PDA

View Full Version : Pneumatic power unit design



Evan
03-15-2006, 07:46 PM
I've been thinking of building a power unit for my hobber/slotter/etc gizmo I just completed. I had a slow day today so I designed such a unit. It's air powered with an integral hydraulic damper. The damper is adjustable during the main stroke and also has a cushion at each end of stroke when the main ports of the adjustable circuit are passed.

On 100 psi air it should produce about 470 lbs of force. That is more than enough to operate the hobber unit directly with no advantage and the effort may be controlled with an air regulator.

I'm interested in any comments or ideas you may have, criticism included.

http://vts.bc.ca/pics/pc1.gif



[This message has been edited by Evan (edited 03-15-2006).]

CCWKen
03-15-2006, 08:06 PM
The hydraulic dampening is a good idea. Pneumatic cylinders can be unpredictable at times. Looks good but I see a lot of drilling and tapping! http://bbs.homeshopmachinist.net//biggrin.gif

topct
03-15-2006, 08:07 PM
My first thoughts.

The action you are trying to perform has been done mechanically.

Why such a gooshy(sp) approach?

pgmrdan
03-15-2006, 08:54 PM
Are you sure those little 1/8" thick tabs you're going to run the screws through are going to hold up under 100 psi x 4.7 sq. in.? Those end plates are each going to get about 470 pounds on them too.

Me thinks you've designed a bomb! Sorry.

[This message has been edited by pgmrdan (edited 03-15-2006).]

joeby
03-15-2006, 09:01 PM
Evan,
Would a simple air over hydraulic setup work. You still wouldn't have the hassle of a pump; but it might operate more consistently. Air cylinders seem to "stick/slip" too much for good control.
Just a thought.

Kevin

Evan
03-15-2006, 09:53 PM
A few of the design considerations are that I don't want any runs of line filled with oil going to valves or an accumulator. In this design any oil leakage from the damper ends up in the air cylinder and is exhausted with the air.

I can't see stick-slip happening when the force required is at least 50 lbs and more.

As for mechanical approaches I need to be able to accomodate infinite variable stroke. That's hard to do mechanically.

As for the 1/8" flanges I am pretty sure they are strong enough. That's just a gut feeling and I will have to do some analysis on that. It's easy to change if needed and is a good point. I have always had a very good feel for what is strong enough and how much a particular design can take.

Yep, lots of drilling and tapping. I do a lot of that. I even broke a tap one time. http://bbs.homeshopmachinist.net//biggrin.gif


I would like to find an easy way to adjust both the end of stroke damping and the point at which it occurs without getting too complicated.

Weston Bye
03-15-2006, 10:17 PM
Evan,
Why not just go straight air over oil? Use two pressure reservoirs and a simple hydraulic cylinder. Control the air applied to the top of the reservoirs forcing the oil below into the cylinder. Control the speed with hydraulic flow controls (full flow into the cylinder, metered or adjustable flow out) between the reservoirs and the cylinder.

In addition to the standard air/oil operation, it can be modified with the addition of a hydraulic air driven intensifier; a small hydraulic cylinder driven by a larger air cylinder to provide more force for other applications.

You might also consider constructing your cylinder with a steel tube sandwiched between two stout blocks and held together with long threaded rods.

Your design will require all of your very capable efforts to assure concentricity.

Wes

Evan
03-15-2006, 10:23 PM
A quick calculation shows that it would take about 35,000 lbs of force to shear the end plate from the 1/8" flange all at once. With a circle of screws holding it the strength will be enough. To shear just a 1" section loose would take about 4000 lbs. I think there is enough safety margin.

snowman
03-15-2006, 10:27 PM
evan

why the oil damper?

why not just use an air flow control? i've got a bunch of em...smooth everything out real nice. skip all of the fluid problems.

-jacob

Evan
03-15-2006, 10:28 PM
Wes,

I would be interested to see a schematic of what you have in mind if possible. As for concentricity the only critical parts are the two end plates and they would be machined together. The piston in the damper need only be a clearance fit since leakage is required and can be machined to diameter once mounted on the shaft. Same with the air piston except it would have an O-ring (not shown).

