one half follow the other.

I've taken a look some different mobile equipment, and not knowing exactly what type you are talking about, I've noticed that most of the ones I've looked at have hydraulic systems rather than pneumatic systems already built into the equipment. Switching to hydraulics over pneumatics solves many problems... smaller cylinders, fluid compression, etc. But tapping into an existing system adds a few more... fluid consumption, existing tank size, etc.

Reading into your statement a little deeper and going out onto our shop floor and running a couple of experiments on existing set-ups similar to the schematic I provided above... I've got some additional advice.

"pulling the two front legs of the canopy with pneumatic cylinders" This infers that your retracting the cylinders while lifting the canopy, using the rod-end of the cylinder to do most of the work. First try to revise your layout to push, using the piston-end, rather than pull. In this fashion you have more surface area inside the cylinder working for you.

Assuming you're talking about lifting a metal framed, canvas/fabric covered, canopy over an operator driving some type of earth-moving or tow-motor equipment... I don't see precise alignment between the two cylinders being of prime importance. Also, I don't see the load being such that compression will be a huge factor. By design, making the front frame of the canopy (the part you're lifting) ridged, and it's pivot point relatively sloppy, the schematic above will be more than adequate. Size each cylinder so that, by themselves. they can lift twice the entire load.

I have a few parts laying around in our shop and I threw together a quick experiment (it's a slow day)... a quick summary of the results is... if your air supply flow is greater in CFM then your two (one on each cylinder) ballcheck needle valves will allow through and your main supply pressure is greater than what the back pressure would be of 100% of the load was on only one cylinder, the system balances itself.

The only flaw is without some type of electronic relation between the two you have no way of sensing if one of the cylinders jams... (i.e. sees more resistance then it is designed to overcome.) Sizing each cylinder so that it can do the work by itself eliminates this problem almost entirely, discount erroneous damage to the system.

Anyway, the set-up I made in our shop had two cylinders capable of each lifting about 500 lbs. Set up per the schematic sent previously and ensuring the needle valves on each allowed less than 50% of the total incoming flow I placed 400 lbs. on one cylinder and nothing on the other... each raised at the same rate. Adding 600 lbs more to the 400 overloaded one cylinder by 2x. Running it again, the cylinder with the 1,000 lbs did not move and the empty one raised at the same rate as it did previously.

In summary, this quick little experiment, combined with the over 7 different areas in our plant where I have similar designs in operation shows that it's an inexpensive and plausible solution. Are they precise... no. Does one cylinder start slightly before the other... sure. But, for the applications where they are used, they work perfectly.

In closing, It's really easy to over engineer stuff... bottom line, make the operator get off his butt and lift the canopy himself.

Seriously, I do have to say that I don't think that pneumatic cylinders is the best design solution for your application anyway, I was just letting you know how I have synchronized them together for my applications that don't require exact precision.

Well, gotta run... our next shop floor shift is in and is wondering what the hell I've been up to time to go put my toys away, take care.

I like the idea of making the operator lift the canopy the best.

I agree that if the two legs are tied together to make one rigid front leg then keeping the cylinders in phase is not nearly as big of a deal.

You mention to size the cylinders so that each can lift 2x the load.

Assuming the to front legs are tie together, what if each cylinder is sized so that it can lift half of the load (or slightly more) then, one will not move without the help of the other and they will have to be in synch? Or, will they just fight each other?

The reason for 2x the load, or even having each cylinder sized to handle 1x the entire load would be fine too, is related to the other "guest's" response in post number 21 below. No matter how well you design the canopy, the loading will not be symmetrical as you lift it. If you size each cylinder to lift exactly half the load and lets say you see a 60/40 load distribution, you've (by designed) just stopped one of your cylinders from moving. The cylinder that does move will be the one with the 40% load and as it lifts the canopy it will tilt and now you will introducing lateral vectors of force into your equation. Now, you have to design your frame to withstand twisting and buckling as well. Certainly, once the one cylinder starts to move, it "should" start to pull the other up but your looking at a lot more dynamic variables in this situation that you can design out by simply over-sizing your cylinders from the start.

Getting back to how this relates to the other Guest's response in post 21 is separating the problem into its two components... pressure and flow. Synchronizing your cylinders is purely related to flow, so long as the pressure portion of the problem is properly addressed. Sizing your cylinders and your incoming PSI such that one cylinder by itself can lift the load, your system will be less affected by load distribution and variance as the canopy is lifted and accumulated water runs off the canvas, or the wind catches it, etc. So as the load distribution variances move back and forth from one side of the canopy to the other, your individual cylinders are still within their capability and are not back-transferring forces into the piping system of sufficient magnitude to change the balance of working pressures between the cylinders.

With the cylinders and your incoming PSI properly sized, it's strictly a matter of balancing the flow. You want your main air supply to dead-head against your flow restrictor (in this case a needle valve) that's mounted on your cylinder. Let's say each cylinder needs 50 CFM (100 CFM total) then you should supply 150 to 200 CFM on your main line.

