# Sweep/loft a tapered shape along a path

**URL:** <https://discourse.shapr3d.com/t/sweep-loft-a-tapered-shape-along-a-path/40410>\
**Category:** Modeling Questions\
**Created:** [February 27, 2026, 4:05pm UTC](https://discourse.shapr3d.com/t/sweep-loft-a-tapered-shape-along-a-path/40410 "2026-02-27T16:05:01Z")\
**Posts on this page:** 18\
**Page:** 1

<div class="post-metadata">

**Author:** ![Brekelefuw](https://avatars.discourse-cdn.com/v4/letter/b/258eb7/32.png) [@Brekelefuw](https://discourse.shapr3d.com/u/Brekelefuw)\
**Post date:** [February 27, 2026, 4:05pm UTC](https://discourse.shapr3d.com/t/sweep-loft-a-tapered-shape-along-a-path/40410/1 "2026-02-27T16:05:01Z")

</div>

I can’t figure out how to sweep or loft a taper around a 180⁰ curve.

I draw the path on an axis, and then draw my beginning and ending shape on the axis perpendicular to the path, but I keep getting the error that the start and end path are on a different axis and can’t be completed, or that the loft failed because the resulting body won’t be valid.

Heres a photo of what I’m trying to do:

[Tapered - Album on Imgur](https://imgur.com/a/EImLLv1)

---

<div class="post-metadata">

**Author:** ![Oregonerd](https://sea2.discourse-cdn.com/flex002/user_avatar/discourse.shapr3d.com/oregonerd/32/8012_2.png) [@Oregonerd](https://discourse.shapr3d.com/u/Oregonerd)\
**Post date:** [February 27, 2026, 4:12pm UTC](https://discourse.shapr3d.com/t/sweep-loft-a-tapered-shape-along-a-path/40410/2 "2026-02-27T16:12:55Z")

</div>

Shapr falls flat on it’s face sometimes. I figured wrongly that scaling the end would have, you know, scaled the length of the body. If you could combine line segments into a single spline… or is this a limitation of Parasolid?

 ![FlatOnItsFace](https://us1.discourse-cdn.com/flex002/uploads/shapr3d/original/3X/7/e/7e2230b065e758a5e42902eb499d4c4aa0cdc283.jpeg)

---

<div class="post-metadata">

**Author:** ![Brekelefuw](https://avatars.discourse-cdn.com/v4/letter/b/258eb7/32.png) [@Brekelefuw](https://discourse.shapr3d.com/u/Brekelefuw)\
**Post date:** [February 27, 2026, 4:31pm UTC](https://discourse.shapr3d.com/t/sweep-loft-a-tapered-shape-along-a-path/40410/3 "2026-02-27T16:31:01Z")

</div>

I’ve had less success on other similar shapes with splines, because it will fail to generate if the shape being swept collides with itself along the path.

Most of my work in shapr3d is making dies for bending tubes, and it can be very frustrating to figure out how to work around the errors that it gives me when I’m sweeping the path.

---

<div class="post-metadata">

**Author:** ![Oregonerd](https://sea2.discourse-cdn.com/flex002/user_avatar/discourse.shapr3d.com/oregonerd/32/8012_2.png) [@Oregonerd](https://discourse.shapr3d.com/u/Oregonerd)\
**Post date:** [February 27, 2026, 4:36pm UTC](https://discourse.shapr3d.com/t/sweep-loft-a-tapered-shape-along-a-path/40410/4 "2026-02-27T16:36:31Z")

</div>

The workaround is to sweep the first straight section, scale the end, select that end and sweep and scale, rinse and repeat for each section? I’ll try it. But you would have to scale by percentages if the segments aren’t the same length? I’ll try it in a bit.

AI is gonna make all this obsolete.

---

<div class="post-metadata">

**Author:** ![Xdrakosha](https://sea2.discourse-cdn.com/flex002/user_avatar/discourse.shapr3d.com/xdrakosha/32/24881_2.png) [@Xdrakosha](https://discourse.shapr3d.com/u/Xdrakosha)\
**Post date:** [February 27, 2026, 11:01pm UTC](https://discourse.shapr3d.com/t/sweep-loft-a-tapered-shape-along-a-path/40410/5 "2026-02-27T23:01:17Z")

</div>

Maybe that will help.

---

<div class="post-metadata">

**Author:** ![Oregonerd](https://sea2.discourse-cdn.com/flex002/user_avatar/discourse.shapr3d.com/oregonerd/32/8012_2.png) [@Oregonerd](https://discourse.shapr3d.com/u/Oregonerd)\
**Post date:** [February 28, 2026, 12:00am UTC](https://discourse.shapr3d.com/t/sweep-loft-a-tapered-shape-along-a-path/40410/6 "2026-02-28T00:00:50Z")

</div>

Any idea why the results are different using the scale tool?