Evan
03-15-2006, 10:35 PM
"why the oil damper?

why not just use an air flow control? i've got a bunch of em...smooth everything out real nice. skip all of the fluid problems"

I want to run it automatically once the damping is set using solenoid air valves I have. This is intended to be a nearly zero cost project. I also don't want fluid problems and I tried to minimize that possibility with this design.

Exactly what do you mean by air "flow control"?

wierdscience
03-15-2006, 10:44 PM
Looks good,I do have two suggestions thou.

The cylinder end caps I would simply turn an o-ring groove on the od and the id,then machine a step in the ID of each end of the cylinder deep enough to accept the ends caps plus enough for a 3" internal snapring,that would save you a lot of drilling and tapping and eliminate the bolt on packing glands.

The other thing would be make the back end cap and the dampener cylinder one piece with it's end cap retained with an internal snapring as well.

Idea being to eliminate as many screw connections(potential leaks)as possible.Just a thought.

If you need any chromed ID cylinder tubing let me know,yours for the postage.

Evan
03-15-2006, 10:59 PM
Snap rings that size are no problem to find around here so that sounds like a good idea Weird. This is heavy equipment central and we have at least half a dozen machine shops in this town and related suppliers because of the logging industry.

As for machining things from one piece I am trying to avoid that simply to save material. I have the stock but hate to make lots of little chips from what used to be solid material.

Thanks for the offer of cylinder tubing but I can easily get that from the scrap bin of my favorite job shop.

Rustybolt
03-15-2006, 11:19 PM
I would like to find an easy way to adjust both the end of stroke damping and the point at which it occurs without getting too complicated.


Use a spring.

tsmartin_98
03-16-2006, 12:05 AM
Interesting design. I'm guessing you must have a valve of some kind in the piston on the hydraulic side. I'm also wondering about having the hydraulic side piston going past the ports for the needle valve instead of being centered between them.
Might that cause uncontrolled motion until the piston passes the ports or maybe cause the whole thing to lock because hydraulic fluid is incompressible?
Just some thoughts here.
TS

wierdscience
03-16-2006, 12:21 AM
Evan,a thought just popped into my head,why not make the hydraulic dampner a simple hydraulic cylinder,that way you could run the air cylinder up to full system pressure and have full control of it's movement simply by controlling the flow of oil from one side of the piston to the other?The tranfer ports could be machined directly into the end caps and connected via a couple short grease gun hoses and a needle valve.A differential needle valve would be primo of course allowing for a higher return stroke speed without losing the original forward stroke speed setting.

Too much caffiene for me today http://bbs.homeshopmachinist.net//biggrin.gif

snowman
03-16-2006, 12:34 AM
flow control is exactly what it sounds like, it limits the flow of the air...just slows things down and keeps things from slamming back and forth.

email me jrunyan02 @ yahoo . com

barts
03-16-2006, 01:35 AM
The problem w/ air flow control is that if the load is very
uneven, the movement of the cylinder will be as well...

- Bart

Evan
03-16-2006, 01:36 AM
I like that idea wierd. No hoses though, I would build the valve body right on the unit directly. That still poses the problem of automatic control though. The air type systems with a large diameter cylinder relative to the stroke are pretty easy to control. It is very similar to a rivet squeezer used in the aircraft sheet metal business. I have used those many times and they are very controllable with just a hand valve.

"Might that cause uncontrolled motion until the piston passes the ports or maybe cause the whole thing to lock because hydraulic fluid is incompressible?"

If you look at the piston plate on the hydraulic side you will see a small port at the top. That is to allow oil through when the piston goes past the ports at each end.

That provides the cushion effect during the last 1/8" of travel. Also, since the idea is to act in a similar fashion to a shock absorber the piston will not be tightly sealed in the bore. It needs to leak in order to move and the needle valve is just to control the majority of the "leakage".

