Specified Monin-Obukhov Length #1418

hgopalan · 2024-12-20T16:08:23Z

There are three ways to specify non-neutral stratification for ABL simulations: (i) specified heat-flux, (ii) specified temperature or heating/cooling rate and (iii) specified Monin-Obukhov length. This PR adds the capability for the third methodology in which the Monin-Obukhov length is kept constant. This method is used often in RANS simulations to do wind turbine siting for non-neutral cases (faster convergence rate) and minimize the mismatch between wall boundary condition and inflow profile.

This method is also robust for running SBL cases in which the non-linear iteration for MOL may not converge.

Please check the type of change introduced:

Bugfix
[X ] Feature
Code style update (formatting, renaming)
Refactoring (no functional changes, no api changes)
Build related changes
Documentation content changes
Other (please describe):

Checklist

The following is included:

new unit-test(s)
new regression test(s)
documentation for new capability

This PR was tested by running:

the unit tests
- on GPU
- [X ] on CPU
the regression tests
- on GPU
- on CPU

hgopalan · 2025-02-24T06:07:56Z

Need additional test cases for immersed forcing

moprak-nrel · 2025-02-24T18:33:31Z

amr-wind/equation_systems/temperature/source_terms/DragTempForcing.H

+    amrex::Real m_gamma_h{5.0};
+    amrex::Real m_beta_h{16.0};
+
+    static amrex::Real stability(


This function seems to be getting repeated implementations across multiple files? Just have one and call those?

I did not know how to do that across namespaces so created one at every place.

You could just put it in a common place and call it across the places you need. For instance you could add these in some type of wall function utilities. We don't want to be repeating code.

moprak-nrel · 2025-02-24T18:35:27Z

test/test_files/abl_specifiedmol_neutral/rans_1d.info

Is the rans_1d.info file identical across the cases? You might be able to just have one and reuse it for the other tests?

No they are different for each case.

moprak-nrel · 2025-02-24T18:39:45Z

amr-wind/equation_systems/icns/source_terms/DragForcing.cpp

    const amrex::Real kappa = 0.41;
    const amrex::Real cd_max = 1000.0;
+    amrex::Real non_neutral_neighbour = 0.0;


Can you const these with a ternary expression?

moprak-nrel · 2025-02-24T18:42:14Z

amr-wind/equation_systems/icns/source_terms/DragForcing.cpp

-                uTarget * ux2r / (tiny + std::sqrt(ux2r * ux2r + uy2r * uy2r));
-            const amrex::Real uyTarget =
-                uTarget * uy2r / (tiny + std::sqrt(ux2r * ux2r + uy2r * uy2r));
+                ustar / kappa * (std::log(0.5 * dx[2] / z0) - non_neutral_cell);


Are these calcs/the model going to be in the documentation? Not sure how to interpret non_neutral_cell and non_neutral_neighbor.

The walk through talks about the details already. The non_neutral are stability function corrections.

moprak-nrel · 2025-02-24T18:44:20Z

amr-wind/equation_systems/temperature/source_terms/DragTempForcing.cpp

    const auto& geom = m_mesh.Geom(lev);
    const auto& dx = geom.CellSizeArray();
    const amrex::Real drag_coefficient = m_drag_coefficient / dx[2];
    amrex::FArrayBox ref_theta_fab(bx, 1, amrex::The_Async_Arena());
    amrex::Array4<amrex::Real> const& ref_theta_arr = ref_theta_fab.array();
    m_transport.ref_theta_impl(lev, mfi, bx, ref_theta_arr);
+    const amrex::Real gravity_mod = 9.81;


Do you want to be using the value of gravity from the input file? We have incflo.gravity or some such right?

moprak-nrel · 2025-02-24T18:50:44Z

amr-wind/equation_systems/temperature/source_terms/DragTempForcing.cpp

+        //! We do not know the actual temperature so use cell above
+        const amrex::Real thetastar =
+            theta * ustar * ustar / (kappa * gravity_mod * mol_length);
+        const amrex::Real surf_temp =


Is this some type of interpolation to get to surf temp? Wondering why this is using theta2 vs say theta or the temperature at the bottom cell?

For terrain, the value at the point above the cell is computed from the solution as it is not a wall. However, we need to force a MOST boundary condition as a forcing term. So we correct the theta at the cell using the cell above it by computing what is the MOST solution. We already had this in place for wind speed. It gets more complicated with temperature.

moprak-nrel · 2025-02-24T18:51:21Z

amr-wind/equation_systems/tke/source_terms/KransAxell.H

+    amrex::Real m_gamma_m{5.0};
+    amrex::Real m_beta_m{16.0};
+
+    static amrex::Real stability(


Same comment as repeated definition of this function

moprak-nrel · 2025-02-24T18:54:23Z

amr-wind/wind_energy/ABLWallFunction.cpp

+                    if (zeta > 0) {
+                        psi_m = -m_gamma_m * zeta;
+                    } else {
+                        amrex::Real x =


This variable name is confusing, this is not plo + i*dx.

