Reply With QuoteA die set? Three plate set. Top fixed... bottom fixed... middle plate moves on bushings on two... three... or four posts.
Referring to the attached drawing.
(And attempting to walk that fine line between too much information, and not enough.)
This is a air over hydraulic press that I am building.
The Anvil sits on a movable plate, which is raised by the jack
and is the only moving part of the press assembly.
I need an idea on how to keep the moving plate perpendicular (horizontal)
to the plane of travel (vertical).
A die set? Three plate set. Top fixed... bottom fixed... middle plate moves on bushings on two... three... or four posts.
Two thoughts: First, the natural inclination is to use linear bearings or bushings. In my opinion, these are the most commonly misused tool in the designer's toolbox. They are meant to handle direct loads only, but many people put them in situations where they are resisting a twisting moment - a recipe for binding and failure. They are intended to keep a point on a linear path, not to keep a body from rotating. If you want to use them for anti-rotation, you must have two bearings on a common rail (to dissolve the moment into individual forces), and the farther you can spread them out the better. The side load each bearing will see is a direct function of the twisting moment and its distance from the other bearing.
Second, another method I have used successfully several times in the past is shown below. Attach two chain loops to a common shaft. Neither chain can move without the other moving an equal amount. Then attach one corner of the platen to each chain. You're using the torsional stiffness of the shaft to guarantee alignment of the platen. No need to drive the chain. Its not high precision, and not meant for all applications, but it works well where its appropriate.
Let us know what you end up doing.
Dayle,
No need to try to reinvent the wheel. As RW suggests the die set was designed for exactly this operation.
Do some research here...
http://www.lempco.com/products.html
Maybe even invert the jack?? A couple of springs to pull it back??
Thank you for the suggestions.
The dual chain idea seems like a great idea.
Easy, tolerant of dirty environments, repairable.
But, a few more moving parts than the die set idea.
(More places for clothing / etc. to get caught.)
The die set option seems to depend on tight tolerances
sliding on polished surfaces (when using plain bearings),
combined with the rigidity of the moving plate.
This press will be used in a dirty environment,
so I am a little concerned about binding and wear.
But there is a ball bearing type which should work just fine.
I would like to build rather than purchase;
I will redraw the press using the die set idea and ask for your comments again.
Thanks.
I have added 4 posts, one in each corner of the moving plate.
The moving plate will slide vertically along each of the guide posts.
Any guidance on how close the tolerance should be?
I am thinking the holes in the moving plate should be about .002 inch larger than the guide posts.
The holes should also be slightly chamfered to help resist binding.
Or am I going about this all wrong?
More info please. Stroke? Maximum forces involved? Anticipated moments from off-center forces? Level of precision required?
Am I seeing this correctly - no bushings at all? Just steel on steel?
Way back I built a couple of fixtures that utilized three plate sets doing what appears to be, a similar operation. The operation requires precise alignment of a male and female component for an injection process. Before you run out and buy plate and have it cut... ground... and machined... and buy posts.... and machine... and bushings. blah...blah...blah... You ought to look around for a set that's built. The cost of an off the shelf unit might surprise you. There are a boatload of folks whipping these up. Danly is one that pops to mind. My guess is that unless you have all the capabilities in house you might spend more building one from scratch. At minimum you're gonna want to incorporate off the shelf bushings I'd think.
Regarding the bearings... The ball bearing type plane linear bearings wouldn't be my choice for a dirty environment. Better off with a greasing schedule on plane bearings with a crossed over pattern of grease grooves. (I like to type the word groove... always has me saying groovy in my head...) Far as clearance goes... like JBoggs mentioned need more info. The application I built involved the unit living in a heated box and heat strips on the plates for approx. 180 degree environment. After assembly I found I needed to hone my bushings out a breath. For a starting point you could safely go with .001 per inch of diameter but in all likelihood you'll probably end up poking them out a thou or two afterward to keep the set from binding up. (That assumes a precision set that was machined very exact as a set.)
Then there's the pressure on the plates and how thick they're gonna need to be... And how big will your posts need to be. If the plate bends some and you're gonna bind. To reduce the thickness requirement of the die set plates you can cheat a little and spread the load. The cylinder rod could mount to a base that bolts under the plate that covers a large part of the plate suface. ... Beam bending cntered load vs. spread over beam thing... plus the added strenght from the plate. Your molds or whatever will do the same for adding strength and spreading the load.
Along the line of Dave's suggestion about inverting... I was wondering if you might want to consider doing the work on the "first floor" rather than the second and leave your jack upright located on the "second floor?" This way you have the larger surface area of the base of the jack pushing on your moving plate and could bolt your rod through the top plate to a larger load sharing plate making the thing nice and rigid. That'd depend partly on where the operator is gonna be and what's comfortable to load and unload etc. Oh... and if the "jack is double acting or a just an actual bottle jack?
Good idea to toss in some extra holes for pins while you're machining the plate too. This way your tools can be accurately located and should you toss some other tooling in there you might get lucky and be able to utilize some pin hole you already have in place.
Good Luck!!
Bob
Dayle,
You are headed in the right direction. The bearings on the sliding plate will need to be more than just holes in the plate as that will easily can't and lock up with just the single centered thrust from the jack. Make them as long as you can accommodate within the movement range. If the "Anvil" plate is moving then I would make some bearings hanging down at least 5".
For sliding tolerance and fit refer to the "Bible"That's what you paid the 60-bucks for it for.
"Stroke?" about 4 inches.
