Please login with a confirmed email address before reporting spam
Posted:
5 years ago
20 mars 2020, 11:27 UTC−4
The reason I am trying this is that I have a problem with phase change and steep gradients. From this blog "https://www.comsol.de/blogs/introduction-to-numerical-integration-and-gauss-points/" I know that it should somehow be possible, I just can't figure out how.
"Why is this important? In finite element analysis, you may encounter fields that exhibit sharp local gradients. Some examples are problems with phase transformations or at the onset of plasticity in solid mechanics. Integrals that are computed over elements containing these kinds of jumps may have significant discretization errors. Also, the convergence of the solution can be impaired. Small changes in the solution can significantly change computed residuals when individual Gauss points change their states.
In such cases, it may be better to select a lower polynomial order than the default for the shape functions used to discretize the field. The lower resolution can be compensated by using a denser mesh. The effect is that the inevitable jumps will be confined to smaller elements having fewer integration points."
The reason I am trying this is that I have a problem with phase change and steep gradients. From this blog "https://www.comsol.de/blogs/introduction-to-numerical-integration-and-gauss-points/" I know that it should somehow be possible, I just can't figure out how.
> "Why is this important? In finite element analysis, you may encounter fields that exhibit sharp local gradients. Some examples are problems with phase transformations or at the onset of plasticity in solid mechanics. Integrals that are computed over elements containing these kinds of jumps may have significant discretization errors. Also, the convergence of the solution can be impaired. Small changes in the solution can significantly change computed residuals when individual Gauss points change their states.
>
> In such cases, it may be better to select a lower polynomial order than the default for the shape functions used to discretize the field. The lower resolution can be compensated by using a denser mesh. The effect is that the inevitable jumps will be confined to smaller elements having fewer integration points."