# Kangaroo force not recognize curve

**URL:** <https://discourse.mcneel.com/t/kangaroo-force-not-recognize-curve/210773>\
**Category:** Kangaroo\
**Tags:** windows, kangaroo, kangaroo-2\
**Created:** [October 9, 2025, 9:10am UTC](https://discourse.mcneel.com/t/kangaroo-force-not-recognize-curve/210773 "2025-10-09T09:10:52Z")\
**Posts on this page:** 7\
**Page:** 1

<div class="post-metadata">

**Author:** ![tzz](https://sea3.discourse-cdn.com/mcneel/user_avatar/discourse.mcneel.com/tzz/32/668628_2.png) [@tzz](https://discourse.mcneel.com/u/tzz)\
**Post date:** [October 9, 2025, 9:10am UTC](https://discourse.mcneel.com/t/kangaroo-force-not-recognize-curve/210773/1 "2025-10-09T09:10:52Z")

</div>

I am trying to simulate a soft film that is anchored by a several lines, but having problems with curve 5 and 6. It works fine when curve 5 and 6 are connect at the vertexes, but it does not work when curve 5 and 6 land in the middle of curve 4. Does anybody know why? Is it because of that the end points of curve 5 and 6 must land on vertices of the big rectangle boundary?

Much thanks, any help is appreciated.

 ![企业微信截图_17600001822519](https://global.discourse-cdn.com/mcneel/uploads/default/original/4X/8/e/5/8e532bb7551cb36cd397e0e4d9ae663ba5b8ca57.png)

 ![企业微信截图_17600005046218](https://global.discourse-cdn.com/mcneel/uploads/default/original/4X/0/6/7/0678c2e5d05c1363bee6ba313c5d6b94e0f995a7.png)

![]()

 ![企业微信截图_17600003213184](https://global.discourse-cdn.com/mcneel/uploads/default/original/4X/4/5/6/456874aca441b79d7389392030acf88f88ceae88.png)

 ![企业微信截图_17600003946213](https://global.discourse-cdn.com/mcneel/uploads/default/original/4X/8/8/d/88dc199773c97758ede6fdc3f57ed4d611e5e7c1.png)

 ![企业微信截图_1760000741685](https://global.discourse-cdn.com/mcneel/uploads/default/original/4X/a/0/0/a00b765be1b6f52b04fe622f6365f9c3e7bd6eb5.png)

![]()

[Attractor curves.gh](https://discourse.mcneel.com/uploads/short-url/rAw5v6GG7r3dDkaGaOlk3rFEUIK.gh) (16.4 KB)

---

<div class="post-metadata">

**Author:** ![Quan\_Li](https://sea3.discourse-cdn.com/mcneel/user_avatar/discourse.mcneel.com/quan_li/32/679256_2.png) [@Quan\_Li](https://discourse.mcneel.com/u/Quan_Li)\
**Post date:** [October 9, 2025, 9:37am UTC](https://discourse.mcneel.com/t/kangaroo-force-not-recognize-curve/210773/2 "2025-10-09T09:37:15Z")

</div>

![image](https://global.discourse-cdn.com/mcneel/uploads/default/original/4X/8/9/3/8933439b8d394156826fd7072465e7b20aba9a71.png)

---

<div class="post-metadata">

**Author:** ![tzz](https://sea3.discourse-cdn.com/mcneel/user_avatar/discourse.mcneel.com/tzz/32/668628_2.png) [@tzz](https://discourse.mcneel.com/u/tzz)\
**Post date:** [October 9, 2025, 9:44am UTC](https://discourse.mcneel.com/t/kangaroo-force-not-recognize-curve/210773/3 "2025-10-09T09:44:21Z")

</div>

I internalized base curves

---

<div class="post-metadata">

**Author:** ![Quan\_Li](https://sea3.discourse-cdn.com/mcneel/user_avatar/discourse.mcneel.com/quan_li/32/679256_2.png) [@Quan\_Li](https://discourse.mcneel.com/u/Quan_Li)\
**Post date:** [October 9, 2025, 9:54am UTC](https://discourse.mcneel.com/t/kangaroo-force-not-recognize-curve/210773/4 "2025-10-09T09:54:32Z")

</div>

These points have nothing to do with the mesh.

 ![image](https://global.discourse-cdn.com/mcneel/uploads/default/original/4X/d/a/a/daaf8126d1342412ef0d8320df7ef3e4dc5f4b30.png)

---

<div class="post-metadata">

**Author:** ![Quan\_Li](https://sea3.discourse-cdn.com/mcneel/user_avatar/discourse.mcneel.com/quan_li/32/679256_2.png) [@Quan\_Li](https://discourse.mcneel.com/u/Quan_Li)\
**Post date:** [October 9, 2025, 10:07am UTC](https://discourse.mcneel.com/t/kangaroo-force-not-recognize-curve/210773/5 "2025-10-09T10:07:25Z")

</div>

Let me be more explicit. These red - marked points are not mesh points, while the others are.

How does K2 determine which points are mesh points? It does so by measuring the distance. Specifically, the non - red points are extremely close to the points on the mesh. Therefore, K2 includes them in the simulation. In contrast, the red - marked points are too distant from the mesh vertices.

 ![image](https://global.discourse-cdn.com/mcneel/uploads/default/original/4X/0/c/1/0c1fabd0596c3137653331debc959a31f7880fe9.png)

---

<div class="post-metadata">

**Author:** ![tzz](https://sea3.discourse-cdn.com/mcneel/user_avatar/discourse.mcneel.com/tzz/32/668628_2.png) [@tzz](https://discourse.mcneel.com/u/tzz)\
**Post date:** [October 9, 2025, 10:36am UTC](https://discourse.mcneel.com/t/kangaroo-force-not-recognize-curve/210773/6 "2025-10-09T10:36:59Z")

</div>

Thanks for your reply! It makes sense, however, looking at the images below, does that mean that points on curve 5& 6 in image 2 are closer to the mesh, and points on curve 5&6 in image 1 are away from the mesh, determined by K2? Is there a way to adjust the threshold of distance that K2 is calculating? Is there a way to make image 1 work? Should I cull the points on the mesh that are close to curve 5 and 6, and have those points pulled toward curve 5 and 6? (Image at the bottom wit the yellow zone)

 ![comparsion](https://global.discourse-cdn.com/mcneel/uploads/default/original/4X/7/0/9/709163564993d31ab2f76bace815bc8be9b61f8d.jpeg)

 ![3](https://global.discourse-cdn.com/mcneel/uploads/default/original/4X/c/8/d/c8d3a8d1dd5103b515ec3192568ad7696029fa25.png)

---

<div class="post-metadata">

**Author:** ![Quan\_Li](https://sea3.discourse-cdn.com/mcneel/user_avatar/discourse.mcneel.com/quan_li/32/679256_2.png) [@Quan\_Li](https://discourse.mcneel.com/u/Quan_Li)\
**Post date:** [October 9, 2025, 10:46am UTC](https://discourse.mcneel.com/t/kangaroo-force-not-recognize-curve/210773/7 "2025-10-09T10:46:41Z")

</div>

> Is there a way to adjust the threshold of distance that K2 is calculating?

Try this.

 ![image](https://global.discourse-cdn.com/mcneel/uploads/default/original/4X/4/0/e/40eae16a2b8883f1b7be93f00ba64937c07fb5a3.png)

> Should I cull the points on the mesh that are close to curve 5 and 6, and have those points pulled toward curve 5 and 6? (Image at the bottom wit the yellow zone)

I will try this way.