Diameters and roundness....

[[Li...]]

So are a machine shop with various machining equipment. I ran into a situation where this roughly 8" diameter by .500" thick plate had a center bore machined to a size of 2.13975±.00025" (2.1395" - 2.1400") During our prove out of the operation I discovered that the size I got was varying a lot. Evaluation method of LSQ/Maximum Inscribed/Minimum Circumscribed was all over the place and would show out of spec conditions (oversized). This led me to ask Calypso what the roundness of this bore was. As I expected the result was high. Not sure what the tolerance is supposed to be but if the dwg only give a tolerance of ±.00025" I set the tolerance for roundness at .0005". The M.E. that was in charge of creating the program to machine the parts was not overly concerned about roundness because it is not on the dwg., but I tried to tell him that this will affect his size. Me and the operator running the machine worked together to get similar results with some Mitutoyo tri-mics he was using (we did not have gage pins-GO/NO-GO at the time of making these part) and CMM results. Once we had our adjustments aligned, we were within a couple tenths of each other, tri-mics and CMM.

Moving forward to yesterday and late last week we got these parts back from outside processing. Etching and chemical coating. One of the parts was used for FAI inspection and it was found to have this bore oversized. We are not sure what has caused this. The operator and I did a lot of trial and error to get comparable results between gaging methods and we ran every single part on the CMM. We now have a GO/NO-GO gage for this bore and the NO-GO falls in.

I was running some of these over the CMM yesterday and by changing the evaluation method I can get these bores in spec, close to the high limit of 2.1400" but yet the NO-GO goes.

So, the results will PASS the parts, but hard gages will FAIL the part.

Someone talk to me about evaluation methods. After experiencing this I don't want to ever use anything other than LSQ.

[[Ma...]]

Something to consider - 2point diameter allows you to use (E) - envelope - look at it.

[[Ow...]]

I won’t dive into the evaluation techniques, but to help others understand the problem, I have a few quick, vague and suggestive questions.

(1) Is the plate with the hole in it machined/ground flat? If it is not flat, are you securing the part on the CMM with some type of clamps to keep it from moving or shimming to where it won’t wobble? 3 pointing would be the preferred method to ensure the part is not distorted or moving while measured.

(2) Temperature, the Coefficient of Thermal Expansion can be tricky to configure and is almost never exact. However, just 5 degrees fahrenheit difference on steel material, when and where measured is more than likely going to change good to bad and bad to good with that small of tolerance.

(3) is the probe proven to have no wear spots on it and have you checked it with two or more different probes to see if the result changes?

[[Ja...]]

I would tell the M.E. about ASME rule #1, The Envelope Principal. The form of an individual feature of size is controlled by its limits of size unless there is an additional form tolerance.

[[Ke...]]

Please sign in to view this quote.

Outer Tangential for "fitting condition" (like go/nogo)
LSQ to study repeatability of the mfg process.

OT for ID = Max Inscribed
OT for OD = Min Circumscribed

[[Mi...]]

In a situation like that where the person machining the part refuses to listen, I would simply report the effective size (Inscribed or Circumscribed depending on ID or OD) and make them figure it out.

Or ask the engineer to put an internal Circularity/Cylindricity callout on the control plan, since the person running our machines usually wouldn't understand something like ASME Rule #1 so for our process we would tighten the size tolerance and add an internal form control.

[[Da...]]

We run close tolerance bores all the time, and roundness is very important. We will check roundness on parts that don't require a roundness evaluation. Set the tolerance for roundness at 1/2 the tolerance for size and set to outer tangential. Check your filter and outlier settings. And look at the roundness plot to see what is causing the OOR condition. A small bit of dirt in a bore that is using OT will result in a size change. You can SEE that on the roundness plot! You can also see 3 point out of round caused by chucking too tightly.