In summary, let's say your total calculated load (including canopy weight, cantilever lifting, wind, wet material, water accumulation within the folds of the retracted canopy, and your safety factor) comes to 1,000 lbs. Size your system PSI and cylinders such that each cylinder can lift 1,250 lbs. Since a 50/50 load distribution is optimal, worse case would be a 100/0 distribution. With one cylinder seeing 100% of the 1,000 lbs and the other seeing 0%, each are still within their designed capability and will move as fast as you let them.

Regarding balancing, remember that the extend and retract functions are separate. To avoid fighting against each other, set your lift canopy extent speed slower then your retract speed... Set your output (exhaust) needle valves located on the rod-end of your cylinder slightly more open then your incoming valves that are located on your piston-end so you don't fight against them.

In closing, many other repliers have offered suggestions for better solutions... air over oil, piped in series with an oil buffer, hydraulic, contacting local pneumatic distributors for design ideas, etc... all of which would provide you with a more precise synchronization. My final suggestion is that you consider all of them and list your own pros an cons related to cost, complexity, effectiveness, durability, the quality/feel of the equipment you are building (read as the difference between a Ford Focus, F-150, Crown Vic, Mustang, and GT) all are vehicles and will get you from A to B but some "feel" better then others and some will last longer than others and some cost WAY more than others. All these things will guide you to "your" best solution.

Sam

You are right but it is possible to determine with a quite simple mechanical device the difference in strokes and use it for the necessary corrections of the 2 flows.

i too have agonised over this conundrum.

I finally realised the problem was the contols work on pressure and pressure varies depending on the loads. you need a volume control.

My solution was to make a device similar to a gear pump. you have a single 'pump' gear tor meter each line and join them externally with gears or common shaft so they must rotate together, exactly metering the fluid in equal proportions.

such "flow dividers" are used since years in hydraulics for a synchronized movement. They can be supplied with up to 6 or even 8 flows!

These flow dividers are usually gear type escapement types on a common shaft so every turnallows thge same flow to each line. If any line jams, the back pressure stops them all.

In fact not the same flow since the back leaks are function of the pressure differences across the gears. This is the reason they are used only where a coarse synchronization is required and of course as you very correctly comment no jam could occur. But if a jam occurs it is normal that synchronized rams stop all of them at same time if not there is no synchronized movement anymore.

How do you think a Steam locomotive operates? So it is steam and not air,but the cylinders are synchronized with no electronics.

http://www.trainweb.org/tusp/ace_det.html

The different suggested solutions are not working properly since they are not based on the error measurement. In fact if you want to synchronize 2 cylinders you MUST use a system which determines continously the stroke difference and corrects the admission or the outlet from cylinders accordingly. It can be done and it is not too complex. The use of needle valves requires a load compensation you cannot make working in a correct way and every time friction or loads are not exactly the same on both rams the synchro mouvement ist not any more synchron which could lead to a frame deforamtion or a stop.

Which stroke ? Which diameters ? which speeds? Which loads?

I may send you a schematic if you give you mail.

Hydraulic cylinders in series would seem to be correct, as the stroke is automatically the same for same size cylinders and the force automatically equalizes and you get a uniform rise.

You do need a way to synchronize them at the start by fluid +/-. Any leaks of course would lead to an imbalance, and would have to be dealt with. End screw adjustors would be useful as well.

I just realised by Series you mean Cylinder 1 input connects to source, its output side connects to input of cylinder 2 input whose output connects to Cylinder 3 input etc. This works well with hydraulics. Pneumatics however are different as the pressure will vary between cylnders depending on the shaft resistance compared to the other shafts.

Yes, it needs a non compressible fluid, air/gas will not work

It is not true the fluid must not be incompressible in order to synchronize the cylinders. The difference is that you consider only PASSIVE solutions and the best solution is an ACTIVE one.

In my application it works very well and reliably. In my opinion active solutions in this situation soon fail as they are running in a dirty environment. The hydraulic series method is fool proof, unless you get a leak/break and in that case the other methods will also fail.

That said you can make an active solution work as well, but at what costs? The series hydraulic has the most basic of costs.

You are too pessimistic. In general the environment for pneumatics is not as dirty as for hydraulics. From an other point of view even if dirt is present it depends where you place the "sensitive" sections. I have the feeling that you had a problem of filtration. I worked for many years in servo hydraulics and servo pneumatics and I had no troubles. There are 2 basic conditions : design the system right from the dynamic point of view and provide the correct fluid filtration or the right quantity of lubricant depending if you are in hydraulics or pneumatics. Unfortunately too many suppliers of servo fluid systems are not qualified enough and sell incorrect solutions. This is one of the reasons many fluid systems were replaced with electric solutions. Electrical engineers have more knowledge than mechanical engineers in active servodrives.

I am basically a mechanical engineer so that I know about what I speak.

Planning on using 2.5" bore x 15" stroke cylinders at about 60 psi.

"In fact if you want to synchronize 2 cylinders you MUST use a system which determines continously the stroke difference and corrects the admission or the outlet from cylinders accordingly. It can be done and it is not too complex."

A schematic on how you accomplish this would be very helpful, my email is capt@ou.edu