Modelling trees with cfd

A subforum specific to the development of the OpenFoam-based workbenches ( Cfd https://github.com/qingfengxia/Cfd and CfdOF https://github.com/jaheyns/CfdOF )

Moderator: oliveroxtoby

thschrader
Posts: 1394
Joined: Sat May 20, 2017 12:06 pm
Location: Germany

Modelling trees with cfd

Postby thschrader » Tue Jul 17, 2018 1:29 pm

Hello everyone,

I want to calculate the pressure-distribution on my little house.
But the house is surrounded by trees. The trees have an effect on the
wind-flow, but it makes no sense to model the trees as solids.

I found this article
https://pdfs.semanticscholar.org/a0e6/9 ... fea0e1.pdf
They use darcy-forchheimer law for simulating the windbreak-effect of rows of trees.

Idea: use the porous-zone dialog in the cfd-wb.
But how must I interpret the input-dialog?
Is d=mu/k and f=1/k1 ?
What does the principal-direction mean?
When I estimate the „porosity“ k1 of a tree with 0,5, f=2.0. Is that right?
The vector e2 is missing in porosityProperties-dict.

I set up a turbulent test case, the „tree“ is the pink porous zone. I set the
turbulence-intensity to 15%, so I neglected the linear term in the darcy-forchheimer law.
Which gives zero for d in input dialog. Correct?

Comments welcome
Thomas
darcy_forchheimer_law.JPG
darcy_forchheimer_law.JPG (64.46 KiB) Viewed 218 times
result.JPG
result.JPG (82.83 KiB) Viewed 218 times
User avatar
oliveroxtoby
Posts: 253
Joined: Fri Dec 23, 2016 9:43 am
Location: South Africa

Re: Modelling trees with cfd

Postby oliveroxtoby » Tue Jul 17, 2018 2:02 pm

thschrader wrote:
Tue Jul 17, 2018 1:29 pm
Hello everyone,

I want to calculate the pressure-distribution on my little house.
But the house is surrounded by trees. The trees have an effect on the
wind-flow, but it makes no sense to model the trees as solids.

I found this article
https://pdfs.semanticscholar.org/a0e6/9 ... fea0e1.pdf
They use darcy-forchheimer law for simulating the windbreak-effect of rows of trees.

Idea: use the porous-zone dialog in the cfd-wb.
But how must I interpret the input-dialog?
Is d=mu/k and f=1/k1 ?
In this notation, f = 2/k1 and d = 1/k.
thschrader wrote:
Tue Jul 17, 2018 1:29 pm
What does the principal-direction mean?
One can have anisotropic drag in which the coefficient is different in different directions. Then the principal directions are orthogonal directions in which there is no rotation (i.e. drag parallel to velocity) - e.g. normal and parallel to a screen of tubes. If the coefficients are the same in each direction, they are irrelevant.
thschrader wrote:
Tue Jul 17, 2018 1:29 pm
When I estimate the „porosity“ k1 of a tree with 0,5, f=2.0. Is that right?
No, I don't think so; k and k1 in this formula appear to be permeabilities, not porosities.
thschrader wrote:
Tue Jul 17, 2018 1:29 pm
The vector e2 is missing in porosityProperties-dict.
No, it is not necessary. OpenFOAM calculates it orthogonal to the other two.
thschrader wrote:
Tue Jul 17, 2018 1:29 pm
I set up a turbulent test case, the „tree“ is the pink porous zone. I set the
turbulence-intensity to 15%, so I neglected the linear term in the darcy-forchheimer law.
Which gives zero for d in input dialog. Correct?
I don't think d should be ignored...