Evan
03-16-2006, 01:42 AM
"You might also consider constructing your cylinder with a steel tube sandwiched between two stout blocks and held together with long threaded rods"

I missed that comment. Yep, done that. That's how the cylinder I made for the snow plow on my land rover is made. However, it needs to handle 2000 psi, not just 100.

wierdscience
03-16-2006, 02:01 AM
Automatic control?You want it to hit a limit and return right?

Okay,how about extending the rod right through the back end cap and threading it for an adjustable stop.That stop could move forward with the tools advance and at a pre-determined point trip a relief valve and dump the air pressure from one side of the air cylinder piston to the other,similar to a "D" valve on a steam engine.If exact control is required,which would be very nice, a seperate "hard" stop could be used in conjunction with the relief system to eliminate over travel.

It would be super neat,if the two stops were built on the turret principal.Just think,stops for 10ths,100ths and 1,000ths.If that could be tied in with the indexing,boy would that be sweet.No more "TOO long 100ths on the first set of leadscrew dials.

Caffiene still not worn off http://bbs.homeshopmachinist.net//biggrin.gif

abn
03-16-2006, 02:08 AM
I don't know enough to help...I'm just wondering how you are going to control stroke...magnetic switches and the solenoids you speak of???

I ask because I have an air cylinder at home, I also have some solenoid switches and I'd like to have an infinitely variable stroke as well...

I was going to start a thread on it, but I figured I should probably do some web searching first.

Scatterplot
03-16-2006, 02:11 AM
<font face="Verdana, Arial" size="2">Originally posted by Evan:

"Might that cause uncontrolled motion until the piston passes the ports or maybe cause the whole thing to lock because hydraulic fluid is incompressible?"

If you look at the piston plate on the hydraulic side you will see a small port at the top. That is to allow oil through when the piston goes past the ports at each end.
</font>

There's still the matter that without something else changing volume, you have a constant volume of liquid and a varying volume of chamber. At full stroke, the chamber will be at maximum volume. At minimum stroke, the chamber will decrease in volume since the piston shaft will now be in the oil. No matter how many holes you punch in that thing, it doesn't matter what side the oil is on the cylinder still won't move.

A possible remedy to this is either the aforementioned accumulator, or to simply make the bore go all the way through, with the other end sticking out into the air. This way there's always a constant "bore volume" in your hydraulic chamber.

abn
03-16-2006, 03:37 AM
Good call scatterplot, you always have to accomodate the shaft volume...just like on any other hydraulic damper.

As an aside, a while ago i spent some time reading this site and there are some fun pieces on damper history and design:

http://www.taylordevices.com/tayd.htm

In the papers section:

http://www.taylordevices.com/papers/history/design.htm

Also interesting there is:

http://www.taylordevices.com/MachinedSprings.htm

Anyway, I don't know if this is a help or a hijack, but it's presented for your enjoyment http://bbs.homeshopmachinist.net//smile.gif

(edited for spelling)

[This message has been edited by abn (edited 03-16-2006).]

Forrest Addy
03-16-2006, 03:39 AM
The oil damper (for want of another term) is a good idea. Air as a driving medium is compressuble. Air cylinders are notorious for stalling until sufficient air pressure builds to move the load (as in a shaper when a tool enters the work) and lunging if suddenly unloaded (as when the tool emerges from a cut). The oil damper coupled to the ail cylinder rod will keep the tendancy to lunge and stall under close control.

One snag: Apparently you want the oil in the damper to recirculate. Problem is theres an area differental in rod ende Vs piston end. Your design will work if you extended the rod out the end of the tamper so the piston area is the same on both ends but you'll still have to account for espansion. Maybe a little surge tank.

Also energy will be expended in the oil end because us turbulent/shear flow through the metering valve. Have you come up with a way to deal with the heat?

david_r
03-16-2006, 04:06 AM
Good catch Scatterplot.

Evan,
Looks like a fun project.

If you are going to build the hydraulic valving right on the body, double it. Make two pairs of holes. Your originals plus a pair at the extreme ends. These will allow you to adjust your end damping rate.

Put your air ports on the cylinder instead of the end caps.

How are you sealing that rather complicated union between the air cylinder and the hydraulic cylinder and the other face bolted pieces such as the endcap on the hydraulic cylinder?