"Maximum forces involved?" the jack is capable of developing 12 tons of force, and I use all of it.
"Level of precision required?" I was pretty careful during machining of the punch / anvil.
Each piece was machined independantly.
The plunger will fit, and work, with punch #1 on the front left corner or you can rotate the plunger
180 degrees, putting punch #1 in the back right corner, no problem. In order to acheive that kind of
flexibility, all machine operations were kept to +/- half a thousandth.
Having said that, this is essentially a hand operation.
The anvil can slide around a bit on the moving plate so that it is correctly positioned
under the 40 punches.
The punches are a sliding fit in the anvil.
The operation involves filling the 40 holes in the anvil with a damp powder compound.
The plunger presses the compound into .5 inch high by 7/16 diameter cylinderical pellet.
It takes the entire 12 tons of force to achieve that.
Currently, I am using a HF hydraulic (manual) press to create these pellets.
I want to automate it slightly, by using an air actuated hydraulic bottle jack.
And I am willing to dedicate a machine to this operation,
so I am building my own version of the HF press, but smaller and (hopefully) better quality.
Besides being able to withstand the vertical forces involved,
really the only other trick is to keep the moving plate from binding during the operation.
HF achieves that by making the fit so sloppy it is impossible to bind.
But this can put a rather significant tangental load on the plunger/anvil.
(the moving plate will move to one side about 1/4 inch, but the force is still applied vertically,
creating a off-axis force.)
Bushings and bearings are by far my weakest area of knowledge of machine design.
I will do some research on what is available and how to apply them.
"Along the line of Dave's suggestion about inverting..."
I see the logic of using the jack's built in base to spread the load.
I will move the jack to the "second floor" and re-draw.
Since the jack is essentially just a bottle jack,
I will add some springs to assist with the 'return'.
"Then there's the pressure on the plates and how thick they're gonna need to be... "
I have not done any calculations on that part yet,
but I have pretty thick steel plate (1.25") on hand.
The frame was going to be build out of .25" wall tubing. (but I need to do the calcs)
"Good idea to toss in some extra holes for pins while you're machining the plate too."
I wish I would have thought of that.
The plunger/anvil is done (and works fine), but I might add a indexing / guide pin hole.
"For sliding tolerance and fit refer to the "Bible"".
Oh yea. I forgot about that. Thanks.
Dale,
I gotta ask...
Are you making pellets for a heater or furnace out of sawdust?
Ron
Same idea, differant application.
I will post some pictures of the press in use (when it is done),
and also a picture of the pellets in use.
Didn't they have black powder pellets for faster reloading towards the end of the Civil War?
I have re-drawn the powder press to include sleeve brearings,
and to show it in operation.
The longest bronze sleeve bearings I could find are 4".
I think I have them in the proper location.
Do I?
Any other comments about the likelyhood of this thing working (or not)?
Hi Dayle,
It looks do-able. Have you thought of building a prototype that makes say, two pellets at a time to test the process? Seems like a lot of work to go to and have the pellets crumble as they are pushed out after the removable plate as been extracted to complete the stroke. They may need impact to bind the powder not just steady pressure. Just guessing though.
I am not too worried about the pellets crumbling.
I am using essentially the same process now, but using a standard bottle jack.
The 'new and improved' version will use an air actuated bottle jack.
Thanks for the input.
Time to start building.
Just an observation...
The bushing parts extending past the thickness of the moving plate really won't be adding much toward rigitizing your moving plate. Or maybe better said... Any unsupported portions of the bushing aren't going to do much. If your prints are scaled, I really doubt you'll need that addtional length to keep the plate on track with four posts anyway.
Dayle,
I hate to disagree with RW, but I would certainly not go below about 2" of bushing material. Specially if the brass wall thickness is less than 1/8". Over the platten area you have no control over how well or evenly the powder will fill the holes. Using a point source of force to raise the moving platten, coupled with unknown load distribution, jacking could be a real possibility.
As RW alludes though, unsupported bushing would not be the best option with thin walled bushes.
You could try using these. Maybe just one per corner or maybe one on top of the sliding plate and one on the bottom for each corner.
http://www.grainger.com/Grainger/PBC...-Bearing-2CPL3
Ok, thanks.
I was wondering about that.
And I see how the thick wall flange of those kind of bearings would be better.
Since I understand the concept, I think I will just make the 8 supporting collars
instead of spending the $72 or so for each.
I found out something interesting today;
if I use UHMW instead of brass for the sleeve bearing,
it will only cost about $7 for the raw UHMW stock needed for all four guides.
hhhmmm... I wonder about just making the whole 'flanged sleeve bearing'
out of UHMW. Maybe with a 3/8" wall or so.
Still inexpensive, easy to machine, and low friction coefficient.
There must be some info on that stuff somewhere around here...
One problem with UHMW material is that it does not like pressure. It will move away from pressure, as in cold-flow. Typcailly UHMW bearings have a metal backing to contain them. Yes they are slippery, but not good for load unless constrained so the material cannot elongate and squoosh (technical bearing termOriginally Posted by dalecyr
Given the closing pressures and no control over sideways forces other than the four sliding bushes as the powder is compacted, I would be using a very stable and stout bearing system. With the brass bushes any lateral forces will be taken by the plate thickness. With UHMW the load will squoosh grooves and cause misalignment problems with the powder rods.
With real bearings designed for that purpose, $72 a corner is a lot cheaper than replacing bent powder rods and sliding bushes a few times until you do it he correct way.
Posting Permissions
