Measuring threads

[[Cl...]]

My part is a cylindrical donut. It has a threaded ID (10-56 UNS-2B/.173 minor Ø). Datum -A- is the ID concentric to the threaded ID.

Datum -A- is a Ø .191 cylinder that is only .025" in depth. The print calls out .001 runout of the threaded cylinder back to -A-.

What in your opinion would be the best strategy for this?

[[Ri...]]

I would base the Ø.191 on tooling for its cylindricity and axiality, it's too shallow to get a reliable axis.

A helical scan of the threads, minimum 3 helix and a circle for the Ø.191 hole.

[[Cl...]]

I'll start there. Thanks!

[[Lu...]]

I often reference this video when scanning threads, helps me out a lot.

 

Goes over how to find the gradient/slope of the threads so you scan at a consistent point during the helix path

 

 

Edited September 16

[[Cl...]]

Thanks Luke. What about ruby size? I was thinking something in the range of 1 to 1.5mm Ø?

[[Ri...]]

With a Ø.173 and Ø.191, I would use either a Ø1.5mm or Ø2mm stylus and set retract to about 0.03937.

Edited September 17

[[Ch...]]

Please sign in to view this username.

Here is a very good Lunch-N-Learn about measuring threaded bores from 2011.  It covers several different strategies and compares accuracy vs time.  

Threaded_Holes_metrology_conf_2011.pdf

[[Cl...]]

Second question about this part. I'm evaluating the threaded ID for perpendicularity and runout both back to -A-.

As previously mentioned, Datum -A- has a depth of only .025". I'm apprehensive about using such a shallow cylinder

as the datum. I also don't feel comfortable using a circle. If I do use the circle, can I constrain it (not sure if that would

be the Z axis) in the characteristic. Datum -A- as a cylinder gives me a result of .035" for both, which I know is wrong.

Edited September 17

[[Wo...]]

I've had a similar situation some time ago. I think we've contacted the customer and asked if we can use a physical gage to check concentricity. Just a pin gage with a thread at the end. Even using helix strategies the results were not consistent due to short datum.

[[Jo...]]

You can capture the feature as a cylinder and recall it to a circle for more consistent results.  Its not the compromise it first appears!

 Any method that doesn't use thread plugs ($$$) is really locating just the minor diameter of the thread and almost nobody cares about that.  But repeatability is difficult with thread plugs. Make sure you know what they are "really" looking for before spending too much time going down the rabbit hole.  

[[Cl...]]

I added some of the suggested strategies to my program and ran a type-1 (10 replicants on/off the fixture each run)

For the perpendicularity, the best result was 28% (capability 0.71). For the runout 41% (capability 0.49)

Edited September 18

[[Za...]]

Please sign in to view this quote.

for short threaded bores, this article may provide some further insight into how to program
https://portal.zeiss.com/knowledge-base?id=599380