Dimension question - NX5

[Simon205]

Hi,

I'm pretty new to NX so this might be a simple question, but in drafting how to you dimension from a flat to the furthest point of a curved surface?

Imagine modelling a cylinder with a flat on one side: add dimension from the flat to the outermost point of the cylinder's radius.

Thanks,


Simon

 

[cowski]

Make sure you have the tangent snap point selection tool turned on before you pick the curve. If the flat is parallel to the X or Y you may be able to use the quad point of the arc (if it is a true arc).

If the curve is a spline, then it gets a bit trickier, but still possible.

 

[hudson888]

There is a tangency snap point that ought to do it.

Cheers

Hudson

 

[Simon205]

Thanks for the replies, but the tangency snap doesn't appear to work. The shape in question is (was) and ellipse that has been modified, but safe to assume that I'm after a dimension that would snap to a spline or other organic shape.

Any ideas?

Thanks,


Simon

 

[hudson888]

Have a look at using either a control point or a point on line method then. It may help rather than using an inferred dimension if you force it to a horizontal or vertical type supposing that is wanted you wanted to do. I hesitate to mention these dimensioning method which despite working quite well may be less precise for establishing the widest or narrowest point of the geometry in question.

Cheers

Hudson

 

[cowski]

Unfortunately, unless the geometry is an arc - the tangent snap point may not work for you. But there is more than one way to skin a cat so let's dig around the toolbox that NX provides...

Isocline curves may come in handy here (see attachment). In the top pic I created a block with a curvy top surface (I extruded a spline to trim the block). In the next frame down I added isocline curves, picking X as my direction vector and entering 0 degrees. This creates an associative line on the surface that I can now dimension to. Finally, the last frame just shows that if the shape changes, the isocline curve updates and the dimension follows.

I also used the same spline to create a revolved surface and was still able to dimension to it, but I have not tried complex freeform surfaces.

 

[hudson888]

Downside Isocline curves are non-associative when working in master model method with your drafting. The exception being when you link the geometry used to create them. Either way you may be confronted with need of managing that data using either layers or some other complicated technique that you may have preferred to avoid. Certainly on occasion this will be quite possibly the only accurate solution that is available to you.

Best Regards

Hudson

 

[cowski]

To keep them associative you would need to create them in your model file and add them to the reference set. Yes, it will need some managing but it is a workable solution.

 

[hudson888]

cowski,

Another method that would work. Not to criticize your solution by any stretch but to explain why I never went there is simply that we have a system whereby in quite a few cases the drawings are finalized by junior staff after the senior modelers have approved the math model designs. It means that the drafter cannot be assumed to own write access to the models. So yours inst such a bad idea just one that I wouldn't have thought of due to that likely obstacle.

Cheers

Hudson

 

[Simon205]

Thanks for the replies guys, much appreciated.

Adding Isocline curves will work in this case so I'll try working with that. I'm just amazed that the dimension manager won't snap to the outer-most extent of the geometry without having to mess about adding extra entities.

Cheers,


Simon

 

[hudson888]

Simon,

Don't be mistaken that you can still put a dimension along the curve ant get very close to the point where it would snap to even if it is only by eye. Unless you need perfect technical accuracy I would not both messing about with the isocline curves, but for the very best accuracy in your case you may have to.

Cheers

Hudson