Continue to Site

Welcome to MCAD Central

Join our MCAD Central community forums, the largest resource for MCAD (Mechanical Computer-Aided Design) professionals, including files, forums, jobs, articles, calendar, and more.

difficult pump assembly

soboyle

New member
I'm having problems getting mechanism to allow motion for an assembly, see attached pic. The cylinder block spins around its own center axis, and the bearing spins around its own center axis, the piston spins with the cylinder block but stays mated with the bearing face, so that it comes in and out of the cylinder bore. I've tried every combination of connections I can think of, and can't get motion after they are assembled. Any suggestions on how to make this work?


Any general suggestions for getting difficult connections to work?
 
you sometimes have to place one connection in to an assy then use another connection type when you place that sub-assy into the main assy.


make sense?
 
Can you give a more specific answer, because I've tried every trick I can think of, like using combinations of connections, like bearing and ball. Currently the cylinder block and bearing are each parts of seperate sub assemblies, and I am attempting to assemble the piston in the top level assy. Should I try to partially assemble the piston in the cylinder block sub assy, then after assembling the cylinder block assy into thetop assy, complete the piston constraint?
 
if i undertand you correctly , yes put the pistons in the cylinder block with a slider connection.. then add the cyl-pist sub assy to the main assy with a pin


then your pistons stay in the bores while they go in and out while the block turns..


and are these pistons going to move independent or as a set? cuz if it a set then pattern the pistons and assembly with just the one slider connection.. and they will all move together... you'll have to play with your zero ref to get the piston where you want on start
 
I've solved this problem in the past - typical swash-plate compressor, although mine had the plate rotate while the cylinder block was stationary. How exactly are the bearing race and pistons connected? There should be some type of "elephant's foot" that provides a ball joint between each, one for each piston. If this is the case, simply look to this exact hardware to find your solution. If your model works exactly as you have depicted it in the image, where the bearing race edge contacts the piston bottom directly, the only options I can imagine are a cam pair or some type of slot arrangement.<?:namespace prefix = o ns = "urn:schemas-microsoft-com:eek:ffice:eek:ffice" />

A cam using surfaces may not work unless you try to create an extra curved surface on the bearing race that would satisfy the
 
Thanks, I have done this before with a different type of piston pump, but this swashplatetype is giving me a hard time. I'll try some of the things you suggested.


This is a low cost light duty piston pump, so the piston's spherical ends run right against the bearing face, so no intermediate piece to use as a ball connection like I did last time. My best success to date is to use both a bearing and ball constraint, but I cant get proper motion out of it, but at least it moves, all other connections I have tried so far have just locked it up. What I really need is a simple tangent connection, so I can hold that piston against the bearing face.


One trick I tried is to offset a datum plane from the bearing face, then use that datum to locate points where it intersects the piston center lines, then a ball connection from the piston to the point, I thought that would work, but no movement after that connection. Frustrating! This stuff should be easier.
 
Try creating a curve in the assembly parallel to the bearing face were the center of the piston spherical radius crosses. Then create a point at that location in the piston and use a slot connection.
 
PRStockhausen, thanks for that tip, the pump is variable displacement, and the bearing tilts from 0 to the 15 degrees shown in the diagram. Can you give an example of a curve that will work in all postions?
 
From what you say I take it that the angle from the bearing to the block changes while the block is moving. Without more detail of how things move, I could not determine a curve. I have at times used apart with only an axis at each of the intersections of the default datum planes and a point at the origin. I would assemble these parts such that with a few of them I would be able to constrain the actual solid parts so they moved properly with each other.
 
That is correct, the bearing changes angle as cylinder block rotates. I thought something simple like a straight line curve, from the centerline of the cylinder block piston bores out touching the face of the bearing would work. The curve end away from the bearing would be at the center of the sphereof eachpiston. But the curve will not follow the cylblk when it is spinning. It seems that what is making this not work is the references I have to make from the cylblk to the bearing when sketching the curves. I've tried making references to only assembly datums as well, with the same results. This seems like a relatively simple motion to want to do, yet it is apparently so difficult. If anyone has suggestions I am all ears!
 
Create a plane or surface in the bearing offset by the spherical radius of the piston. Then create a general mate aligning a point at the center of the radius in the piston to that plane or surface.
 
YES! (doing a victory dance), YES, that worked! I offset a datum plane in the bearing part, not the assembly, then used normal constraints, not connections, point on plane (from the piston sph center), and an axis align. Simple, and I was thinking much more complex when it wouldn't work with similar attempts earlier. Many thanks PRStockhausen.
 

Sponsor

Articles From 3DCAD World

Back
Top