Repeatability

[[Re...]]

I have a 65mm diameter plug that I have mounted on a RT1 rotary table that I am measuring with an Accura cmm with a Vast XT head.
1. Measurement plan measures a circle diameter on the plug, 2 times - one with RT and the other without RT.
2. Same plan ran and re-ran without unchucking plug.
3. Same measurement strategy on both circles other than the RT.

Zeiss claimed that we had too much airflow in the lab going across the RT even though we were within the temperature specifications. We had work done on our HVAC system to where the airflow across the RT is minimal and the temperature is now controlled much tighter. However I'm not satisfied with the measurement results with the RT. All of the studies that I have done show that the measurements are inversely correlated with the temperature when using the RT.
My latest study shows a range of 0.0008mm with a temperature delta of 0.27°C without the RT.
A rage of 0.0036mm with a temperature delta of 0.27°C with the RT. - This seems unexceptionable to me.
This study of 15 runs took 1hr 35min with the HVAC cycling 3 1/2 times.

When the airflow was not as controlled as it is now, we were seeing a range of 0.0088mm with a temperature delta of 0.56°C with the RT.

Can anyone tell me what this range should be with the data provided.
The attachments show the study, allowable linear for cmm, and RT info.

RT study.JPGAccura.JPGRT1.JPG

[[Re...]]

FYI - I'm running Calypso 2018, ServicePack 3 Version 6.6.1606

[[Re...]]

This cmm and rotary table is due for calibration in a month. Is this repeatability something that would be checked for during calibration?

[[Ri...]]

How are you performing the RT Axis calibration?

[[Re...]]

I'm calibrating the rotary table with the truing bar method. 1.187 Dia x 12" lg

[[Ow...]]

Fortunately, I never need to look at the 4th decimal place metric.
However, size aside, I have noticed form differences (roundness) when comparing RT vs no RT and when it's important, use other methods to verify the characteristic.
It would be interesting to see if the variation follows variation in form.

[[Re...]]

Unfortunately I have to measure to +/- 0.005 mm on a good number of parts. This seems to be possible with out the use of the rotary table. It requires calibration often. With the rotary table I struggle to measure to this tolerance even when calibrating often. The measurements with the rotary table become unreliable as soon as there is a small change in temperature. When I say small change in temperature, I'm talking about 0.1° C. Without the rotary table I feel confident in my measurements up to a 1° C delta. I just need to know how much of a difference in measurement that there should be between using the rotary table verses not using the rotary table. I would not think it should add 4.5 times the error with only a 0.27°C delta. If it should, then I need to convince management that we need to get a more accurate rotary table. But no one seem to be able to give me the tolerance that I should be able to measure to with an RT1.

[[Re...]]

A few months ago we had a Zeiss technician come in to analyze the situation. His conclusion was the we had too much temperature fluctuation and too much air flow across the rotary table. Our temperature fluctuations at the time were within the Zeiss specifications, but we did have air flow in the direction of the rotary table. We have since had our HVAC system improved, and now the temperature is controlled much tighter and there is no air flow in the direction of the rotary table. However the problem still exist. I just need to know what the cause of the problem is. Maybe these rotary tables are that sensitive to temperature changes. If so, then I do not understand how Zeiss could have sold our company such a device with such variance compared to the cmm itself.

[[Ri...]]

I'd be interested in your results if you did a Rotary Table axis alignment on the part. One that is performed each time the part is ran. It will increase the cycle time a little bit, but I feel like that is the root of your problem.

I've held form tolerances using the rotary of 0.01mm (full blow MSA/GR&R).

Most of my struggles over the years has always been with getting a good rotary table axis alignment. The suggestions that I've gotten from colleagues has been to perform the axis alignment using the part, and doing it every time.

This may not be your case.

[[Me...]]

Please sign in to view this quote.

^ This X 1,000,000

[[Re...]]

