Angular alignment of two concentric parts?

[DesignJunkie]

Hello,

I have an assembly that contains two concentric rings. Neither of these parts has any flat surfaces that I can use to constrain them to each other to achieve angular alignment. They do, however, contain matching bolt hole patterns. Here is what I have been trying. I've attached a screen capture for your viewiing pleasure. I start out by giving the base part a fix constraint. Then I bring in the second part and use the touch/align constraint using both part centerlines to make them coaxial. Next I apply a distance constraint so that the two ring faces are separated by 1/2". So far, so good. Finally, I try to apply another touch/align constraint using centerlines of matching bolt holes. This last step makes all of my constraints invalid. Does anybody have suggestions?

Thanks!

-DJ

 

[jerry1423]

You can change the ref set to "entire part" and constrain to the datum planes, and then change the ref set back to "model" (or whatever the ref set was), and you should be OK.

 

[DesignJunkie]

Thanks Jerry!

This method only works for me when I first bring the part in. Can you change the ref set after the fact?

 

[jerry1423]

Yes, I think so.
I haven't done that in a long time, but it was brought up in this forum not too long ago.

 

[JohnRBaker]

Yes, all you have to do is either select the Component from the screen or from within the Assembly Navigator, press MB3 and select the 'Replace Reference Set' option, and select 'Entire Part'. If you don't have any convenient Datum Planes (or a Datum CSYS located at the axis of the rings) already in your model, just add one to each part and go from there. After the Constraints have been added, repeat the above steps, only this time select the 'Model' Reference Set. Even though you will no longer see the Datums, they will still be used to constrain the Components in the Assembly.

John R. Baker, P.E.
Product 'Evangelist'
Product Design Solutions
Siemens PLM Software Inc.
Industry Sector
Cypress, CA

http://www.siemens.com/plm

http://www.plmworld.org/museum/

To an Engineer, the glass is twice as big as it needs to be.

 

[DesignJunkie]

Thanks John,

I appreciate the tip! Can anybody explain to me why my method won't work? Using planes to achieve angular alignment is great and I can already see that it will be invaluable in the future. But my method (if it worked) would also serve as verification that the bolt hole patterns are in fact identical. It's a method that I've used successfully in several other CAD packages. I understand that when you switch packages you need to prepared to let go of the dogma of the previous package, but this one will be hard for me...

 

[ewh]

Try applying your constraints in a different order... center diameter, bolt holes, then your distance constraint. If your hole pattern diameters are not identical your constraints will not work.

"Good to know you got shoes to wear when you find the floor." - [small]Robert Hunter[/small]

 

[DesignJunkie]

ewh,

These are good bits of advice - suffice it to say that that's where I started too! I did find that the two bolt pattern diameters varied by .0005". After rectifying this problem I still can't constrain these pieces the way I would like. There may be other iterations of constraining that I've yet to try, but I think I got most of them. Will keep looking...

 

[JohnRBaker]

Try this. Don't align the centerlines or the angles of the two parts at all, but rather just align TWO sets of Holes since in reality, if you were actually assembling something like this, it would be the location of the HOLES which would determine the alignment of the parts.

John R. Baker, P.E.
Product 'Evangelist'
Product Design Solutions
Siemens PLM Software Inc.
Industry Sector
Cypress, CA

http://www.siemens.com/plm

http://www.plmworld.org/museum/

To an Engineer, the glass is twice as big as it needs to be.

 

[DesignJunkie]

Ok,

I designed an experiment and got some interesting results that I'll share with you. I created a simple disc with two holes in it on a 2" B.C.D. I then assembled two of these discs together as per my previous posts, using two instances of the same part file. Now my technique of constraining the bolt holes works. In fact, I tried a number of different iterations and they all worked. So clearly there is something wrong in my other assembly, some fine difference that I'm not picking up on. Thanks everybody for the help and suggestions! I guess I need to go back and study my assembly a bit more carefully.

-DJ

 

[rrrocket]

Try this for an approach. My company is still on NX4, but the solution should still be valid using Assembly Constraints, rather than Mating Conditions. Align the two components centrelines and Mate or Distance two faces of each component as before. Then select the ‘Angular’ constraint, and from the drop-down menu, select ‘Orient’, rather than ‘Planar’.

Select two of the hole faces, one in each component, that you wish to ‘clock’, then hit ‘Apply’ or ‘OK’. This obviates the need to play around with Reference sets or to align datum planes.

Incidentally, I always try to align the holes first, as this forms a good assembly check, and reveals things like a PCD mismatch. If the two sets of holes do not align, use the above method – often the error can be as simple as a rounding error if an imperial part is used in a metric assembly.