What holds your air piston on the rod? I think I might see a set screw there. Is that going to hold up to 470lbs of force? I think snap rings might be a better solution there.

joeby
03-16-2006, 06:48 AM
Evan,
I see what you mean about fluid leaks, another one like this could mean a lengthy lathe cleanup!

Kevin
http://vts.bc.ca/pics/spray1.jpg

JCHannum
03-16-2006, 07:22 AM
For a relatively simple application such as this, an off the shelf air cylinder with flow controls and cushioned ends is more than adequate.

Many are available with built in magnets on the piston and external hall effect switches for stroke control.

Bimba, Compact Products, Allenair and SMC are a few of the hundreds of manufacturers. Do a search for them, and look at their catalogs. Many will have cut away drawings you can copy if you want to build your own.

By the time you find and buy the proper seals and odd pieces, it might be cheaper to buy a complete cylinder.

Rustybolt
03-16-2006, 09:25 AM
Looks like you're reinventing a Shreader-Bellows hydracheck if you're going to control the stroke.

Evan
03-16-2006, 09:36 AM
Definitely rev 1.1 coming up. Scatterplot is right on the money on that problem and I totally overlooked it. Thanks for seeing that, thats why I posted this here. I don't think I would have noticed that problem until I tried it.

Extending the shaft out the end will fix that and gives me the chance to put an adjustable hard stop on it. I'm still not sure about the reversing idea wierd, I have a dual solenoid air valve body that runs on 110ac that is trivial to hook up with solid state relays and use for control.

http://vts.bc.ca/pics/aircontrol.jpg

JC, I have most all the pieces I need to make this. The short length of 3" tubing is all I need to find. You wouldn't believe the material that goes in the scrap bin at the local job shop I frequent. Last saturday I was in there and saw an eight foot length of what looked like perfectly good 2" chromed hydraulic cylinder shaft in the bin. I'm guessing it was slightly bent. Too bad it wouldn't fit in the back of my truck.

Besides JC, it's no fun buying something and then trying to modify it to fit the intended application. I would rather build.

Weston Bye
03-16-2006, 09:45 AM
Evan,
Here is a link to an animation that is very similar to what I am suggesting.
http://www.tox-us.com/products_powerpackage.htm

I have implemented a couple of these units in the past 6 months.

Years ago, I was involved with a machine for pressing wheels onto axles for Lionel locomotives, put together from cylinders, fittings and valves all piped together. It worked, but with the inevitable minor leaks and drips. When I build one for my workshop, (Evan, you are truly inspiring!) I will adapt some of the ideas from Kozo Hiraoka's feedwater pump systems from the various locomotive construction articles he has done for Live Steam Magazine. (Bookstore on the HSM website)

Wes



[This message has been edited by Wes1 (edited 03-16-2006).]

hitnmiss
03-16-2006, 10:26 AM
Nice idea and good comments.

I don't have much to add but a simple idea stolen from my r/c racing days:

They had oil damper shocks and one trick hop up I saw added variable damping as a function of stroke... A tapered rod fitted in the piston oil hole would close the oil metering hole in the piston as the shock traveled. Picture your needle valve above poking into the piston metering hole... One on each end of the stroke would give you the cushioned end stroke I think your after.

Evan
03-16-2006, 10:42 AM
"Picture your needle valve above poking into the piston metering hole... One on each end of the stroke would give you the cushioned end stroke I think your after."

Yep, I considered that. However, then I need to stop the entire piston assembly from rotating. Although the link to the device it powers will do that it isn't precise enough and I would like to avoid that consideration by making it irrelevant if it is rotated. This also gives more freedom in how it is mounted.

John Stevenson
03-16-2006, 12:40 PM
I think I'd be more concerned about a stonking great cylinder arangement hanging off the end of the slotting device.
You say you will get 460 pounds of force, what are you getting now with a decent pull and the mechanical advantage of the lever?

What about a remote geared motor driving a variable stroke crank linked to the slotter head by a truck handbrake cable.

Light, portable, adjustable for speed and stroke, quiet, no clattering compressor and easily made from scrap components.

