Completely different results for CMy after switching to 3.45.4

[Kilian]

Hi everyone!

I am currently considering upgrading my OpenVSP workflow to the newest version 3.45.4 since I use the VSPAERO solver a lot.

Setting up a test case, I found that my results for CMy differ greatly. In fact, while VSP 3.44.5 gives me a stable aircraft (negative CMy slope), VSP 3.45.4 gives me an unstable aircraft with a positive slope for CMy.

I use VLM for 3.44.5 and only thin surfaces for 3.45.4. The settings are exactly the same. This is not allowed to happen, right? What could be the reason?

Will attach screenshots.

[Kilian]

This is OpenVSP 3.44.5:

This is OpenVSP 3.45.4:

[Zach Hazen]

And I trust you double checked your moment reference location?  Can you trace the moment disparity to pressures using the viewer?  Do the meshes look very different?

[Kilian]

The meshes only differ at the intersections. The results change when changing the tessellation. However, I want to run a grid convergence study and I noticed those weird results shown above when running a rather coarse grid.

[Rob McDonald]

Can you provide a simplified test case that exhibits the change you're seeing?

Rob