When I was trained in GD&T, the instructor said to think about how you would put the piece into a fixture for inspection.
Set the part down on the table, using 3 points of contact. This becomes your primary datum.
Slide the part to 2 points of contact on a side. This becomes the secondary datum.
Push the part to a singlle point of contact to keep it in place. This is your teriary datum.
Simple, effective and constrained.
You want to do as many inspections as you can from a single set of datum callouts as each addittional set requires a new inspection setup and/or fixture.
"Wildfires are dangerous, hard to control, and economically catastrophic."
Even better: think about how the part is used. Make datums relevant to most important functional aspects of part. If you do not know what is important to the part, you have bigger problems than getting your drawings finished.
No, basically, you let the nature of how the part actually functions in the assembly lead you to the datum framework! You are only deciding if you are the designer of the assembly!
Frank
I agree, Frank! But I was just riffing off of Cowski, who noticed your choice of words (divining, rather than defining). I just thought it made for a funny joke. (Bad engineering humor, I suppose.)
John-Paul Belanger
Certified Sr. GD&T Professional
Geometric Learning Systems
I agree, I could not think of a better word. I am saying the nature of how it mounts in the assembly leads us to our framework, as the designer we do need to use this to design solid assembly conditions, but I got the strong impression this is just someone looking at a part after it is designed, saying how should I put datums on the part. The old way was to let processing determine datums.
Frank
