[bhargav1195]

Hello Everyone,

I am trying to simulate the flow and heat analysis of high-rise (33 Floor) building using buoyantboussinesqpimplefoam.

Physics of the problem:
Every floor has two ODU(Out-Door Unit). I have registered two inlet and two outlet of ODU in the domain.Boundary conditions are
1) Pressure: Inlet - zerogradient
outlet- fixed value 0
2) Velocity: Inlet - fixed value 1
outlet- zerogradient
3) Temp: Fixed value at inlet
With this BC, even when I try to run simplefoam, solution is diverging. I have attached screenshot and my case folder in following link for reference

https://drive.google.com/open?id=122...PHKf712bvsksNh


Basically flow coming out from the outlet of ODU is at higher temp and due to natural convection it should create thermal stacking effect and flow should move towards upper floors.

ERROR
1) I have fixed flow rate at Inlet and Outlet of the ODU and I am giving fixed rate at Inlet but I am confused what should i give in velocity BC at outlet? should I give fixed flow rate or zero gradient?
2) Which BC for velocity and Pressure I should give to the upper surface of the floor from where flow should be defuse to upper floor? should I give freestream condition or zerogradient?
3)As the hight of floor increases, the inlet temperature also varies due to higher temperature flow coming from bottom floors due to thermal stacking. How to incorporate this varying temperature at inlet in buoyantboussinesqpimplefoam.

Can anyone suggest me appropriate boundary conditions and solution for this problem. Once again, I have uploaded all the case file and model geometry picture in the following link for reference

https://drive.google.com/open?id=122...PHKf712bvsksNh

Thank You

Regards,
Bhargav

[piu58]

I tried to run your case. It is not complete:

--> FOAM FATAL IO ERROR:
keyword PIMPLE is undefined in dictionary "C:/OpenFOAM/17.02/Z0Pilz-3.0.x/run/Test/system/fvSolution"

file: C:/OpenFOAM/17.02/Z0Pilz-3.0.x/run/Test/system/fvSolution from line 22 to line 74.

    From function dictionary::subDict(const word& keyword) const
    in file db/dictionary/dictionary.C at line 648.

FOAM exiting

[bhargav1195]

Hello Uwe,

Thanks for giving some time in trying to help me. In controldict file I have given simplefoam as a solver name. I have set all conditions with respect to buoyantboussinesqpimplefoam and simplefoam. But in your error message it is written that PIMPLE word is undefined. So is it anyway possible that you are trying to run the PimpleFoam? Or am I mistaken here?

Thank You again.

Regards,
Bhargav

[piu58]

> I am trying to simulate the flow and heat analysis of high-rise (33 Floor) building using buoyantboussinesqpimplefoam.

If I try to uses ..SimpleFoam I get

--> FOAM FATAL IO ERROR:
Unknown patchField type noSlip for patch type wall

Of course, this one I could repai by myself. But I ask you to give a case which ic checked, so it runs.

[bhargav1195]

Hello Uwe,

I am really sorry for this error. One bracket was missing in the P/0 file which I have rectified and now it is running. Th file I have updated in the shared folder. After few iterations epsilon and k will diverge and solution will stop. I really need help to adjust boundary conditions so that case runs as per our requirement.

Thank you so much for giving your time to help me.

Regards,
Bhargav

[piu58]

I found more errors and cured them:

--> FOAM FATAL IO ERROR:
Unknown asymmetric matrix solver PBiCGStab

Valid asymmetric matrix solvers are :

4
(
BICCG
GAMG
PBiCG
smoothSolver
)

and

--> FOAM FATAL IO ERROR:
keyword p_rgh is undefined in dictionary "C:/OpenFOAM/17.02/Z0Pilz-3.0.x/run/Test/system/fvSolution.solvers"

file: C:/OpenFOAM/17.02/Z0Pilz-3.0.x/run/Test/system/fvSolution.solvers from line 22 to line 25.

    From function dictionary::subDict(const word& keyword) const
    in file db/dictionary/dictionary.C at line 648.

FOAM exiting

But I have it running now.

This lead to a questions: Do you have any minimal experience with OF? Your problem is not a elementar one and requires some knowledge. If you don't have that I recommend to simulate much easier (but connected) cases first.

[bhargav1195]

Hello Uwe,

Thank you so much again.

