Vector Orientation Issue

Hi,

helping out a friend, and am not really a big help, it seems, as I'm asking for guidance once again on here ;P

I'm trying to construct normal planes from a list of vectors. The vectors themselves are derived from two lists of divided curve points. Everything goes well up to this point, but when I now construct normal planes from the vectors, one axis direction is not uniform - in other words, some of the vector values are negative, and I can't figure out why. Oh and yes, the whole construct resides in one coordinate space quadrant.. I had it at the origin once, going into several quadrants at once, which didn't work, either ;)
Can anyone enlighten me, please?

thanks for reading!

Attachments:

Replies to This Discussion

Permalink Reply by Max D on December 9, 2009 at 8:15am

update:

I've worked around the issue by constructing the needed planes differently - that is, by not using "plane from normal".. Still, I wonder what the original problem is!

Attachments:

LouverSketch.ghx, 258 KB

Permalink Reply by Luis Fraguada on December 9, 2009 at 8:21am

Hi,
Yes, I often avoid this component for the same reasons...sometimes you can align the planes (align components) with a guide plane.

More often than not I find myself using plane from 3 points...

Permalink Reply by Max D on December 9, 2009 at 8:52am

yes, plane from 3 points is in the workaround, too.. though I am not really sure whether the plane component is the problem, as it's fed the weird values from the vec2pt component - which due to whatever reason produces the negative numbers. huh!

Permalink Reply by Benjamin Golder on December 9, 2009 at 2:47pm

Luis is right. The trick to this component is that it only determines the positive z value for the planes, and the positive x and y directions for the planes can essentially be rotated in any direction about this z axis. Therefore if you combine it with the "align plane with a guide vector" component you can tell it which way to point it's positive x vector. Looking at your image, I would guess that it might be helpful to use the curve tangent vectors at the evaluation points as your guide vectors.

The negative numbers listed in the output from the vec2pt shouldn't be a problem as far as I can tell. All that they indicate is that your vectors point very slightly in the global negative x direction, which does not seem like an error.

Permalink Reply by Max D on December 9, 2009 at 5:06pm

thanks Luis and Benjamin! I'll try the "align plane with a guide vector" gizmo and see how it goes. But what you say makes sense - I think I was expecting the planes to simply behave uniformly, despite the x and y directions being essentially undefined through the normal plane component. The tangent vector alignment seems like a good idea! ;D

cheers,

m

RSS