Need ideas for a fixture with expanding diameter

[geesamand]

Hello,

I'm designing an inspection fixture that must fit into two pilots. Each pilot is about 8" dia x 0.75 thick with a diameter tolerance of +/- .0015". I need to easily mount a bearing into each pilot while maintaining center to the pilot +/- .0001.

I don't have room for an expanding chuck (too large and heavy) and expanding collets too.

The best idea I have is a ring that is just a few thou under size of the pilot, with single split running from the center to the OD. Then a tapered pipe plug or taper pin centered on the split to make it expand. Then mount the bearing in the center with reliefs around it so the expansion is distributed over the diameter. Only thing is I'm not sure how well this concept will hold center.

I'm interested in any comments about the direction I'm started on, or other suggestions.

Thanks,

David

 

[MintJulep]

How about something like a keyless hub?

http://www.fennerdrives.com/trantorque/

 

[MikeHalloran]

It's most important that the ID and OD of the spacer be machined in the exact same setup, so concentricity is likely. Then a single radial split with some sort of wedge inserted should work.

Consider aluminum or Delrin for the spacer material.

A pipe plug is not a bad idea, but be gentle; it gives you a lot of mechanical advantage.



Mike Halloran
Pembroke Pines, FL, USA

 

[geesamand]

I've used Trantorque before, and they work well when the dimensions agree. Unfortunately in this case they are too small for the 8" diameter. I dismissed the larger locking ring options since they have so many bolts and are designed for such high loads.

For the split spacer you're right aluminum or delrin could work, but I was considering a 1/2" steel plate. It now occurs to me that I can test the effectiveness before I invest in the whole fixture by buying just the spacer plate first, then CMM it with the pin/plug fitted and confirm concentricity. I figured on a wire EDM to make the spacer plate so that perfect concentricity is assured.

Thanks,

David

 

[ajack1]

Is anything with splits, tapers and multiple pieces really going to hold a tolerance of +/- 0.0001?

 

[geesamand]

I'm open to other options. I'm not a fixture designer so what else is there?

David

 

[MikeHalloran]

I was thinking a lathe and a saw, like machinists have used to make 'pot chucks' for millenia.

A steel spacer with a EDM surface may scratch the bearing races.
Steel is also going to be stiff, so you may need extra sawcuts, depending on how much adjustment range you need.





Mike Halloran
Pembroke Pines, FL, USA

 

[BillPSU]

There are several suppliers of hydraulic expansion mandrels which would probably work for this application. The expanding mandrel expands through hydraulic pressure on the inside of the expansion sleeve. They can be activated by a hand tightened screw or by power. I've used this type of mandrel to locate on the honed ID of a cylinder barrel to generate a steady rest turn OD on the outside of the barrel. A couple of companies are Rohm and Koenig-mtm. There are probably others.

Bill

 

[geesamand]

I'm thinking I could install an o-ring on the outer diameter of the male pilot, so that the squeeze of the o-ring will provide a centering force. Then we tighten the bolts to the flange to prevent it from shifting.

Has anyone tried this kind of thing? Did it work for you?

Thanks,

David

 

[MikeHalloran]

An o-ring is not going to be stiff enough, or uniform enough, to meet your stated spec. The tolerance on every dimension of an o-ring is around +/-.003". They aren't all that bad, but they aren't within 'tenths' either. Additionally, the ID and OD both may have flash ridges.

You _can_ use an o-ring as a sort of leakproof hydraulic fluid, by trapping it in a cavity with no free space and then moving one of the cavity boundaries; they don't compress easily.

Have you got a machinist in your shop?
Ask him to show you how to make a 'pot chuck'.



Mike Halloran
Pembroke Pines, FL, USA

 

[Benedictj1]

Are you trying to measure the coaxiality of two cylinders?

 

[geesamand]

The coaxiality of 3 short cylinders, in fact.

The intuitive approach is to replicate the machine, and build a fixture that rotates within two of the cylinders. For the reasons discussed above, it's hard. I spent a lot of time thinking within this paradigm.

I'm not going toward an approach where the fixture is stationary and the part will rotate. The two cylinders that form the datum centerline can be expanding, and if you set the clearance to allow a tiny slip fit, we think the part should rotate consistently around that axis to allow inspection of the 3rd cylinder. The inherent issues with this method - namely - wear on the part, wear on the fixture, friction due to part weight can all be managed. I'm still debating whether it's necessary for the fitted diameters to expand - if the user keeps the component pushed firmly toward the indicator, the clearance has little effect.