---

<div class="post-metadata">

**Author:** ![Brekelefuw](https://avatars.discourse-cdn.com/v4/letter/b/258eb7/32.png) [@Brekelefuw](https://discourse.shapr3d.com/u/Brekelefuw)\
**Post date:** [February 28, 2026, 4:13am UTC](https://discourse.shapr3d.com/t/sweep-loft-a-tapered-shape-along-a-path/40410/7 "2026-02-28T04:13:20Z")

</div>

That might work. Can you explain the formula you used in the scale box?

---

<div class="post-metadata">

**Author:** ![Xdrakosha](https://sea2.discourse-cdn.com/flex002/user_avatar/discourse.shapr3d.com/xdrakosha/32/24881_2.png) [@Xdrakosha](https://discourse.shapr3d.com/u/Xdrakosha)\
**Post date:** [February 28, 2026, 5:57am UTC](https://discourse.shapr3d.com/t/sweep-loft-a-tapered-shape-along-a-path/40410/8 "2026-02-28T05:57:01Z")

</div>

> [@Brekelefuw](#):
>
> Can you explain the formula you used in the scale box?

18 mm top circle, 10 mm bottom circle.

18 mm = 100%  
10 mm = ?

10 mm \* 100% / 18 mm = 55.5556%

55.5556 / 100 = 0.5556 (scale factor)

---

<div class="post-metadata">

**Author:** ![Xdrakosha](https://sea2.discourse-cdn.com/flex002/user_avatar/discourse.shapr3d.com/xdrakosha/32/24881_2.png) [@Xdrakosha](https://discourse.shapr3d.com/u/Xdrakosha)\
**Post date:** [February 28, 2026, 6:14am UTC](https://discourse.shapr3d.com/t/sweep-loft-a-tapered-shape-along-a-path/40410/9 "2026-02-28T06:14:28Z")

</div>

> [@Oregonerd](#):
>
> results are different using the scale tool?

I can only guess.

Left (scale tool), right (sweep with scale)

 ![image](https://us1.discourse-cdn.com/flex002/uploads/shapr3d/original/3X/2/e/2e72bf24ae98097f8f47674ab924449e35d3b456.png)

I assume that scale tool applies only to the part until the edges, selected on the next screeshot.

 ![image](https://us1.discourse-cdn.com/flex002/uploads/shapr3d/original/3X/3/a/3acf62671004c9d88f5123c221b171a3119f4581.jpeg)

But why exactly this is happening only S3D team can explain I think.

---

<div class="post-metadata">

**Author:** ![Oregonerd](https://sea2.discourse-cdn.com/flex002/user_avatar/discourse.shapr3d.com/oregonerd/32/8012_2.png) [@Oregonerd](https://discourse.shapr3d.com/u/Oregonerd)\
**Post date:** [February 28, 2026, 4:51pm UTC](https://discourse.shapr3d.com/t/sweep-loft-a-tapered-shape-along-a-path/40410/10 "2026-02-28T16:51:57Z")

</div>

Well logic says once the body is swept, it’s _should_ be a single body. So scaling the end should be uniform along the body. It selects as a single body. So the scale should be uniform no matter which method, logically.

Seems like a bug to me!

---

<div class="post-metadata">

**Author:** ![SCALEMODEL](https://avatars.discourse-cdn.com/v4/letter/s/77aa72/32.png) [@SCALEMODEL](https://discourse.shapr3d.com/u/SCALEMODEL)\
**Post date:** [February 28, 2026, 5:19pm UTC](https://discourse.shapr3d.com/t/sweep-loft-a-tapered-shape-along-a-path/40410/11 "2026-02-28T17:19:55Z")

</div>

> [@Oregonerd](#):
>
> Well logic says once the body is swept, it’s _should_ be a single body.

But why is there 3 segments?

---

<div class="post-metadata">

**Author:** ![SCALEMODEL](https://avatars.discourse-cdn.com/v4/letter/s/77aa72/32.png) [@SCALEMODEL](https://discourse.shapr3d.com/u/SCALEMODEL)\
**Post date:** [February 28, 2026, 5:23pm UTC](https://discourse.shapr3d.com/t/sweep-loft-a-tapered-shape-along-a-path/40410/12 "2026-02-28T17:23:57Z")

</div>

Also to add to this!

Lots of tools before did not require or should not require you to go back and adjust in the history.

---

<div class="post-metadata">

**Author:** ![SCALEMODEL](https://avatars.discourse-cdn.com/v4/letter/s/77aa72/32.png) [@SCALEMODEL](https://discourse.shapr3d.com/u/SCALEMODEL)\
**Post date:** [February 28, 2026, 5:28pm UTC](https://discourse.shapr3d.com/t/sweep-loft-a-tapered-shape-along-a-path/40410/13 "2026-02-28T17:28:41Z")

</div>

![image](https://us1.discourse-cdn.com/flex002/uploads/shapr3d/original/3X/7/8/7871c5e87080f7f0b00ec41a43738f89cff18c1c.jpeg)

but to scale!

 ![image](https://us1.discourse-cdn.com/flex002/uploads/shapr3d/original/3X/9/b/9bd8c427c0f3fa881db0355ab9e71dc6dd7978b7.jpeg)

you could of gave us that before the process.