Sir John.

JCHannum
03-16-2006, 01:30 PM
<font face="Verdana, Arial" size="2">Originally posted by Evan:
.Besides JC, it's no fun buying something and then trying to modify it to fit the intended application. I would rather build.</font>
The suggestion was to look at the existing information for design purposes. If you wish to fab, that is up to you. The application does not need anything exotic, adapting an existing, simple design will do the job.

Take a look at the cushioned cylinders and flow controls They can be easily fabbed if that is your thing.

The technology is existing and proven.

Evan
03-16-2006, 02:00 PM
John,

It's not very big, only 3" diameter and 6" long. It has a perfect place to attach to on the base that fits the cross slide and could be held with only two bolts, removable for hand operation in a minute. A piece of 3/4" plate would fit precisely to the existing part of the hobber and provide plenty of support for the power unit.

http://vts.bc.ca/pics/hobber1.jpg

"You say you will get 460 pounds of force, what are you getting now with a decent pull and the mechanical advantage of the lever?"

Estimate with 7 or 8 to one advantage from the handle it takes about 10 to 15 lbs of force on the handle to cut .020 deep in aluminum.

[This message has been edited by Evan (edited 03-16-2006).]

CCWKen
03-16-2006, 10:17 PM
I hate to imply version 2.0 at this time but why not simply go mechanical? A screw shaft powered by a motor. You can start and stop it anywhere. I'm not so electronically inclined so I'd be using simple micro switches or the like. Set up correctly(one NO, one NC in each directional travel), it would run all day--back and forth.

If it was Dave, he'd have the thing computer controlled. http://bbs.homeshopmachinist.net//biggrin.gif Some simple counter could be made to count strokes and pause to allow you to move to the next index. Hit reset and it cycles again.

Evan
03-17-2006, 01:12 AM
I have considered that. I have been looking around my stash to see what I have in a motor. I have one that is perfect except it is too slow. It's a hefty gearmotor with tons of torque and still small enough to work but it only does six rpm. I have a spare garage door opener and will have to look at that but they have terrible duty cycle so it probably won't do.

A nice big stepper motor would be cool. Computer control to the thousandth of an inch. Hmmmm....

[edit]

Hey! I have a spare motor off a camper jack rated for 5000 lbs, 12vdc. I have some spare 1/2" 10 tpi acme shaft laying about too...

[This message has been edited by Evan (edited 03-17-2006).]

john hobdeclipe
03-17-2006, 10:55 AM
Eliminate the cushion chambers at each end of the oil damper. I don't think you'll need them even if you use the full stroke. And if you set up a control to limit the stroke, then they won't neccessarily come into play anyway.

Make the air cylinder diameter larger, to apply more force. Then adjust the oil needle valve to absorb a greater percentage of this force. I think that if the flow control absorbs about 3 or 4 times as much energy as the cutting tool, it would prevent any lunging or jumping as the tool breaks through at the end of the cut.

Stick with this pneumatic power...If anything goes wrong and the cutting tool hits something solid that it can't move, there won't be as great of a risk of damage as there would be with a motor driven mechanical linkage.

Exhaust the cylinder air across the oil damper to help dissipate the inevitable heat buildup. Leave room in your design for cooling fins to be added if needed.

Be sure to provide a means to bleed ALL of the air out of the oil cylinder when you fill it.

Does that add up to 2 cents worth, or shall I babble some more?

Evan
03-17-2006, 11:05 AM
Good advice John. Most I have considered already. I was planning on cooling fins on the damper. I still like the pneumatic design for the reasons you mention. In particular, the ratio of the cylinder size to the stroke does play a big part in the controllability of air as I mention in the case of the rivet squeezers. I do have a space constraint but could probably fit a 4" cylinder in the available space.

One of the reasons I would like the air design is simply because it seems like a cool project and if I plumb in some air from my big compressor in the garage it will make little noise as long as I put a muffler on the exhaust.

I'm still working on rev 1.1 of the air design and am not quite happy with it yet. I'm going to build it as well as an electric linear actuator because I can see other uses for both.