[evan247]

Hi all,

I'd like to mesh a channel whose upper side is a convex curve (think of y=x^2) and study the BL separation along that side. In order to resolve BL better, I used the "Geometric 2" law in edge parameter settings, with the spacing 2 to be "1e-5" and ratio 2 to be "1.2", so as to get a Y plus (based on the physics) of 1.0. Channel width is around 1.0 (meters), and the convex curve is around 0.6m in arc length. The curve is created by an input of 500 points.

But here goes the problem: the best I can do for Spacing 2 is 9e-5, if I wanna have a mesh of good quality (angle > 18 deg, determinant > 0.3 say). If I push it a bit further, extremely small angles and determinant start to pop up.

I tried to zoom in (a lot) to see what's going on for these troubled elements: I do spot some weird intersection points where there supposed to be only 2 lines crossing each other. (see last pic)

I'd appreciate it if anyone can help suggest possible causes for the meshes' low quality. Is it possibly due to precision as I'm forcing the spacing to be very small?

Thanks very much.

[kad]

Try to lower Triangulation Tolerance under "Settings->Model". Choose a value that is e.g. factor 0.1 or 0.01 smaller than your smallest setting for spacing.

[evan247]

Originally Posted by kad

Try to lower Triangulation Tolerance under "Settings->Model". Choose a value that is e.g. factor 0.1 or 0.01 smaller than your smallest setting for spacing.

Thanks Kad. I tried and now it works!