Hi JC Biggs, I've called in the cavalry. I have issued a plea for help over at our LinkedIn group! Link Hopefully someone'll have an answer for you soon
cool... i feel kinda slow for not knowing how to figure this myself. i thought there would be some simple way to model it.. but i dont think so...
The way the problem is described, it can not be done. If the distance between the rollers is constant, then the orientation of the rollers with respect to the rotational axis of the disc must vary. If the axis through the center of the rollers is fixed parallel to the axis of rotation of the disk, then the distance between the roller centers must change- or the diameter of the rollers must change (i.e., deformable materials). Here's an illustration of the issue: [[url]http://cid-80cbf5359a7240ac.photos.live.com/self.aspx/Pictures/SwashPlate.png][/url] Hopefully that link works. The maximum distance will occur where the derivative of the function defining the plate surface is maximized or minimized (in this case at 0 and 180 degrees, the derivative of the sine function being the cosine function). The illustration shows what would happen as the plate is oriented at the 180 degree point vs. at the 270 degree point- the difference is significant if the axis of the roller centerlines is constrained to the axis of rotation of the plate. If the distance between the rollers is fixed, then the axis between the centers of the rollers must oscillate such that it is always perpendicular to the plate at the point of contact.
I don't seem to be able to grasp the procedure for adding images to the post. If you want to see the image, send me a PM and I will send it to you...
it is possible.. i found this axialvectorenergy.com not what i needed the design for, but same concept.
It is possible, within limits. You need a rotating plate that varies in thickness around the circumference. The trick is to work out how the thickness should vary. I'll see if I can work it out over the next couple of days.
Where is the force for the rotation coming from, and what is holding the ball bearings in place? Is the plate going to always be going the same speed as it spins or will it fluctuate? If it fluctuates then you won't be able to have a single path for the bearings to follow, you'll need a concave bowl impression on both sides to allow for all points of contact that would be covered as the speeds change, and that might lower the strength of the plate to unacceptable degrees. Why couldn't you just put pistons at the far ends of the ball bearings that give and take, so the plate can maintain its thickness all the way through? If you had the ball bearings attached to pistons and made the join between the balls and the pistons slightly flexible, it would add momentum during spin instead of drag. I think. Just trying to visualize in my head and that's the questions/answers I came up with.
I am not sure that i understand how what you want to do is different than a regular swashplate, which is usually a flat plate mounted at an angle. The double sinusoidal face cam drive shown in the video is similar, but it can have at best line contact with the followers. If you are going to use roller followers you will need to either allow them to pivot or vary the thickness f the plate. I dont know how to vary the thickness mathematically, but it is fairly straightforward to do graphically in 3D CAD. This is how I did it in SolidWorks. 1 - create a cylinder with a vertical axis 2- create the path you want, two times, each displaced vertically from the other 3- create a circular profile and sweep it along each (upper and lower) of the paths, with the profile(section) normal to the path. This essentially gives you offset curves a constant distance from your original curves. 4-use the "radiate surface" function (under mold tools) to extend each of these paths as a surface. level lines on this surface will all intersect the cylinder axis 5- fill in between these radiated surfaces using the extrude function, using the surfaces as the start and stop conditions for the extrude. 6-cut away the cylinder this is very similar to a winding cam I helped design. In that case it was a single follower riding in a slot. But I did follow the above steps and create the part you describe, which I could send to you if you have solidworks and could use it It is even easier to machine. just get one surface of the cam done, and then make a follower with one roller and one cutter. The roller follows the face, and the cutter (if it is the same diameter as the follower) will cut a perfectly matching opposite surface.