# IntersectSelf inconsistent on curves

**URL:** <https://discourse.mcneel.com/t/intersectself-inconsistent-on-curves/130222>\
**Category:** Rhino\
**Tags:** windows, rhino7\
**Created:** [September 15, 2021, 5:11pm UTC](https://discourse.mcneel.com/t/intersectself-inconsistent-on-curves/130222 "2021-09-15T17:11:05Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![Will\_Wang](https://sea3.discourse-cdn.com/mcneel/user_avatar/discourse.mcneel.com/will_wang/32/513624_2.png) [@Will\_Wang](https://discourse.mcneel.com/u/Will_Wang)\
**Post date:** [September 15, 2021, 5:11pm UTC](https://discourse.mcneel.com/t/intersectself-inconsistent-on-curves/130222/1 "2021-09-15T17:11:05Z")

</div>

![image](https://global.discourse-cdn.com/mcneel/uploads/default/original/4X/7/3/9/739d577f7840b62fd1b4aee62ec78c65d1da5f69.png)

as you can see, both curves have a crossing and a tangent conditions.  
yet the command `IntersectSelf` only finds the crossing on the left  
FYI, the curves are constructed by make curves tangent to an existing curve, and joining the segments.

is this a tolerance issue? why the inconsistency?  
[forum\_selfX.3dm](https://discourse.mcneel.com/uploads/short-url/tEs92xqOhqEIWr1OgqzLqxvyEtj.3dm) (27.1 KB)

---

<div class="post-metadata">

**Author:** ![pascal](https://sea3.discourse-cdn.com/mcneel/user_avatar/discourse.mcneel.com/pascal/32/479395_2.png) [@pascal](https://discourse.mcneel.com/u/pascal)\
**Post date:** [September 15, 2021, 10:18pm UTC](https://discourse.mcneel.com/t/intersectself-inconsistent-on-curves/130222/2 "2021-09-15T22:18:58Z")

</div>

Hi Will - one thing I notice is that for the interesection that fails, on the left, the curve there is a polycurve that joins (i.e. fully multiple knot) right at the intersection/tangent location. My guess is this is what is messing things up here - I’ll check with the developer.  
[RH-65616](https://mcneel.myjetbrains.com/youtrack/issue/RH-65616) IntersectSelf: missing intersection

-Pascal

---

<div class="post-metadata">

**Author:** ![Will\_Wang](https://sea3.discourse-cdn.com/mcneel/user_avatar/discourse.mcneel.com/will_wang/32/513624_2.png) [@Will\_Wang](https://discourse.mcneel.com/u/Will_Wang)\
**Post date:** [September 16, 2021, 12:40pm UTC](https://discourse.mcneel.com/t/intersectself-inconsistent-on-curves/130222/3 "2021-09-16T12:40:18Z")

</div>

thanks for creating the ticket

curve on the right should have the same condition ![tangentcurves](https://global.discourse-cdn.com/mcneel/uploads/default/original/4X/5/2/3/52399e5645e7d67a86122e4dfe3b4eebb93fe175.gif)

both curves were created in the same way - tangent off of existing curve(c0), along tangent first and then along reversed tangent. join the two and finally join with c0