attaching machined part to welded frame to yield precise motion 1

[learningallthetime]

Hello.

I'm new to this forum and come with a question.

There is a welded steel frame structure to the top of which is attached a three wooden sheets (not touching) adjusted to lie in a single plane. The frame has 4 "feet" that can be adjusted to ensure that the plane is horizontal.

There is an aluminium machined part (containing bearings & an optical encoder) from the top of which emerges a vertical shaft to which a horizontal jointed arm (circa 500mm long) is attached. The arm is dimensioned to support 10kg at the end with minimal flex (circa 1mm).

The arm should swing out over the aforementioned horizontal plane (think of it like a table top), as close as possible but without ever touching (am aiming for a conservative 5mm separation).

My question is, how to attach the aluminium machined part to the steel frame during the assembly process such that the motion of the arm is properly adjusted? OR such that it is in the future adjustable?

A couple of thoughts so far:

- initially I thought that I'd create some kind of "adjustment mechanism" so that the tilt of the part could be adjusted post assembly. Was unable to come up with anything that was not tremendously complicated (and chunky, given the mass at the tip of the arm).

- then decided that the way to go was to create some kind of jig to hold the part in perfect alignment during assembly at which point to fix it right once and for all.

- however even then, not clear how best to make the attachment since predrilled holes would be no good. (Considered bolting an aluminium plate to a steel one, welding one to the part and one to the frame with everything clamped... Bit of a shot in the dark.)

The idea is to come up with a method that can be done routinely.

As an aside I don't know the extent to which welded steel frames might distort over time, shifting the plane and necessitating some further adjustment of the arm? Is that just a matter of dimensioning the frame such that no loads (or knocks) could possibly take it near yield strength?

Any input much appreciated.

 

[BUGGAR]

Here I go again. Have you looked at automotive hub unit bearings? Usually, one flange is tapped and you could use adjusting screws in these holes to level the unit. Pick one with three flange holes and you can level it like a theodolite. They come with preloaded bearings. Tell Jimmie Johnson that Bob sent you.

 

[learningallthetime]

Excellent @BUGGAR

That's two good suggestions.

- Seems that automotive-hub-bearings handle an impressive range of loads. Great one stop solution. On the downside they are quite chunky (outer diameter 13cm upwards?). (By contrast could choose a 20mm rod and find bearings with an OD not much greater.) Does anyone know of some really small hub bearings?

- As far as I can tell, theodolite-type levelling involves 3 adjustment screws with ball joints at one end. Love the mechanism! Whereas a theodolite is basically in balance, here the screws will be taking a load and indeed the entire load (50Nm...) might end up more or less concentrated on one screw. So that would mean chunky bolts with chunky (custom?) ball joints. Hmmm.

 

[3DDave]

Nothing like a picture to clarify the needs of the design.

 

[learningallthetime]

@3DDAVE here's a sketch. The system supporting the arm should be as slim as possible (circa 50mm) in the horizontal plane, though it can be quite tall. It's mounted on a vertical bar that's presently 20mm wide and 40mm deep (some flexibility there).

http://i.imgur.com/bO9X5Sq.jpg

@BUGGAR axle hub bearings are perhaps "double direction tapered roller thrust bearings" which can indeed be sourced in more modest dimensions? Good call. The central problem remains, however either the aligned attachment of the part or some kind of levelling mechanism. I'm guessing a theodolite system would want 3 circa M10 bolts with ball joints on the ends which is definitely a bit chunky for a more-or-less one-off adjustment. The good news is they could be short since only a small adjustment should be required.

 

[GregLocock]

You haven't said how wide the base is. Shims are a no nonsense solution,often used where reliability is crucial,loads are high and adjustment is infrequent.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[learningallthetime]

@GregLockock The base is 1m wide and 0.6m deep. The structure 0.9 high.

How would you use shims to adjust? The issue would be attaching the part/adjusting such that the arm is horizontal in all directions.

 

[SwinnyGG]

You would use a simple bolted flange with shims to adjust the arm parallel to your reference surface. This will be by far the most robust, reliable, and cheapest solution.

 

[learningallthetime]

Excellent. Thanks!

In order to allow for a certain angle of tilt adjustment:

a) would 3 bolts in the flange be the optimal number (since that defines a plane, as with the theodolite with its levelling screw system)?

b) presumably the holes must be drilled loose to allow for the bolts to sit at an angle. Is that simple trigonometry based on the max angle and the thickness of the plate or is there a special best practice?

 

[Tmoose]

My builder's level has 4 leveling screws (with rounded tips) so leveling occurs while sighting in 2 distinct and separate steps in each of 2 orthogonal directions. That is much easier to accomplish than with 3 screws, where all 3 are kind of linked at times. Looks like MAybe David White instruments all used to use 4, not 3, despite the stability "built in" to 3 point support.

If using wheel hubs there are tapered shims available to correct rotor runout that might be handy truing up your scanner arm.

https://www.nucapperformance.com/brake-align.php

 

[learningallthetime]

@Tmoose that is great info.

This forum is excellent!

So 4 bolts facilitates adjustment....

Am I right in thinking that a bolt/shim arrangement is likely to be stronger than a levelling screw with ball joint arrangement of the same bolt size? Because whereas each levelling screw is like a solitary stalk, each bolt/shim is in compression and moment loads will be somewhat distributed across them?

By the same reasoning 4 bolts should also reduce stresses.

Watched the brake hub shim installation video and that answers a lot of questions about how to design the shims. Putting two and two together (so to speak!) it is beginning to look like 4 bolts arranged orthogonally combined with shims machined from an aluminium sheet, tapered in one direction, that can stacked and arranged orthogonally as necessary, might actually do the job. And way better than the horseshoe shims I'd envisaged.

Since the lateral alignment of the part is not critical, I guess I can just make the bolt holes loose in anticipation of tilting alignments. Perhaps its better that the holes of both plates be loose-fitting so that pressure concentrations can be shared between them?

 

[3DDave]

Doesn't look like the sketch made it.

 

[learningallthetime]

Sorry @3DDave, tried again to no avail so have added external link (

http://i.imgur.com/bO9X5Sq.jpg

).