Linear measurement of a curved part

[vin300]

Below is a part proposed to be manufactured. The inspection of this part poses a problem, because vernier callipers only measure distances between parallel surfaces.

I know there is some photographic method of measuring distances, however I would like to know is there no way of physically measuring the diameter along the curvature.

 

[Ranger Mike]

classic method would be an open set up using granite surface plate and a digital height stand to measure various points along the long axis. Coordinate measurement machine is the typical solution. google CMM.

 

[jrmichler]

If you just want to measure the diameter in the middle, you could use a point micrometer:

 

[Tom.G]

Build a jig to keep the product directly under a Dial Indicator .

 

[vin300]

Thanks. I'll have to enquire about the feasibility of these methods. I think using the micrometer would be cumbersome.

 

[Tolklein]

Are you manufacturing hundreds of these parts? And would they all need to be measured for accuracy? If so then then a Tom.G suggested, having some kind of jig built would be my suggestion, though I would use two laser distance sensors to give an accurate thickness, or in this case diameter, in this case I assume you would have to make multiple measurements along the edge. The cost of such a setup would only really be feasible if you are producing enough quantities to justify, also greater accuracy = greater price.

 

[vin300]

I'm being told the CMM isn't absolutely necessary, an optical profilometer that is used to measure roughness would do fine. I'm about to find out soon if it does.

Are you manufacturing hundreds of these parts? And would they all need to be measured for accuracy? If so then then a Tom.G suggested, having some kind of jig built would be my suggestion, though I would use two laser distance sensors to give an accurate thickness, or in this case diameter, in this case I assume you would have to make multiple measurements along the edge.

It is a large quantity but I'm only going to inspect some samples.
In the beginning I had mentioned physical measurement, but I'm seeing what works.

 

[Ranger Mike]

i have sold cmms since 1980 and was product manager for the original Profilometer. patented product.
what dimensions do you need to measure and to what accuracy?
it total measurement time a consideration?

 

[Baluncore]

It looks to me like a round bar is machined to a reduced diameter over a length near the centre, with a conical transition on either side.
I see no reason why a Go/NoGo profile template cannot be made. Exactly what shape it it?
What is the material, is it machined in a lathe or injection moulded? How about putting some dimensions on it.

 

[vin300]

Done. They were skeptical at first. Then one person used the profilometer. The bar was kept on a V block over the device table. The table could be controlled by a computer and joystick on two axes. By taking reference of the end face, diameter at any distance was measured by focusing the image.

It looks to me like a round bar is machined to a reduced diameter over a length near the centre, with a conical transition on either side.

It is not conical but circular.