Sweep1 produces strange results

Hi,

I need to work with sweep1 and noticed some strange effects. Searching older discussions seems that many users had to struggle with that component.

The main problem is, that the GH sweep does not produce the same results as the rhino sweep. Even if you work with the curves created from GH the resulting surfaces look different.

I made some experiments to find out the problem, but I failed.

First I created a Text (surfaces) to sweep the contours along a created path.

If the path (railcurve) is straight there is no problem.

When I create a curvefillet as rail, you can see that contours which have the same rail produce an inconsistent surface. (definition sweep1-A/B)

In both defintions the rail starts at the area center of the contours.

As I read in older discussions that GH makes a difference between closed and open curves I created railcurves starting at the start point of the closed contourcurves and sweeped them along the new rails (definition sweep1-C)

And while you are reading this post you can guess that this strategy was not successful.

If anyone out there had the same problem, please share your knowledge!

And if anyone has an idea to test, please let me know.

best,

teejot

Attachments:

Replies to This Discussion

Permalink Reply by Chris Tietjen on February 1, 2012 at 6:45am

Attachments:

Sweep1Rail.gh, 13 KB

Permalink Reply by Thomas J. Mrokon on February 1, 2012 at 8:30am

Hi Chris,
thanks for your solution. It works in this particular case, but what is the explanation for that?
I will create my railcurves much more complex and therefore I need an advise because I don´t want to end up with similar results.
The problem could depend on the data structure because I modified my solution to get the same data-structure input for the sweep operation like in your definition.
It worked but I am still unsatisfied with that "workaround". There are still some strange results depending on the curve type. If its periodic or not.
It would be great to get an detailed description of how the sweep-component in GH works different from the sweep-operation in Rhino.

Thanks!

Permalink Reply by Chris Tietjen on February 1, 2012 at 7:25pm

Here's a revised definition that is a little bit more compact.

In your 1-A definition (which is the only one I looked at) you're not recognizing that the centroids need to be taken for each closed curve forming the letter and not the surface as a whole. Then you need to extrude those closed section curves along a rail curve that originates from those respective centroids. If you're going to turn the curve through 90 deg. and maintain the proportionate lengths of the 'legs' of the curve on each side of the fillet you'll find the intersection points or 'kink' location in the bisecting plane (in this case at -45 deg) of the two 'legs'. The other thing to remember about the sweep components is that the rail curve has a start and an end (even if it's periodic) and it makes no sense to have the section curve located at the end of the rail curve. The rail has to lie ahead of the section curve just like the destination has to be ahead of the train otherwise 'you're already there.'

I can't tell you anything about how the sweep component in Rhino differs from the one in GH. Best solution for now is to just post examples where a problem exists. Usually a system for handling the type of work you want to do will develop.

Attachments: