How to simulate center plane datum from width datum feature on CMM?

[bxbzq]

I think this is a very simple question but I don't find the answers.
I understand for slots, one may use adjustable parallels to fit into the slot to serve as datum simulator. But for outside width feature, what are the tools?

 

[Belanger]

The datum feature can be simulated by two parallel surfaces that close down equally, similar to a vise-type thing.
Or, if the CMM is scanning enough points on the outside surfaces, you might not need a physical tool to simulate -- the CMM could calculate the center of the actual mating envelope (see Fig. 4-13 of the Y14.5 standard).

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

 

[SeasonLee]

Self-centering vise is a good choice for the external FOS, you may easily find out the datum center plane by the opposite threads spindle design (right hand and left hand threaded spindle), it will always makes a centric clamping, there are several brands on the market, following is the one of them for your reference only.

http://www.wiltontools.com/us/en/p/super-precision-self-centering-vise-4-jaw-width-1-1-2-depth/11713

Season

 

[bxbzq]

Thanks for the info.

Belanger,
I'm not from inspection area, but I learned the proper way to establish datums is to hit on the simulator surfaces, not on the part surfaces. Do you agree?

 

[Belanger]

Yes, I agree. The datum is derived from the datum feature simulator.
I just meant that sometimes in a pinch you could derive a datum from the surface itself by doing the mathematical analysis (if enough points are sampled on the surface) to find the tangent plane or envelope.

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

 

[Tarator]

Belanger,

I do have a question... With a hard gage, let's say, with a self-centering vise, the part would sit on its "high" points inside the vise... but on CMM, how do you guarantee you "scan" the high points? or can the software take care of that?

 

[3DDave]

It depends on the material condition for the feature and where it sits in the order of precedence of a referring callout as to what needs to be done to simulate it.

In many cases it isn't required to find the exactly perfect high points, only to be certain that the finding is within the bounds of accuracy required to be sure the check is a good one.

For example, on a simple (as anything can be simple) length measurement: If the tolerance is +/-1.00 unit and the accuracy is only .25 units on the measuring equipment, but the measure is .25 from nominal, then it is likely the part is acceptable and lies somewhere between 0.00 and .50 of nominal. If a different copy of the part is at .999 and the measure with the .25 accuracy is .99, then one can't be sure, so a more accurate method is required for that copy.

The theoretical gauge is perfect, but practical applications are always a trade-off between cost, time, and likelihood of accuracy.

 

[Belanger]

It certainly depends on how many points the CMM scans across the surface -- that's the downside of using a CMM without any simulating equipment. So as mentioned earlier, we should really use two parallel plates that close down (or expand) equally until making contact. The CMM should them probe those plates; this better ensures that we are getting the highest points.

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

 

[Tarator]

Thank you all! I appreciate it!