Two-axis Gimbal?

[ginsoakedboy]

Please view the uploaded PDF file. Will this arrangement work as a 2-axis gimbal? If yes, is there a software to perform kinematic (rigid-body) simulation of this arrangement? Can Unigraphics NX 6.0 do a motion simulation for this?

 

[IRstuff]

As a severely limited goniometer, perhaps, but hardly a full gimbal. It's doesn't even appear to have much azimuth motion.

TTFN

FAQ731-376
Chinese prisoner wins Nobel Peace Prize

 

[MikeHalloran]

It's not clear how you intend it to work, but in any case it probably won't work like you're thinking.

You could get a few milliradians of rotation around a remote center, provided that you arranged to slide the three elements relative to one another, and greased the sliding surfaces of course. That would work, except that your clearances are all wrong; the internal arc has to be shorter than the outer arc of each sliding pair, in order to allow the sliding to take place.

You could get a few degrees of rotation about a center within the pictured assembly, if the inner arc were actually smaller than the outer arc, so the inner arc would roll on the outer arc. There would need to be some kind of elastic tension arrangement near the hole to hold the stack together.

As drawn, you could get a few degrees of rotation of the parts about a center within the assembly, with the complication that the left and right rotations would be about centers displaced from the hole. I.e. the centers for rotation, four in all, would lie along lines perpendicular to plane of the image at the ends of the longer arc. Again, rotation would require an elastic tension member holding the assembly together.



Mike Halloran
Pembroke Pines, FL, USA

 

[ginsoakedboy]

IRstuff and Mike,

Thank you for the comments.

I only need very small rotations, like 2° max, but I need it to function as a 2-axis gimbal within that range. The arrangement is intended to be used for equalizing the compressive load acting on top face by adjusting the orientation if the compressive load is not perfectly vertical.

I do realize the need to have difference between radii of inner and outer arcs. Even if I made that change, the model doesn't "feel" right to function as a gimbal. What other changes would be needed to make it better?

What do you mean by ... "with the complication that the left and right rotations would be about centers displaced from the hole. I.e. the centers for rotation, four in all, would lie along lines perpendicular to plane of the image at the ends of the longer arc."??

Also, I don't understand "It's doesn't even appear to have much azimuth motion".

 

[IRstuff]

Perhaps you should back up and explain which two axes you're refering to, whether that really suites what you seem to be desiring.

From your drawing and comments, I'm guessing that you have 2 perpendicular grooves that the plate above is supposed to rock on, resulting ina nutation. If so, my comment about azimuth is moot. However, since your two grooes are offset, their rotation centers are not coincident, and therefore , the structure would not behave as you expect or desire.

I'm wondering why you don't simply have a ball and socket, which would have the center of nutation in one place, and closer to the structure, and allow for a much larger degree of nutation. Typical commercial ball and socket constructions move up to about 40 degrees.

TTFN

FAQ731-376
Chinese prisoner wins Nobel Peace Prize

 

[ginsoakedboy]

The attached image shows the different parts of the assembly. The 2 axes are X and Y. The vertical axis is Z.

In what sense are the grooves offset? What changes would I need so the structure behaves as desired?

 

[ginsoakedboy]

The grooves are at exact right angles to each other.

 

[IRstuff]

Are you kidding? They're offset vertically. Fundamental rule of any decent gimbal is that axes of rotation intersect, which yours clearly do not.

TTFN

FAQ731-376
Chinese prisoner wins Nobel Peace Prize

 

[ginsoakedboy]

I see. I didn't know that.

Ok, so the axes of rotation for both the curved surfaces need to be co-planar.

Also, radius of curvature of these surfaces is much greater than the "height" of the rings in my solid model.

Is there any requirement that the rotational axes need to lie close to or roughly within the plane of the gimbal rings themselves?

 

[MikeHalloran]

Get a few bars of soap and carve a model of it, and tell us how much water you need to apply to make it work.


No, I'm not kidding.


Mike Halloran
Pembroke Pines, FL, USA

 

[ginsoakedboy]

Vw (volume of water) will be proportional to the torque? Brilliant!