Cylindricity of .0003"

[[Li...]]

I was asked today if our CMM(s) can accurately measure a cylindricity callout of .0003"
I don't see why we couldn't measure this and have reliable data.
The length of the cylinder is .135" and the size is Ø .1562 - .1587
As long as all the points taken fall within a zone of .0003" no matter the diameter. This is form not size.

What am I missing on their (Engineering and CNC programmer) concerns?

To me, this is exactly what CMM are for.

Please let me know your thoughts and concerns. How do I test this to prove we can? The SQE talked about doing a Gage R&R
We have some air gage masters (ring gages) that I think would be a good artifact to measure.

Cylindicitry of.0003.png

[[Ro...]]

If i read my wall poster wall right, Cylindricity of .0003 means the form must fall within 2 concentric cylinders with Radial separation of .0003, so its similar to a .0006 profile.
I don't see a problem with that.
I would say that a hole that small and only 1/8" deep should be pretty easy.

[[To...]]

Please sign in to view this quote.

wouldn't .0003 cylindricity actually be a profile of .0003 (or +/-.00015)

[[Ja...]]

Depends on your machine and whether or not you're try to achieve a TAR of 4:1 or better. A Prismo wont do 4:1, I'm pretty sure. An Ultra should but you might need guard banding. But I'm far from an expert, so take all that with a grain of salt.

[[Cl...]]

https://fractory.com/cylindricity-gdt-explained/

[[Li...]]

So I grabbed a ring gage master out of our Gage Crib and ran it for Cylindricity. It is over .0003" but in spec for its diameter intention. I used helix and spiraled 10 times. I expected to see better results from this gage. I am going to try a different one.

[[Ke...]]

If you have the documentation for your CMM (usually in a black binder), it should list the measured and allowable form error for that specific CMM, I believe as the parameter RONt, which will include the settings used for that test, and may give you a better idea of your CMMs capabilities and settings to use.

[[Ro...]]

Please sign in to view this quote.

The words "radial separation" is what makes me think that the tolerance would effectively double in a profile.
.0003 per side?
I'm probably wrong......

fdgdfsgdsfgdsfgs.PNG

[[Ow...]]

Please sign in to view this quote.

Seems like everybody has different views on helix scans but, try using several circles and lines and see what you get.
In my experience, helix scans are prone to non-stable vector issues and the ability to filter out the beginning and end of the scan is challenged.

[[Th...]]

Please sign in to view this quote.

Cylindricity tolerance describes the width of the zone like other ASME form tolerancing, so it would be similar to profile in interpreting the plus/minus allowance off true. I think the use of "radial separation" could be misleading, and ASME uses it as well in Y14.45, so I am relying on the drawings in 14.5 here.

Please sign in to view this quote.

Yeah, I think the helix path strategy trades robustness for time savings.

[[Li...]]

I was a little shocked that a "Master" ring gage did not give better results for cylindricity. That being said the diameters are spot on.
The perpendicularity is a little over a .0001" I'm thinking of changing the helix scan to 8-10 lines or several circle scans within the .135" cylinder it created. Another troubling equation here is that I have ran this a couple times and used different probes and get different results. At one of the trials cylindricity was in spec(under .0003") and another time it slide up to .001"

[[Je...]]

Please sign in to view this quote.

.

Lincoln,

Yes, CMMs can verify a cylindricity callout of .0003", but it depends greatly on the hardware and measurement strategy. While a Prismo, Micura, or other higher-end CMM is preferred, a Contura can do it if equipped with a VAST sensor. I think it would be difficult to justify an RDS/XXT's capability for this task.

Also, are you measuring for a short run such as tooling verification, or will this be continuous production? If the volume of measurements is low, you can afford to slow down the measurement strategies and add more scanning paths.

For cylindricity, I would definitely recommend circle paths as opposed to a helix. I typically take a minimum of three circle paths, and for a depth of .135 and a tight tolerance, I would probably increase to five paths. It's nice to be able to visually depict the form of a cylinder at various depths, and this can be separated more easily with circle paths as opposed to a helix.

Once again, if you're not concerned with cycle time, you can afford to scan slower with greater point density, and both of these should be set more conservatively than the Zeiss cookbook suggests. Also, set the probing dynamic to 50% or less to make sure that deflection is minimized.

I'm interested to hear how this turns out for you. Thanks for sharing about it.

.

[[Ni...]]

Please sign in to view this quote.

This. Setting the probing behavior of passive sensors from standard to sensitive and setting the probing dynamic to 50% or 25% is going to shoot the accuracy way up.

[[Wo...]]

This looks like a typical "Lee Plug" requirement.
I get tons of these, and always measure them as cylinders with 3 circle scans - top middle bottom.
You can then clearly see if the counterbore is not tapered etc.
Or even better - measure 3 circles and then recall feature points into a cylinder.
I've never had luck using helix strategy.

A machine with VAST/ VAST XTR should measure that easily.
I wouldn't trust an RDS to check it.

[[Tr...]]

Since I started, I've always wanted to use the helix path because it's cool and whatnot. 🤠 But, I've found that circles or lines give better results. If you want to collect more data, use more circles/lines (as many as you want and/or have time for). Use the biggest probe you can. And, stay clear of the chamfers on the ends of the cylinder. No way a ring gage should have 0.0003" cylindricity. I looked at some of my calibration history and they almost always are less than 0.0001"