Best-Fit of angled Cylinders.

[[Mi...]]

I'm not 100% clear on what's happening with best fit of bore pattern so maybe that's the cause of my question.

But, is there any reason I can't simply measure these angled cylinders and fill out the best fit of bore pattern like usual? Calypso let's me without any warnings. AMSE not ISO if that makes any difference here.

I'm being told it likely won't work, but it'll be awhile before I actually have the part to test.

Edit: These cylinders are evenly spaced around a hub/shaft

angled.png

[[To...]]

Please sign in to view this quote.

Many years ago, I had a similar application. It involved creating a Geometry Best Fit Alignment of the cylinders but beyond that, I won't be of much help. Sorry.

Just thinking out loud here. Instead of using the cylinders in the BFBP, maybe use intersections of cylinders and top face. That might give you results that give you a better picture of where you are at.

[[Ro...]]

Hi Michael,
The main problem I see is that since the holes are not all in the same axis, you need to create an alignment for each hole that is rotated to look down the axis of the hole. BFBP will best fit them, but the results will be out because it's viewing them all from one direction, and the length of the axis changes location when looking at them like that.

As Tom noted, you can project the cylinder to a point, but your reducing an axis to a single element. I would normally create a best fit of the geometry, allowing for any unconstrained DOF to best fit, and them report them individually to that alignment, but using the Special button to align to the basic dimensions that define their location. You'll end up with one alignment per hole, but they will be best fit as a group, per the drawing.

[[Mi...]]

Please sign in to view this quote.

This sounds like what needs done...

You're talking about using "Geometry Best Fit" under Resources>Utilities?

Do you have any screenshots of this by chance?

As I read that a few times I think I understand, but screenshots always help.

[[Ro...]]

I would do it with the BF Bore Pattern and use that alignment for the holes. BFBP will constrain each datum in order of precedence to the previous one, based on the order you put in the datums. That's what you want. It can be done with Geometry Best Fit, but that creates an alignment based on a previous alignment and doesn't constrain based on datum precedence. It best fits them all at once. You could build a series of alignments, but BFBP does it all for you.

Then you would need to copy each cylinder you used for the BF pattern to report the position. Create a Position characteristic for each hole, Rotate using the Special button in the Position characteristic to the nominal angle of the hole to the centerline. You could/will have 2 Special entries, 1 for the rotation around the hub axis to align to the hole, depending on your starting alignment orientation when you created the BFBP, and 1 for rotation of the angle relative to the hub axis centerline.

Select the BF alignment instead of individual datums, and away you go. Rinse and repeat for each hole.

[[Mi...]]

Please sign in to view this quote.

Thank you very much! Pretty sure I got it...

One last question, on the individual positions do I need to attempt to offset the coordinate system so the Z actually looks directly down the hole? I'm pretty sure the X nominal takes care of it but just wanted to clarify.

This seems to be giving the right answer, I created it from the saved data on a program with similar holes, but this program uses a number of secondary alignments to do the rotating. So I was able to directly compare. This overall gave a slightly better result.

Position.PNG

[[Ro...]]

Please sign in to view this quote.

That's a judgment call, IMO. I would report it based on the BF alignment origin, because that honors the datum precedence. Since the holes aren't parallel to the hub axis, the distance from the axis changes based on where you measure it from.

That being said, some customers want to see a specific number, even if it's not the actual distance from the datums themselves. As long as you start with the correct alignment, you can move/rotate Basic distances to produce whatever origin they want. It will still geometrically be in the same place, just with a different reference point.