Swirl curves with in a defined zone

[adrian.yiu (YAdrian)]

Hi There,

I am trying to do a 2D swirling pattern.
A bit like a tornado.
However, after multiple attempts, I failed to keep the green spiralling curves within the geometry itself and not overlapping.
Could someone help point me in a direction on how I can make a nice spiralling pattern?

Thank you very much in advance.

260709_Swirl_Partial.gh (17.8 KB)

[andy.vanmater]

This is a fun problem. In order to make sure the spirals sit between the curves you need to map out each ‘step’ and then walk along the steps. have a look here…

Spiral_between_Curves.gh (28.0 KB)

[laurent_delrieu]

You can also use some field technique, so you can choose the angle.

Point flows on Meshes

Hello! I had been trying to figure out how to achieve a somewhat vector-field like line flow on a mesh. My ambition is to somewhat evenly cover most of the mesh surface with the flow lines while reducing overlap / crossing over of the lines as much as possible. My approach uses the looping in Anemone to create incremental movement of points which are then interpolated. My starting point was a typical water-flow simulation where the points populated on a mesh are moved down on the z-axis. The …

Gradient Flow on Mesh

A new tool that will make a particle flowing along the maximum gradient on a mesh. If no weights on mesh vertex, the tool will use the height of the vertex, so by default it is a Rain Flow. A step is per face so you don’t need to input a step distance but rather a maximum number of steps. [MeshGradientFlow_bunny] [MeshGradientFlow_bunnyScript] [MeshWeightFieldLine_1] [MeshWeightFieldLine_2] [MeshWeightFieldLine_3] [MeshWeightFieldLine_4] [MeshWeightFieldLine_5] There will be an optio…

flow.gh (18.0 KB)

Scaling is just here because I have a bug with geometry that have big sizes

[Joseph_Oster]

@andy.vanmater

This is a fun problem.

Maybe, but I’m rusty and have already spent too much time on this. I have something more elaborate but this very simple code demonstrates my problem.

Why is BzSpan producing a straight line? Can anyone fix this?

260709b_Swirl.gh (12.7 KB)

R7 or R8, no plugins.

P.S. I can imagine that one or both of the tangent inputs (‘At’ and ‘Bt’) need to be reversed, but I’ve tried everything and it still fails.

OMG!! I got it by adding an expression to both tangent inputs:
-x*5000

How ridiculous!! The immense size of the initial curves was the problem!

260709c_Swirl.gh (12.9 KB)

[Joseph_Oster]

@Joseph_Oster

I have something more elaborate…

OH NO I lost the more elaborate code

[Joseph_Oster]

This isn’t the same as the code I lost but the general idea is the same. Needs more work…

260709d_Swirl.gh (16.9 KB)

Not fun at all

[adrian.yiu (YAdrian)]

thank you so much all for the suggestions!
this is exactly what I’m looking for.
Thanks a lot everyone!

[Joseph_Oster]

There are several problems with the curves you provided.

• Of the four curves you posted, I used two (after Join - cyan group).
• The curves are HUGE, which (with my negligence) contributed to problems yesterday.
• The inner (smaller) curve was not closed, it had a small gap, too small to notice.
• The two curves had opposite directions, requiring Flip Curve on one of them (orange group).
• Seam was required to align point pairs (orange group).

Yesterday I used Range to get ‘t’ values, which is generally a bad practice. It resulted in uneven distribution of points on the inner curve.

So I used ‘t’ values from Divide Curve instead (gold group).

260711a_Swirl.gh (47.3 KB)

The teal-colored group (‘HACK1’) is an attempt to externalize the silly expression I used on tangent inputs (-x*5000) by using curve lengths. It might break in some circumstances?

Below the gold group (unseen) is a disabled purple group (‘HACK2’ with “faint” wires) that tries and fails to use an alternative method using Connect - ignore it