Composite Profile

[Scott Brown]

Hello Everyone

I have a question regarding composite profile. I have attached a rough sketch for reference. Can someone please explain how to properly dimension the lower segment of this composite profile. I am using legacy dimensioning so I would like to see the correct method based on legacy. This is a stamped part with 4 arcs (trim edges) that need to be controlled to the main datums and also as a feature set. A pipe will weld to these 4 edges. I have added edge points to all 4 arcs and have dimensioned the T values of each point relative to ABC. For the lower segment I have created a 2D best fit alignment in the Y+ workplane allowing translation along the X and Z axis. Under this alignment I dimension the edge points to check the lower segment. Should this be a 3D best fit?? Please ignore the datum modifiers for this discussion.My main focus is creating the correct best fit alignment to check the lower segment.


Thanks
Scott

[Rich P]

I think there drawing is looking to constrain the form error of each of the 4 places to a max deviation of 0.5 individually
Dimension each scan area to profile form only 0.5

[VinniUSMC]

Yes, just dimension each of the 4 locations as "Profile Form Only" to 0.5. It's not exactly perfect, but it's more than good enough.

[Scott Brown]

Doesn't the comment "4 places" indicate a pattern requirement. (all 4 arcs relative to each other) instead of individually??

Thanks
Scott

[Schlag]

I would agree. I would make a "set" of the 4 arcs / places and dimension it that way. Might be overkill but your are for sure covered that way. Also, by using legacy and not EXACT, you can not taking advantage of the the MMB. That's fine as you realize its possible to reject "good" parts by not using this bonus. The good news is that you will never accept bad one by not using the bonus.

[Rich P]

The upper feature control frame controls the pattern location.
a +-0.5 window or envelope is created at each of the 4 places from datum orientation.
Each section individually or joined together is constrained by the datum structure. report each surface individually for clarity in my opinion but a constructed feature set would create the same result from a pass fail perspective. Each section has a simultaneous requirement by sharing the exact same FCF by definition.

The lower portion each section creates a 0.5 max boundary of deviation envelope so the data would be fitted and is a refinement of upper FCF, that if you report each as a group you are limiting the fitting to a total 0.5 error of the 4 places at which point then FCF with datum should be 0.5 not 1.0