Two Air Cylinders in Series? Parallel?

You nailed it SolarEagle. That's the idea except can't we afford a custom machine like that. We have a lathe. I saw that video too. That's where I got the double roller idea. The difference is that the tube we'll be spinning is much longer and bigger in diameter and will be spun over a mandrel. How were you able to post my photo to page when I couldn't? Thanks!

I used to have a buddy that formed brass and aluminum by hand, just had a big spinning mandrel and leaned on it with a forming tool....lost art now I think...anyway, what are your concerns?

We used to do that in industrial art class with aluminum..., high school industrial arts class of this type when the way of the dodo bird.

Yeah I just copy and pasted it...One thing I think is that you have to work the metal, it's formed a little at a time and repeated as you see the guy doing in the last video...and I can tell you it's not as easy as he makes it look...so precise control of the force applied is necessary, as well as a sense for how much of a step you take with each pass, it's an art form really....to maintain uniform thickness you have to migrate the metal, how do you do that, practice practice practice....I would think you need mechanical control of the force rather than pneumatic, just think it would be too hard to control with the variability and dexterity required....You definitely will need to build a jig...I would incorporate a force measurement capability for repeatability....

I've tried manual/mechanical and had success in the past with thinner wall and smaller diameter employing a rotary cam plate set-up. It worked well. It just didn't work in this case. The operator ended up hanging his full weight from the lever to get the material to lay down on the mandrel. We need more power and give the guy a break. Here's what I did in the past that won't work here. Again, I can only post a link. The point is I need more force. I want to use air cylinders. Can I do that per the first photo? I need two working together. https://photos.app.goo.gl/4kNyg9RbXYiQufHr8 Here: https://photos.app.goo.gl/H9ZrGxwkozWwQ39Q6

Yes. The mandrel never changes it's location but the tube is fed out on to it incrementally so only a short lengths are being worked at first and then longer lengths as the process progresses. Eventually the rollers will make a full sweep at the end... but we digress. Can I employ two pneumatic cylinders to do this with some control? How? What do I need beyond I've already stated? I'm completely new to pneumatics. https://photos.app.goo.gl/H9ZrGxwkozWwQ39Q6

How is that application different? At 120 psi from a continuous available compressed air source into 4" dia air cylinders x two gives me well over 1500 lbs force per cylinder. Based on my manual experience, that should be more than plenty. The problem I keep getting hit with is timing. Will they arrive at the target at the same time with the necessary but controlled force? How can I make that happen? It's a short stroke from such comparatively big cylinders. 1-1/2" travel max. Will it work? What do I need? Please help me.

Maybe you could use opposing screw threads like in a self-centering vise to position the tool against the pipe, then have a pneumatic plunger that goes through the positioning mechanism that holds the tool, apply the pressure, so you have the tool clamped on the surface of the pipe before any pressure is applied...sort of a 2 stage positioning and clamping system...

http://www.nookindustries.com/Product/ProductLine/Acme-and-Lead-Screw-Products

This is pretty cool, bit pricey though....

https://www.mscdirect.com/product/details/97440168?fromRR=Y

I think the key is side-to-side freedom of movement of the entire assembly...so the clamping assembly moves side-to-side, self-centering itself when clamping...this assures equal pressure from either side...then when centered, the assembly is locked in place and only the clamping surfaces can move....at least this seems reasonable to me....

That's exactly what I thought when I saw the model. GA.

Providing you are confident in the cylinders being able to generate the required force. filling the front end of the cylinders with oil, from a container connected to the front of each cylinder (it might even work with one container connected to the front of both cylinders) with give very fine control of the cylinders as they advance.

I'm sure there is an illustration of a typical hydro/pneumatic circuit on the net but if not I would be more than happy to give a more detailed explanation of the circuit although it might get a bit long winded.

John

Well you need to gear down the clamp to increase the torque....something maybe like a two sided vise that would tighten from both sides with a fine threaded worm gear assembly...