Richard,

I feel as though you are correct in regards to getting a good calibration on the rotary table. A good calibration is needed to get the correct actuals. But what I do not understand is the range that I am seeing from run to run in my studies. And how this range is directly related to temperature change. The Zeiss technician saw the same thing with his calibration methods. My studies have been back to back runs with out moving the part, and measuring the part with and without the use of the rotary table, and using the same strategy on both. I would think that even a bad calibration would not show the range that I'm getting. And as many time that I have recalibrated for these studies, I would have thought that at least one of them would have been a good calibration. At this point Zeiss will have to rectify the situation when they come in next month for our annual calibration. Their will be no sign-off by me on the rotary table until they can show me that the rotary table is somewhat repeatable. I just wish that I knew to what value it should be.

[[Kl...]]

I've done a lot of MSA for CMM's with rotary table located shopfloor or lab.

My practial experiance flows into the attached document.
It does not claim to be complete and 100% scientific, but all acceptances tests and complaints had been fulfilled with customer satisfaction.
Sometimes it was hard for the customer to accept the extra cost of better air conditioning or extending the measurement time through the integrated rotary table qualification.

I can only fully agree with Richard's remarks.

I hope the content is clear and understandable, additions and corrections are welcome.

RotaryTable_en.docx

[[Si...]]

Please sign in to view this quote.

There is a new SP v .20

[[Re...]]

Any advice as to what may next step should be. Retirement is currently not an option.

[[Ch...]]

Have you tried the RT alignment from the part yet? Then repeating study?

[[Kl...]]

Hey Reid,

make a test by RT-axis on part.

Put temperature sensors on "air" not on "part" to monitor the temperature around CMM during the repetitions.
PRISMO Ultra and ACCURA 2 with Option TFK are able to report each sensor serperatly. I prefer the Gaussian for circles
inside the test. Please send me the database, for deeper analysis.

I've put and example into attachment

Klaus
"Once Metrology, always Metrology" 🤣

RT_Axis_Part.zip

[[Re...]]

Chad, I have not yet tried the RT alignment from the part yet. However Klaus has an attached program in this thread that I'm hoping to find time to execute today, that does align to the part each run.

[[Ch...]]

I am also planning on trying Klaus's program today but it has already been busy so far.

[[Re...]]

Thanks Klaus for the data - Your program worked fine, however there was one characteristic that needed the tolerance adjusted. The study showed me that my environment is not controlled as well as I had thought. Hanging the temperature probes of from the rack exposed to the air gave far different results as to being attached to the rotary table. I believe that I may have an airflow issue affecting the temperature. I have a precise temperature device within 3 feet of the cmm that is showing the temperature to be within spec of 1° delta/hr. When the probes are attached to the rotary table they also monitor the temperature to be within spec. However when the temperature probes are hanging from the probe rack I'm getting a 2° C range for each cycle of the HVAC system, which cycles about 3.5 time per hour. Looks like the HVAC system airflow will need to be worked on once again.

[[Re...]]

Thanks Guys,

I do see better results when using the part to establish the rotary table alignment. It appears that I will need to change all of my programs that have critical callouts to this method even though I will be adding substantial time to the measurement. I assume that one would be able to call up a previous rotary table position and axis that was established by using the truing bar/cylinder method when running a plan that does not use itself "part" for alignment. I do not what to constantly setup to calibrate the RT between plans using the different methods. Also when I use the less accurate method with the 12 inch long cylinder verse the part which has a much shorter cylinder, I would like to know how short the cylinder axis can be and still be reliable? Some of my parts the cylinder can be as short as 8mm. Any pointers would be appreciated.

[[Ke...]]

I started off using the Part to establish the RT axis (since I was measuring 8"+ tubes), but now that I am using the RT to measuring 1" tall bearings, I switched to using the Tooling Ball method which works wonderfully for repeatably measuring raceway diameters.

Wondering if anyone has used the Tooling Ball method, instead of Part (where the parts are long enough to trust for the RT axis), and how the results compared.

[[Ri...]]

Please sign in to view this quote.

I originally did the tooling ball method (but used the reference sphere). I had success with it. As tolerances got tighter, I switch to the high/low tooling ball method which helped a lot. As the tolerances got even tighter, I started using the really tall cylinder, and the part method. That is where I found the most success.