My flow past airfoil case shows horizontal stripes on airfoil surface in Q criterion contour.
The result is calculated by custom solver, using incompressible model, DNS algorithm, structured grid, immersive boundary method to handle coupling of solid.
I am sorry, I am outside of school, so I can’t ssh to vncserver in school server, even though I have connected to vpn. So I can’t provide you a screenshot.
More specifically, I use Resample to Image to draw velocity section. I use NVIDIA IndeX plugin as representation of Resample to Image, take Velocity_X, Velocity_Y, Velocity_Z as the value to be resampled. And I use NVIDIA IndeX slice to show the section.
Thanks for explaining your pipeline. The Resample to Image sampling pattern is introducing aliasing of your velocity field. Some points of the resampling grid fall inside the wing, where presumably the velocity is set to 0, and some are out in the volume outside the wing, so have real values. That sharp transition across grid cell is probably what you are seeing along the wing.
IndeX is supposed to be able to render unstructured grids. It must not be available for your input structured grid. You could apply a Tetrahedralize filter to your input at the cost of a lot more memory use and try to render it with the IndeX representation. That might not be the best solution, but should at least prevent the resampling artifacts you are seeing.
You mention that the velocity field is smooth. However, in the immersed boundary method, the wing surface shape is represented as a staircase. This effect may cause the velocity vector field to appear wobbly when displayed with shadows in volume rendering.
