Runout measurement when datums are hidden by fixture

[romulus102]

I have a part which has to be machined in a fixture. The repair document requires measurement of a runout w.r.t two datums. When the part is placed and held down in the fixture, both the datums locations are hidden by the fixture. In such a scenario, do i just put a DTI on the surface which requires runout to get a measurement, or do i need to clock the datums locations and make sure they are lowest as possible before i proceed to measure the runout. Please refer to the attached image for problem illustration. Any guidance on this matter will be much appreciated.

 

[Belanger]

I'm not sure what you mean by clocking the datum locations to make sure they are as low as possible. The two datums just mean that you fixture against those two features so that they sit evenly on their highest points (I presume B is a diametric feature).
In other words, there's no measurement to do on datums A or B. The only issue may be if a datum is so irregular that the part rocks. This gets into something called multiple candidate datums, but let me know if I'm way off from your question.

John-Paul Belanger
Certified Sr. GD&T Professional
Geometric Learning Systems

 

[romulus102]

Thanks for responding to my question. Yes, B is a diametric feature. I believe your reply answers part of my question i.e. that we don't need to do any other measurements on the datums. Am i correct to state that, once the part is fixtured, it is deemed to sit on the highest points of the relevant datums and then all i have to do take FIR of the feature whose runout needs to be determined.

 

[pmarc]

Yes, you are correct.

 

[pylfrm]

romulus102,

I'd say your question is not answerable without more information about the fixture and the connection to the dial test indicator which presumably allows for rotation. Does the fixture actually clamp on datum feature B and center it, or does it just have a clearance hole? Is the "runout" of the fixture itself known?


pylfrm

 

[mkcski]

romulus102:

This is always a dilemma: how accurate must the simulated datums be. I would secure (chuck in a lathe) your "adapter" plate to the machine tool and machine the datums A & B and not disturb them before mounting the part to it. The tolerance on the "perfect" simulated datums is equal to the precision and accuracy of the machined tool. The best you can do. If the inaccuracy of the simulated datums is a equal to a "large" percentage of the specified feature tolerance in the FCF, you may reject good parts or accept bad parts.

Another item: to be strictly per the "spec", the simulated datum B cylinder must be the "smallest circumscribed cylinder". This means it must not be fixed at MMC but must vary within the size limits of the datum feature, i.e. not datum shift is allowed.

Certified Sr. GD&T Professional