It’s a craft,... for us, we made small bowls... and being kids,... it wasn’t easy.

Maybe you could do it electrically. Connect each tool to a screw thread driven by a geared stepper motor. Drive both stepper motors synchronously so that both tools are driven into the workpiece the same distance at the same time.

Seems like it would be a lot more complicated and expensive but certainly could be done...

https://www.jvl.dk/984/stepper-motor-high-torgue

Probably would need some strain gauges, the motors would still need gearing to increase torque, you would still need a centering mechanism I think, and a quick release mechanism...

Found this.

http://s3.cnccookbook.com/MTLatheSteppers.htm

I have an old Unimat that I used to fool around with. Maybe I'll buy a couple of stepper motors and replace the handwheels on the cross slide and lead screw.

The pressure requirements will change as the sidewalls get thinner,so less pressure will be required as the forming continues.

This will require feedback,either manually or automatically.

In the days of pneumatics they used a valve positioner for precise valve position control.

If you are determined to use pneumatic control,here is an instructive link on pneumatic controls.

https://www.instrumentationtoolbox.com/2016/10/how-pneumatic-valve-positioner-works.html

I could not edit my post #28, even though I had plenty of time left,so here is a corrected version of my post above.

----------------------------------------------------------------------------------------------------

The pressure requirements will change as the sidewalls change thickness so more

pressure will be required as the forming continues.

This will require feedback,either manually or automatically.

In the days of pneumatics they used a valve positioner for precise valve position control.

If you are determined to use pneumatic control,here is an instructive link on pneumatic controls.

https://www.instrumentationtoolbox.com/2016/10/how-pneumatic-valve-positioner-works.html

The tube will be spun down on a solid mandrel that dictates the shape. The pressure applied to the spin rollers does that work. The force can surely be adjusted as necessary during the final passes along the length in the final working of it. The tube is feed out incrementally on to the mandrel and worked in progressive lengths to smaller tapered diameters little by little to longer and longer. I have done this manually/mechanically before. This is not a new process for me. I just need more force now. A constant and reliable force. It's a bigger and longer and thicker walled tube this time. I know the process.

I would suggest using a single cylinder and mechanically linking the two sides together to move in unison,like a the jaws on a bolt cutter.

Here is a link illustrating various linkages.

Keep it simple whenever possible.

Link:https://web.mit.edu/2.75/fundamentals/FUNdaMENTALs%20Book%20pdf/FUNdaMENTALs%20Topic%204.PDF

I have explored that idea but found, as much as I could find, to be inefficient. It's basically applying the working force 90 deg from the applied force by mechanical means. Friction and other laws of leverage and physics get in the way if it. I appreciate your input but I'd still like to have my initial question addressed. Although it's been interesting it looks as though I'm not going to get my question answered here but thanks just the same everybody. I've appreciated and enjoyed your ideas. I was inquiring about applying two pneumatic cylinders to provide the necessary direct force and it's feasibility. Thank you all.

There is no one right way to do this, each person here will have his or her own methods...further we cannot say whether you can do something or not, we don't know your capabilities...If you're asking will an idea work, well the best we can do is a maybe...it all depends on how you go about it.... If you're asking us how it should be done, well everybody will have a different answer, and everybody will be right....because the possibilities are endless....

If you parallel the two cylinders then when the first piston makes contact it will stop until the second one makes contact and then both will apply the same force. Control can be easily achieved through a valve block fed from a pressure regulator drier so you can adjust the air pressure to the needed value.

Hello

You have re created a very old technology. I was looking at info from other companies that did a very similar process over 30 years ago. When I worked for Louisville Slugger in Santa Fe Springs, Ca. Have fun with it .

Ben

I have created no such thing. I have already done this before using the three roller rotary cam plate and lever as shown in the CAD rendering. That's my design and it worked. The job now is twice as big in diameter, twice as long and the tube wall is twice as thick. Please give me some credit. The customer only wants a half dozen of these things and I'm late. This is not my free time but I'm stuck. I need to learn.