Copied from here: https://github.com/Exawind/amr-wind/blob/main/amr-wind/wind_energy/MOData.cpp

Why not just call that function?

Also context matters, that is a standalone small function where x isn't going to get confused with the spatial direction. Which isn't the case in your code.

Why not just call that function?

I want to have the flexibility of using a different stability function in the future within the context of this model. Currently it is set to the same as the one in MODData.H.

In that case, create a function for calculating psi_m/psi_h in your namespace and call the MOData function within that. Or simply just call the MOData function for now, and replace it when you add a different stability function.

moprak-nrel · 2025-02-24T18:56:49Z

amr-wind/wind_energy/ABLWallFunction.cpp

+                            const amrex::Real vv = vold_arr(i, j, k, 1);
+                            const amrex::Real wspd =
+                                std::sqrt(uu * uu + vv * vv);
+                            const amrex::Real theta2 = told_arr(i, j, k);


maybe a more descriptive variable name? you have theta2 to refer to theta at k+1 in a different place.

In this case the value at k is correct so I am referring to it as theta2.

First commit. RANS no support yet.

cfcbed7

hgopalan added the enhancement New feature or request label Dec 20, 2024

hgopalan added 24 commits December 20, 2024 09:39

RANS support and GPU change

92b79fb

Static.

d9478a0

clang

d0f38e8

Merge branch 'main' into mol_driver

cb3e532

RANS model broke

c78b9a9

Clang

f2cc667

Temperature support for non-neutral on terrain

f433f57

clean

aac5326

Fix

781ef4e

Merge branch 'Exawind:main' into mol_driver

c8ed246

Merge branch 'Exawind:main' into mol_driver

d3ee3e8

Update

3978346

consistency

da189d4

Merge branch 'Exawind:main' into mol_driver

40dc226

Restore

eb629ed

Consistenct

250ed57

Consistency

a820f22

Consistency

083851b

Merge branch 'Exawind:main' into mol_driver

c9b527e

Update

87e0d48

Clang

eea2a20

Warning

78d9cae

Merge branch 'Exawind:main' into mol_driver

3c8bb92

Works

491deb3

hgopalan marked this pull request as ready for review February 24, 2025 05:38

hgopalan requested review from marchdf and moprak-nrel February 24, 2025 05:39

tiny

f239478

hgopalan added 2 commits February 23, 2025 22:55

Reg Test

72a2c56

Docs

4d401cc

hgopalan added 2 commits February 23, 2025 23:17

TKE aware

5224856

gpu

8cbac3a

moprak-nrel requested changes Feb 24, 2025

View reviewed changes

hgopalan added 5 commits February 24, 2025 13:07

Clean

1876108

Merge branch 'Exawind:main' into mol_driver

8924cf4

bug

ee32454

Replacing with soil temperature

eb5dfae

Fix

2e2dfed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Specified Monin-Obukhov Length #1418

Specified Monin-Obukhov Length #1418

hgopalan commented Dec 20, 2024 •

edited

Loading

hgopalan commented Feb 24, 2025

moprak-nrel Feb 24, 2025

hgopalan Feb 24, 2025

moprak-nrel Feb 24, 2025

moprak-nrel Feb 24, 2025

hgopalan Feb 24, 2025

moprak-nrel Feb 24, 2025

moprak-nrel Feb 24, 2025

hgopalan Feb 24, 2025

moprak-nrel Feb 24, 2025

moprak-nrel Feb 24, 2025

hgopalan Feb 24, 2025

moprak-nrel Feb 24, 2025

moprak-nrel Feb 24, 2025

hgopalan Feb 24, 2025

moprak-nrel Feb 24, 2025

moprak-nrel Feb 24, 2025

hgopalan Feb 24, 2025

moprak-nrel Feb 24, 2025

moprak-nrel Feb 24, 2025

hgopalan Feb 24, 2025

Specified Monin-Obukhov Length #1418

Are you sure you want to change the base?

Specified Monin-Obukhov Length #1418

Conversation

hgopalan commented Dec 20, 2024 • edited Loading

Checklist

hgopalan commented Feb 24, 2025

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

hgopalan commented Dec 20, 2024 •

edited

Loading