Cylinder orientation Issue?

Thanks for all the insight everyone.  I appreciate it greatly.  I am still having a few issues however.  I have attached a partial of my print and the cylinder issue I'm still having.  

louisd I did your suggestion.  I made two circles in the YPlus plane.  They looked fine.  I then constructed a cylinder out of those two circles and the same thing happened from the first.

I attached another doc, this time with excerpts of the print I'm working off of.  I can reveal more print if need be.  

Datum A is the part underside, Datum B seems to be the hinge centerpoint and Datum C Is a line?  Maybe I have this aligned like garbage?  

I appreciate everyones help and insight.  ThanksDOCX

I'm guessing you are taking two 3-hit circles?  Instead of constructing two separate circles, and then constructing a cylinder from those two (which the demon will only use those circle centroids)...

Create a cylinder with two circle cuts of at least 4 hits per cut.  Adding the extra points, locks down the vector direction. 

The construct cylinder tool is a bit dumb as it doesn't utilize all the hit data of each circle.  it uses the centroids of the circle features.  Collapse-o (Calypso, competitor's software) has a "SHELL" button you can toggle, which forces the circles to be judged by using all hit data, instead of just the centroid.  It would be pretty cool if PCDMIS would someday have such a simple yet powerful option :D

If you are using circle features with the CIR1.HIT[1..8] etc, does that use all hit data or the centroid of the hits?

If you are using circle features with the CIR1.HIT[1..8] etc, does that use all hit data or the centroid of the hits?

In my experience that doesn't play nicely with me within the construct tools.  It wants each hit comma separated before it will work.  I figured that would be too advanced for this user as well.