Sketching elliptical segments

[aukselis-c1connections]

🔍 Before submitting the issue

• I have searched among the existing issues
• I am using a Python virtual environment

🐞 Description of the bug

There are multiple ways to create a circular segment (arc), however, it appears that there is no way to do the same for an ellipse.

Or am I missing something, and there is a way?

If I try arc with points corresponding to an elliptical segment, I get the following error:
"The start and end points of the arc are not an equidistant from the center point."

📝 Steps to reproduce

Try to create an elliptical segment.

💻 Which operating system are you using?

Windows

📀 Which ANSYS version are you using?

2025R1

🐍 Which Python version are you using?

3.13

📦 Installed packages

Not relevant

[RobPasMue]

Hi @aukselis-c1connections -- you are right, currently there is no way to implement an elliptical segment but this is great feedback! Let us work on this API shortly, I'm sure it won't be hard. Just for reference.. what kind of API, as a user would you expect? I've got a few ideas in mind but we would also like to know your feedback on this.

For me the most obvious one is a "start_angle" and "end_angle" entrypoint, so that the user can easily define the segment. The angle would be defined starting on the right-most component of the semi-major axis -- similar to the the "theta" angle shown here: https://en.wikipedia.org/wiki/Ellipse#Polar_form_relative_to_center

Another option could be to pass in two points of the ellipse and the center. This is a fairly common mathematical problem. This implementation could be benefitial especially when you know your start and end points for your segment. We would also have to pass in an option to invert the segment selected (clockwise, or anti-clockwise)

[aukselis-c1connections]

Hi @RobPasMue, Thanks for a quick response. The option with start point, end point, centre, and direction bool is quite convenient. However, it does not work if you choose your points to be antipodal (180 degrees away from each other) (to avoid this problem, you could limit the function to a maximum 90-degree segment angle). I think the most robust would be a total of 7 parameters:

• Centre coordinates ($x_c$​, $y_c$​) - 2 parameters
• Semi-major axis length ($a$) - 1 parameter
• Semi-minor axis length ($b$) - 1 parameter
• Rotation angle of the ellipse ($θ$) - 1 parameter
• Parametric start angle of the segment ($ϕ_{start​}$) - 1 parameter
• Parametric sweep angle of the segment ($Δϕ$ or end angle $ϕ_{end}$​) - 1 parameter

[RobPasMue]

I like that approach for sure! Much more robust. I'll try to work on this soon. Thanks @aukselis-c1connections for your suggestions!

[aukselis-c1connections]

Thanks! In the meantime, are there any known workarounds? Perhaps drawing full ellipses and trimming out what is not needed. But I also seem not to find any trim functionality.

[RobPasMue]

That is a good point. I don't think there are - in my mind, the quick workaround would be to discretize manually the ellipse segment you are interested in (i.e. break your elliptical segment into N points and create straight, line segments between each of those points).

Not ideal, and time consuming -- but a temporary workaround nonetheless.