Using an expression for d in atmBoundaryLayerInlet results in incorrect description upon writing

Summary

As the displacement height d in atmBoundaryLayerInlet boundary conditions is a PatchFunction1 I would like to use it to describe the local ground slope using an expression. When writing the data, the entry for d is replaced by a subDict that is verbatim the original entry for the entire patch.

Not sure, if this is relevant, but it looked close to the following issue: #2101 (closed)

Steps to reproduce

Run a simpleFoam simulation with an expression for the d entry in an atmBoundaryLayerInlet type of boundary condition and look at the output (U, k and omega all had the same issue) at write time.

Example case

The example input and output are shown in the k files in this archive: example.zip

What is the current bug behaviour?

The entry for d gets replaced with a subDict instead of the original input upon writing: 0/k: `
domain {

    type               atmBoundaryLayerInletK;
    flowDir            ( -0.000000 -1.000000 0 );
    zDir               (0 0 1);
    Uref               5.0;
    Zref               60.0;
    z0                 uniform 1.6;
    initABL            true;
    value              $internalField;
    d                  expression;
    expression         "(pos().x() > 0.0) ? 0.0 : (pos().x() < -500.0) ? -5 + 2.5*((-pos().x()-500)/1000) : pos().x()/200 ";
}

25/k:
domain

{
    type            atmBoundaryLayerInletK;
    initABL         0;
    kappa           0.41;
    Cmu             0.09;
    C1              0;
    C2              1;
    flowDir         constant (0 -1 0);
    zDir            constant (0 0 1);
    Uref            constant 5;
    Zref            constant 60;
    z0              constant 1.6;
    d
    {
        type            atmBoundaryLayerInletK;
        flowDir         ( 0 -1 0 );
        zDir            ( 0 0 1 );
        Uref            5;
        Zref            60;
        z0              uniform 1.6;
        initABL         true;
        value           uniform 1.0511;
        d               expression;
        expression      "(pos().x() > 0.0) ? 0.0 : (pos().x() < -500.0) ? -5 + 2.5*((-pos().x()-500)/1000) : pos().x()/200 ";
    }
    value           nonuniform List<scalar> 

8400 (.... ....);

`

What is the expected correct behavior?

The entry for d should be kept verbatim from the input, so the output is: 25/k: `
domain

{
    type            atmBoundaryLayerInletK;
    initABL         0;
    kappa           0.41;
    Cmu             0.09;
    C1              0;
    C2              1;
    flowDir         constant (0 -1 0);
    zDir            constant (0 0 1);
    Uref            constant 5;
    Zref            constant 60;
    z0              constant 1.6;
    d               expression;
    expression      "(pos().x() > 0.0) ? 0.0 : (pos().x() < -500.0) ? -5 + 2.5*((-pos().x()-500)/1000) : pos().x()/200 ";
    value           nonuniform List<scalar> 

8400 (.... ....);

`

Relevant logs and/or images

Environment information

• OpenFOAM version : v2412
• Operating system : OpenSUSE LEAP 15.5
• Hardware info : 192GB Ram, 24 core intel i9-14900K
• Compiler : gcc

Possible fixes