I dont have any experience in OpenFOAM but I do have experience with other CFD Packages. I studied and tried few documents and cases of OpenFOAM and based on that only I am trying to simulate this case. In this case all I want is flow and temperature analysis due to two inlets and outlets. That's all. On top of that I have done the analysis for Single floor and then Two Floor combined and in both case I was able to get the desired solution. But with the same boundary condition I am not able to get the converged solution for 33 Floors. I am creating Mesh in salome. I am able to understand most of boundary condition but few only I am not able to understand

1) The Top surface from where flow will move to the atmosphere, U, P, and P_RGH BC for this TOP surface. As per my understanding, for buoyantboussinesqpimplefoam P_RGH is only required for calculation.
2) Freestream boundary..i.e. surfaces on the sides of the computational domain

can you suggest me some appropriate BC for above case or will it change according to every specific case?

Thank You so much.

Regards,
Bhargav

[piu58]

> After few iterations epsilon and k will diverge and solution will stop.


I repaired the case and it comes to an end:

Time = 24

DILUPBiCG:  Solving for Ux, Initial residual = 8.75483e-009, Final residual = 8.75483e-009, No Iterations 0
DILUPBiCG:  Solving for Uy, Initial residual = 1.50941e-008, Final residual = 1.53925e-010, No Iterations 1
DILUPBiCG:  Solving for Uz, Initial residual = 1.16955e-008, Final residual = 3.88379e-011, No Iterations 1
DILUPBiCG:  Solving for T, Initial residual = 4.70309e-007, Final residual = 7.29866e-010, No Iterations 1
GAMG:  Solving for p_rgh, Initial residual = 0.427255, Final residual = 0.00347759, No Iterations 6
time step continuity errors : sum local = 1.02881e+022, global = -1.27014e+020, cumulative = -1.27014e+020
DILUPBiCG:  Solving for epsilon, Initial residual = 4.90031e-009, Final residual = 4.90031e-009, No Iterations 0
bounding epsilon, min: 2.34381e-023 max: 3.89193e+046 average: 3.28729e+040
DILUPBiCG:  Solving for k, Initial residual = 7.44824e-007, Final residual = 2.61052e-009, No Iterations 1
bounding k, min: -4.60822e+030 max: 1.43061e+045 average: 3.67576e+039
ExecutionTime = 936.233 s  ClockTime = 937 s


SIMPLE solution converged in 24 iterations

End

> dont have any experience in OpenFOAM but I do have experience with other CFD Packages.

Commercial codes have more support of selecting the right computational environment. In OF you need to know more (in fact: Nearly all details).

Even if I have a result now I would not trust it. I am not close enough to your problem.

I strongly recommend to start with a model case, which si of simpler geometry, but contains all the physics. It would be best if you have a reference solution for that case. I would deal with more complicated cases not until having full success with a model case.

[bhargav1195]

Hello Uwe,

Thanks for the help. May i know what changes you have made? Can you share that files? I have also obtained such result but the high value of Bounding K and Bounding epsilon used to scar me. According to you, is it fine to have such value?

Now I will try to run buoyantboussinesqpimplefoam and hopefully I will get desired results.

Thank you so much for help.

Regards,
Bhargav

[piu58]

I did not analyze what happened with the k values. They are much to high, of course.

My changes:

In 0/U instead of noSlip

	type fixedValue;
		value uniform (0 0 0);

in fvSolution:

   "(U|T|k|epsilon|R)"
    {
        // solver          PBiCGStab;
		solver BICCG;
        preconditioner  DILU;
        tolerance       1e-08;
        relTol          0.01;
    }
	
	p_rgh
    {
        solver           GAMG;
        tolerance        1e-7;
        relTol           0.01;

        smoother         DICGaussSeidel;

        cacheAgglomeration true;
        nCellsInCoarsestLevel 10;
        agglomerator     faceAreaPair;
        mergeLevels      1;
    }

    p_rghFinal
    {
        solver           GAMG;
        tolerance        1e-7;
        relTol           0;

        smoother         DICGaussSeidel;

        cacheAgglomeration true;
        nCellsInCoarsestLevel 10;
        agglomerator     faceAreaPair;
        mergeLevels      1;
    }

[bhargav1195]

Hello Uwe,

Thanks so much for sharing the change. I will study about that and will further make some changes to decrease the bounding k and bounding epsilon value.

Your help is much appreciated.
Thank you so much.

Regards,
Bhargav