A few questions on filteredLinear2
 

Hello Foamers:

I'm trying to use the interpolation scheme filteredLinear2. If I want to use filteredLinear2 on the divergence term then I believe the usage should look like:

divSchemes
{
div(phi,U) Gauss filteredLinear2 1 0;
}

It requires two coefficients (above the first coefficient is one and the second one is zero) and I think it is important that I assign appropriate values for the two coefficients. However, I'm not quite sure what values to choose when I want to run a LES case to study the flow over a cylinder at high Re.

I've looked into the source code and included some portion of the code description.

// Scaling corefficient for the gradient ratio,
// 0 = linear
// 1 = fully limited
scalar k_;

// Maximum allowed overshoot/undershoot relative to the difference
// across the face.
// On input:
// 0 = no overshoot/undershoot
// 1 = overshoot/undershoot equal to the difference across the face
// Note: After input 1 is added to l_

It seems like the first coefficient has something to do with limiter that is

0 = no limiter so it's just linear
1 = fully limited meaning it becomes like limitedLinear

Am I understanding it correctly? Please, correct me if I'm not.

Also about the second coefficient it says it has something to do with overshoot/undershoot, which I'm not quite sure what it is about.

Could anyone explain to me the details of the algorithm in filteredLinear2 and recommend appropriate values for the two coefficients?

A.Lee

Hi, I know it’s been a long time since you’ve posted this question, I was wondering if you managed to find any answer for it. I am also interested to know about those coefficients and how to set them correctly for a LES simulation over an airfoil.
Thank you in advance

I discuss the filteredLinear2 scheme a little bit in my thesis (https://drum.lib.umd.edu/handle/1903/21883) around page 38 -- that might be useful ....

This scheme struggled for the scalar transport problem I used but for the momentum equations this scheme might work ....

One issue with this scheme is that it is hard to understand the thinking behind it (for example see equations 2.28 and 2.29 in my thesis) .... other schemes I looked at were pretty transparent but this one was not ....

Cheers ....

Quick update:

The filteredLinear2 scheme did well, when it was applied to the momentum equations, for all the test cases we explored in this recent paper https://www.sciencedirect.com/scienc...45793021003236 ....

Also the description we have for filteredLinear2 in this paper is a little bit better than the description I had in my thesis mentioned above ....