Projection onto Various Surfaces - Unexpected Results

In the Video attached Tools > Project is used to Project a Spline Curve onto various surfaces:

  1. Cone.
    This produces the expected result with almost instant effect, i.e., without any perceivable delay and the projected image is also positioned as expected:

  1. Angled Rectangle Sketch on an appropriate Construction Plane.
    This produces an unexpected result after a noticeable delay and the projected image is displaced in a downward direction causing a portion of the image to miss the ‘target’:

  1. Angled Rectangular Body, of similar size and position to #2 above, with only the surface nearest to the Spline selected.
    This produces an unexpected result after a noticeable delay and the projected image is displaced in a downward direction with similar results to #2 above:

  1. Angled Rectangular Body, as in #3 above, with the whole Body selected.
    This produces an unexpected result after a considerable delay and the projected image is displaced in a downward direction with additional Spline Curves and Straight Edges:

  1. Similar to #3 above using one surface of a Vertical Rectangular Body.
    The Spline hits the target perfectly after some delay, a massive time difference compared to the Cone projection:

  1. Similar to #4 above using the whole of the Vertical Rectangular Body.
    This produces an unexpected result after a considerable delay, the Spline hit the target with additional projections that are all straight Edges/Lines:

This is the Video of the above scenario:

https://share.icloud.com/photos/0m2oVT98mlhjZcBhB9AvEFEew

Note: The above defied all attempts to save after removing the few unneeded frames at either end of the clip.
EDIT: NB. The above Video will only be available until 21 April 2020

1 Like

Stan,

Very interesting to see how the Project tool interacts with various surfaces. In the example of selecting the entire body (as opposed to just one surface) it appears the Project tool tries to project the spline to all 6 surface. I find it interesting the way it reflects back onto the original spline plane.

In the last case, I would guess that the 2 reflected vertical lines are really reflected splines rotated at 90° therefore appear as straight lines because they are only in 2D. Don’t know if you get what I’m saying here but then again it’s hard to describe certain anomalies. Thanks for sharing.

-Mike

I have not used the Project tool simply because I have not needed to. You have my curiosity up on this and here are some of my findings.

The video is self explanatory. The vertical reflected line is in fact the spline at 90°. I took a cube and rotated one side at 10° and then did the spline projection. Notice how the projected spline is squished in the XZ plane. Then I rotated the surface (and new spline) to 10° in the other direction and projected the spline. Note that is a mirror image of the original spline and also squished vertically.

I don’t think this is a bug but rather the way the Project tool works. It looks like an interesting way of scaling a sketch (spline, circle, etc.) in one dimension.

-Mike

@TigerMike
Many thanks for your input.
I bought into the Reflection/Deflection idea when this first surfaced 6 months ago.

Set up, in the manner shown, it was easy to change the target objects as required.
The Cone, sitting on it’s wide base, revealed what I have accepted as the normal S3D response and set me contemplating the reasons for Flat Surfaces taking considerably longer to process.

It is noted that although the two selected/highlighted Lines/Curves appear to be on the same Sketch plane they display in Items as separate planes.

Inverting the Cone does not affect the processing speed but introduces heavy shading, this is not usual IIRC.

This demonstrates the difference in Shading of Cones, the Cone with Helix is the norm while the sudden changes exhibited have only made an appearance in this .shapr File.

ADDENDUM:
Just received your very interesting update, I hope the S3D Team will be able to make good use of this.

These results are perfectly fine - this is how projection works. Currently we are using “closest point” projection, meaning that the projected curve is created by finding the closest point of each point of curve on the surface. We could also use a directional projection, that would give different results, but in that case the direction would have to be defined.
I understand that it can be surprising at the first glance, but it works perfectly fine. Parasolid FTW!