---

<div class="post-metadata">

**Author:** ![Oregonerd](https://sea2.discourse-cdn.com/flex002/user_avatar/discourse.shapr3d.com/oregonerd/32/8012_2.png) [@Oregonerd](https://discourse.shapr3d.com/u/Oregonerd)\
**Post date:** [February 28, 2026, 7:19pm UTC](https://discourse.shapr3d.com/t/sweep-loft-a-tapered-shape-along-a-path/40410/14 "2026-02-28T19:19:57Z")

</div>

> [@SCALEMODEL](#):
>
> But why is there 3 segments?

Because the program interprets/ separates all lines/ splines into easily to described compressed descriptions? E.G the way DXF exports a curve into many smaller arcs creating many points. The solid body is interpreted essentially the same way?

What do I know. Just a worn our old carpenter that worked in an electronics company for a while. 🤣

---

<div class="post-metadata">

**Author:** ![Xdrakosha](https://sea2.discourse-cdn.com/flex002/user_avatar/discourse.shapr3d.com/xdrakosha/32/24881_2.png) [@Xdrakosha](https://discourse.shapr3d.com/u/Xdrakosha)\
**Post date:** [February 28, 2026, 8:18pm UTC](https://discourse.shapr3d.com/t/sweep-loft-a-tapered-shape-along-a-path/40410/15 "2026-02-28T20:18:50Z")

</div>

> [@SCALEMODEL](#):
>
> But why is there 3 segments?

I think that’s because we used 2 lines and 1 arc - 3 in total.

Using spline we can get one piece 🙂

 ![image](https://us1.discourse-cdn.com/flex002/uploads/shapr3d/original/3X/b/3/b3b6a6fcd1704114dc18d831254e5f2a99ab952b.png)

This time scale tool will affect on whole part but still not as we expect

 ![image](https://us1.discourse-cdn.com/flex002/uploads/shapr3d/original/3X/a/d/ad0d7d73ac16d68a43580b4541e69e6797b941d2.jpeg)

Scale tool on the right, sweep scale on the left.

---

<div class="post-metadata">

**Author:** ![SCALEMODEL](https://avatars.discourse-cdn.com/v4/letter/s/77aa72/32.png) [@SCALEMODEL](https://discourse.shapr3d.com/u/SCALEMODEL)\
**Post date:** [March 1, 2026, 2:42am UTC](https://discourse.shapr3d.com/t/sweep-loft-a-tapered-shape-along-a-path/40410/16 "2026-03-01T02:42:57Z")

</div>

I think I was not clear. This seems like a good feature but it’s in the history panel.

---

<div class="post-metadata">

**Author:** ![Xdrakosha](https://sea2.discourse-cdn.com/flex002/user_avatar/discourse.shapr3d.com/xdrakosha/32/24881_2.png) [@Xdrakosha](https://discourse.shapr3d.com/u/Xdrakosha)\
**Post date:** [March 1, 2026, 8:36am UTC](https://discourse.shapr3d.com/t/sweep-loft-a-tapered-shape-along-a-path/40410/17 "2026-03-01T08:36:27Z")

</div>

> [@SCALEMODEL](#):
>
> but it’s in the history panel

To be honest I don’t feel like this is a big deal.

---

<div class="post-metadata">

**Author:** ![Oregonerd](https://sea2.discourse-cdn.com/flex002/user_avatar/discourse.shapr3d.com/oregonerd/32/8012_2.png) [@Oregonerd](https://discourse.shapr3d.com/u/Oregonerd)\
**Post date:** [March 1, 2026, 4:22pm UTC](https://discourse.shapr3d.com/t/sweep-loft-a-tapered-shape-along-a-path/40410/18 "2026-03-01T16:22:56Z")

</div>

Be honest though. This is a work flow you are comfortable with. The reason I dropped Alibre and almost, (It was close), bought a Mac was because Shapr WASN’T history based. They had no Windows version.

Alibre’s forum environment wasn’t friendly and helpful, as engineers tend to be a bit unhelpful for the lesser folk who haven’t “paid” for an education. It was always different here.

My confusion is what do we need a tool bar for if we have a history bar? Seems a lot of the tools could just be moved to history. Instead it seems we have a bit of a kluge of a tool bar that gets less attention and a multifunction history bar.

What should have happened “intuitively” at the scale tool got relegated to the history bar. (Honestly I never thought to look there).

💡 _ **Maybe the history bar should, in real time, expand at the point where we are currently on the new body, to show us our options at that point?** _ 💡

At this point I have to collapse history to get my preferred work flow back in some cases. I’d just prefer a switch to turn history off completely.

One example: When I split a body, they sketch doesn’t reflect the dissection. So then if I want to sweep a face off that body to continue on with a new body, it sweeps the whole length back in the wrong direction as well. On the fly design thwarted.

History can be a bit of punishment now for those of us that aren’t classically trained engineers. It didn’t used to be that way. That was Shapr’s attraction. But here we are 😊