Re: [Libmesh-users] Solution projection

[Roy S. <roy...@ic...>]

On Mon, 6 Nov 2006, Roy Stogner wrote:

> On Mon, 6 Nov 2006, Tim Kr=F6ger wrote:
>
>> * Loop over active cells only, as suggested;
>> * Use HEX27 cells instead of HEX8;
>> * Refine only one cell instead of all;
>> * Repeat refinement 10 times.
>
> Here's an oddity: I can't seem to find a 2D test case that exhibits
> this bug, and in 3D the HEX20 seems immune too.  Is it possible that
> there's something wrong with the HEX27 element specifically?

And oddity 2: I don't have any problems with second-order Lagrange FE
on HEX27.
---
Roy