[Hughtong]

I currently calculating case of wind around buildings and encoutered weired velocity rise. The geometry is as follows, the flow direction is from north to south, with the value of (1 0 0). The mesh contain some grid with high non-orthogonal

Checking geometry...
 78     Overall domain bounding box (-621.9 0.299993 -126.364) (662.617 50.66    62 175.364)
 79     Mesh has 3 geometric (non-empty/wedge) directions (1 1 1)
 80     Mesh has 3 solution (non-empty) directions (1 1 1)
 81     Boundary openness (1.66252e-17 -2.51027e-15 -1.64139e-16) OK.
 82     Max cell openness = 2.35866e-16 OK.
 83     Max aspect ratio = 11.1171 OK.
 84     Minimum face area = 0.000321178. Maximum face area = 103.224.  Face a    rea magnitudes OK.
 85     Min volume = 7.6199e-06. Max volume = 324.355.  Total volume = 1.7993    8e+07.  Cell volumes OK.
 86     Mesh non-orthogonality Max: 76.2032 average: 17.5263
 87    *Number of severely non-orthogonal (> 70 degrees) faces: 17.
 88     Non-orthogonality check OK.
 89   <<Writing 17 non-orthogonal faces to set nonOrthoFaces
 90     Face pyramids OK.
 91     Max skewness = 1.00769 OK.
 92     Coupled point location match (average 0) OK.
 93 
 94 Mesh OK.
 95 
 96 End

However, when I do the simulation in OpenFOAM(I tried OpenFOAMv2306, OpenFOAMv2206 and OpenFOAM-7). The max(mag(U)) keeps rising and finally the case blow up. try `cat log.txt | grep "max(mag(U))"` will show this.

I've tried many ways to solve this but still get the problem：

1. I tried the official case settings in OpenFOAM2306, and use laminar to simulate(All the tests are done in laminar mode)
2. I think the intlet and outlet boundary condition matters, the bc in windAroundBuildings are:

# | | p | U |
# | ---- | ---- | -------- |
# | TOP_2D | symmetry | symmetry |
# | GROUND_2D | zeroGradient | noSlip |
# | WEST_2D | symmetry | symmetry |
# | EAST_2D | symmetry | symmetry |
# | SOUTH_2D | totalPressure| pressureInletOutletVelocity |
# | NORTH_2D | zeroGradient | fixedValue |
# | WALLS_2D | zeroGradient | noSlip |

and I also tried:

# method 1, windAroundBuildings in ofv2306
# | inlet | outlet |
# p | zeroGradient | totalPressure
# U | fixedVaule | pressureInletOutletVelocity

# method 2, simpleCar in ofv2306
# | inlet | outlet |
# p | zeroGradient | fixedValue |
# U | fixedValue | pressureInletOutletVelocity

# method 3, turbulentFlatPlate in ofv2306
# | inlet | outlet |
# p | zeroGradient | fixedValue
# U | fixedValue | zeroGradient |

# method 4, airfoil in ofv2306
# | inlet | outlet |
# p | freestreamPressure | freestreamPressure
# U | freestreamVelocity | freestreamVelocity

# method 5, motorBike in ofv2306
# | inlet | outlet |
# p | zeroGradient | fixedValue
# U | fixedValue | inletOutlet

3. I tried to use some bounded fvSchemes

divSchemes
{
    default         none;

    // div(phi,U)      bounded Gauss linearUpwind limited;
    
    // tttt: try-1
    div(phi,U)      bounded Gauss limitedLinear 1;

    turbulence      bounded Gauss limitedLinear 1;
    div(phi,k)      $turbulence;
    div(phi,epsilon) $turbulence;

    div((nuEff*dev2(T(grad(U))))) Gauss linear;
    
    //tyj add
    div(phi,C)  Gauss linear;
}

4. I also tried to implement some else matrix solver, but I think this will not influence the velocity?

    p
    {
        solver          GAMG;
        smoother        GaussSeidel;
        tolerance       1e-6;
        relTol          0.01;
    }
    pFinal
    {
        $p;
        tolerance       1e-06;
        relTol          0.01;
     }
    "(U|k|epsilon|omega|C)"
     {
        solver          PBiCGStab;
        preconditioner  diagonal; 
        tolerance       1e-7;
        relTol          0.0;
        minIter		2;
     }

    "(U|k|epsilon|omega|C)Final"
    {
        $U;
        relTol          0.0;
    }

}
SIMPLE
{
    residualControl
    {
        p               1e-6;
        U               1e-6;
        "(k|omega|epsilon)" 1e-6;
    }
    nNonOrthogonalCorrectors 2;
    pRefCell        0;
    pRefValue       0;

}

5. MOST IMPORTANT! The case can be simulated in ANSYS FLUENT. I attempted using ANSYS FLUENT to create an initial velocity field, which is physically correct with a maximum velocity magnitude of 3.2 m/s, just as the picture below. I then converted the FLUENT results to OpenFOAM, but encountered same problem!



Any suggestions or solutions to this problem are welcomed !!!It appears that I am unable to solve this particular issue with my current knowledge...

[Hughtong]

anyone can help me - -