If pneumatic pressures are adequate, to get control of the movements do it hydro/pneumatically. This gives you the pneumatic force (if that's all you need) and the ability to control the movement accurately with the oil in the system. It can be much cheaper than a pure hydraulic system.

Best regards, stay safe.

John

In pneumatic valves,they always preload the valves with air to eliminate stiction(static friction) and hysteresis.

This old technology was capable of very precise control of valves,even in nuclear applications.

Digital controls took over due to cost and flexibility of programing via computer.Compressed air costs more than steam to produce.

Some pneumatic valves are still used in certain environs where electronics cannot survive or are dangerous,and they are controlled remotely and electronically by I/P (Current-to-Pressure) transducers via computer.

The technology is not totally dead yet.

I would suggest an adjustable preload spring and a small amount of preload to reduce hysteresis and improve response.

Design piping with as large a diameter as feasible,with the supply source in the center of a "tee" feeding both cylinders.

Calibrate both cylinders to move together with a manual regulator by adjusting the preload springs individually.

If you want even more accuracy,you could incorporate a valve positioner on each cylinder.

Refer to the link on pneumatics that I provided in a previous post.

Or,if you want to go more expensive and high tech,try this:

Some have an internal LVDT for ultra precise position control.

https://www.enfieldtech.com/Products/S2-Cylinder-Positioning-Systems?utm_source=bing&utm_medium=cpc&utm_campaign=S2%20Positioning&utm_term=%2BPneumatic%20%2BPositioning%20Valve&utm_content=Air%20Cylinder%20%26%20Servo

Not the whole drawing but most relevant part:-

Exactly. It's only a sketch but it dissapoints me that people whom I was counting that really understand these systems aren't showing up. That self-centering air vice that SolarEagle posted is pretty much it but too small for my needs but it looks pretty straight forward. What's the mystery there? So far I get that parallel cylinders is the way to go. Whatever other contraptions I'll need upstream I guess I'll have to figure out for myself because nobody here seems to be offering definite ideas about that. This should be basic for those who really know the subject but I have to learn this stuff. Wish me well.

Bassnut - I would look at parallel gripper products for ways to synchronize and balance your cylinders. (Schunk, Zaytran, PHD, SMC...) Buying a large force parallel gripper can be pricey, but you can duplicate the concept yourself for less if you don't need all the bells and whistles.

Most large force grippers I have used or built have racks & a pinion gear linkage to sync opposing cylinders. The racks & pinion gear are only taking the balancing load and don't need to be too large and the basic components can be bought out of a catalog. If you can fabricate and machine what you sketched you can probably build this. You'll spend more in time than you'll save in money, but I'm assuming your time is "free".

Force is a function of air pressure so get a good regulator. You could get fancy and get a soft start valve as well so the rollers don't crash in to the work but build force slowly.

Good luck

Why use 2 cylinders when only one could be enough ?

If you use 2 you are confronted with the possible force differences which ask for a very complex control.

If you use only one you get same force on both sides but you should let the two sides center on the tube + mandrel. The cylinder connects the 2 sides so you are sure to get same force on both roller pairs.

Thanks. nick name. I don't know how to do that efficiently and get the required force. That idea keeps cumming up. The implication is that I'll have to adjust manually using the cross feed wheel on the lathe's traverse. Not impossible but not exactly ideal. The nature of the taper is that both sides have to adjust to the reduction in diameter as it travels. Both need to react. One side can't be static. It has to self-center.

... I'm sorry, I ran out of edit time. The other idea that comes up a lot is the "robot pincher " model. Same deal. Inefficiency. I need to bring to bear around 1K lbs force minimum per side to coax this tube to lay down on the mandrel. It's 4-1/2" dia going down to 2" eight feet down the line. I'm figuring having a head-room of 1500 lbs of force per side utilizing 120 psi of compressed shop air.

