[ROltaito]

Hi,
I'm trying to create a mapped quad mesh on this aerofoil that increases in vertical refinement up to the blade from the inlet and then gets coarser again away from the blade.


When not using a mapped mesh this kind of works in tris, but when I map it (by specifying trailing and leading edge points as the sides in the mapping algo) then the bias switches to give the mess attached. The top sides bias correctly but the bottom sides bias away from the aerofoil. Any ideas please?


Thanks,
Rowan


These are my edge sizing methods:

[siw]

You need to assign some of the edges to Reverse Bias in the Edge Sizing

[ROltaito]

Originally Posted by siw

You need to assign some of the edges to Reverse Bias in the Edge Sizing

Hi,
Thanks for your suggestion, yeah I'm aware of the reverse bias function and have tried to use it accordingly. As you can see in the edge sizing picture, the yellow edge glyphs are biased correctly.


Ah! I also have match control on the top and bottom flat faces, suppressing this makes the bias work. Unfortunately I need match control as this is a compressor cascade and is periodic vertically. Any ideas please?



Thanks,
Rowan

[siw]

Can you omit Match Control and just use the Edge Sizings (with Bias Reverse and Hard) to enforce the node distribution you want and to ensure it is periodic?

[ROltaito]

Originally Posted by siw

Can you omit Match Control and just use the Edge Sizings (with Bias Reverse and Hard) to enforce the node distribution you want and to ensure it is periodic?

Hi Stuart,
I've got the mesh mostly as I need it, without the match control. But is there a way to check that the meshes line up? Is it just a manual thing?


Do they even need to line up perfectly anyway to make the domain periodic in fluent? If it's a question of accuracy then I guess it might have to do.


EDIT: measuring the last element outside edge length on both sides closest to the blade are the same top and bottom. As you say with the hard control, I think they should be the same across the board. There is a little wobble in the middle of the outlet portion versus the nice uniformity of the inlet, but I think this will do. Thanks for your help!



Thanks,
Rowan