You should think so : both sides are connected with the cylinder, both sides can move radial so that when the cylinder pulls them together they come BOTH on the pipe you want to deform. I shall make a sketch since I think, it is the simplest and safest approach for you to have same force on both sides and be sure that the contact is centered on the mandrel.

You can have either matched displacement or you can have matched force, but not both without some serious control systems. Using air over oil and operating the two cylinders with hydraulic would allow you to run the cylinders in series and match displacements.

However, I would suggest something simpler. Use a single cylinder and put both sets of rollers on a sliding base. One set of rollers would be fixed to the sliding base and the second set of rollers is driven by the cylinder. The one cylinder provides the force and the sliding base allows the rollers to float to center on the work-piece and matches that force on both sides of the work-piece.

I have a three roll knurling tool like the photo below that I occasionally use with smooth rolls to spin down metal parts. In fact, I used it to spin down .223 shell casings over a mandrel to make Nagant pistol cartridge shells. Replacing the screw thread with your cylinder would accomplish something similar for your tapered tube.

I would caution that given the amount of tapering you are trying to do, you may need to do it in stages with anneal steps to prevent the tube from forming ripples and folding over laps or tears. You will also see some tube length growth that will need to be accounted for.

Looking at my four roller design, what's the worst that can happen? Notice that each set is attached to dove-tail slides with gib adjustments. You can oil them and adjust them to some degree to compensate for timing by introducing drag. It's maybe not ideal. There's no guarantee that it will be consistent over time. I liked the idea of a slow contact feed mechanism to avoid slamming that JRiversW suggested and allow time for more equal centering.

I think I would just use the air as an assisting force, in other words dial in the pressure to before it begins to actually deform the pipe and then do the shaping manually with a lever the way you were doing it before but with a better mechanical advantage system of levers...that way the work could be better controlled...sort of like power steering...or power brakes....it just seems like the system will be too clumsy to operate properly without a lot of sophisticated controls built in and that seems to me would take quite a lot of time and expense to work well for the operator...I'm not saying that will be the case, it's just my opinion, it's possible you see a clear path to success that I don't see....anyway good luck, and feel free to keep us posted...

The 2 roller supports are free in their dove tail the force loop closes through the tube so automatically they will be centered. On both sides you get same force and if the roller can oscillate they will apply same force even if there is a height difference.

The sketch is only for orientation you can use either one cylinder or 2 in parallel as you wish.

That's amazing nick name! Thanks for doing that! It looks clean and simple. I'm thinking though that SolarEagle had handed me exactly what I needed with that Heinrich self-centering air vice very early on. I kept wondering how and why it worked. I finally looked. Check this out. Scroll to the bottom and look at the video. I can make that with parts from McMaster-Carr.

"The jaws are connected by a gear and rack assembly concealed in the vise base so that they move together equally on both sides of the center, within an accuracy of .001".

Heinrich self-centering air vice

Sorry, nick name. I ran out of editing time and then ran out day time. I just wanted to thank you for your idea and taking the time to draw it and present it. It is both simple and elegant and not difficult to build. It's beautiful. I want to say "ding, ding, ding! We have a winner here!. Close thread." I now have some clear options which I didn't have before coming here. I was stuck. Thanks to everyone who pitched in to help. This was not a waste of my time by any means. This a vital resource which I'll recommend. Please understand that I'm not the owner of this company. I just got handed the task as what usually happens and I have to figure it out. A lot of times I just don't know what to do exactly and it takes time. Ideas need to build and develop. That's why I don't get the big bucks.

Looking at the CAD drawing again.

If both cylinders have a common air supply, the fact that one reaches the tube first should be of little consequence, as it will only develop any force when both cylinders are in contact with the tube.

There is a good chance that synchronising the movements is not necessary.

It maybe worth a try if you have the cylinders and can "cobble up" the mechanicals.

John

You will get imbalanced forces from un-linked cylinders which would push the rotating tube and mandrel off center. How much is unknown and likely not repeatable.