[Eakarach]

I try to simulate stirring tank with rushton turbine. The RNGKE model was selected to simulate the fluid flow in tank. The experimental power number is 0.65. However, the calculated power number is higher than experiment.

Can anyone help me to solve this problems?

[Eakarach]

Can anyone help me?

[jamalf64]

Originally Posted by Eakarach

Can anyone help me?

Please write more details for example: MRF or sliding mesh, second order or first order, baffled or unbaffled tank, ...
How power number is greater than experimental data?

[Eakarach]

The unbaffled tank with diameter of 0.19 m is simulated by MRF method. The second order upwind is adopted to estimate the face value. The experimental power number is 0.65, while the simulated power number is about 1.3. The top of the tank is a lid, which prevent the presence of vortex. The impeller is rushton turbine with diameter of 0.095 m. The rotating speed is 200 rpm.

[jamalf64]

Originally Posted by Eakarach

The unbaffled tank with diameter of 0.19 m is simulated by MRF method. The second order upwind is adopted to estimate the face value. The experimental power number is 0.65, while the simulated power number is about 1.3. The top of the tank is a lid, which prevent the presence of vortex. The impeller is rushton turbine with diameter of 0.095 m. The rotating speed is 200 rpm.

If trend of diagrams(for example axial velocity, radial velocity, tangential velocity vs radial coordinate) is acceptable, I think you must refine your mesh.

[Eakarach]

Originally Posted by jamalf64

If trend of diagrams(for example axial velocity, radial velocity, tangential velocity vs radial coordinate) is acceptable, I think you must refine your mesh.

I plot tangential velocity VS radial of tank. The comparison between experiment and simulation are difference with increasing the radius. The model was simulated with the mesh size of 1 mm all of domain.

[Eakarach]

Moreover, I simulated this case with increasing the Swirl Factor of 0.11 (in RNG Options). The simulated tangential velocity and experiment were very similar but the simulated results doesn't smooth (Oscillation). The residual was also oscillation. The residual was higher (about 0.1) than the residual obtained by default Swirl Factor (0.07), (about 0.0001), which give the deviation of tangential velocity at the higher radius.

Can any one help me to solve this problems?

[jamalf64]

Originally Posted by Eakarach

Moreover, I simulated this case with increasing the Swirl Factor of 0.11 (in RNG Options). The simulated tangential velocity and experiment were very similar but the simulated results doesn't smooth (Oscillation). The residual was also oscillation. The residual was higher (about 0.1) than the residual obtained by default Swirl Factor (0.07), (about 0.0001), which give the deviation of tangential velocity at the higher radius.

Can any one help me to solve this problems?

Are you sure that the solution is converged? Do you monitor some parameters (for example area-weighted velocity magnitude in 2 plane: once above impeller and other below impeller)?
see this tutorial: "Modeling Turbulent Flow in a Mixing Tank"

[Eakarach]

I do this step as reported in the tutorial. However, these monitors show oscillation. Can anyone suggest me about this problem?

[jamalf64]

Originally Posted by Eakarach

I do this step as reported in the tutorial. However, these monitors show oscillation. Can anyone suggest me about this problem?

Ok, change second order to first order, this cause to solution will be converge and monitors dont show oscillation.If in this step calculated data is wrong, change turbulence model.
I hope this help you
Be success
Jamal

[Eakarach]

I'm try as your suggestion but the solution is wrong. However, do you have more suggestion.

Regards
B. Eakarach

[Eakarach]

Dear Jamal Foroozesh

Please send me your email. Then, i will send my case file to you.

Thank in advance

[jamalf64]

Originally Posted by Eakarach

Dear Jamal Foroozesh

Please send me your email. Then, i will send my case file to you.

Thank in advance

Before sending your file try to this:
turbulence: k-e Realizable
impeller wall: moving wall, relative to adjacent cell zone, rotational, speed=0
active gravity
pressure: body force weighted
momentum , and turbulence equation: first order upwind

[jeicek]

Originally Posted by Eakarach

I try to simulate stirring tank with rushton turbine. The RNGKE model was selected to simulate the fluid flow in tank. The experimental power number is 0.65. However, the calculated power number is higher than experiment.

Can anyone help me to solve this problems?

Hello

I am also simulation a CSTR using MRF in openFoam and I need to calculate power number. I do not know how to do this. can anyone help me and ginve an idea?
Thanks in advance

[ghost82]

Originally Posted by jeicek

Hello

I am also simulation a CSTR using MRF in openFoam and I need to calculate power number. I do not know how to do this. can anyone help me and ginve an idea?
Thanks in advance

Hi,
this is the fluent forum, you'd better to write in openfoam forum;
however power number Np is defined by:

Np=P/(N^3*Dt^5*rho)

P=Absorbed power (W)
N=Rotational speed in revolutions per second (rev/s)
Dt=Impeller diameter (m)
rho=Fluid density (kg/m3)

Absorbed power is calculated from simulation, other parameters are known.

Daniele

[jeicek]

Originally Posted by ghost82

Hi,
this is the fluent forum, you'd better to write in openfoam forum;
however power number Np is defined by:

Np=P/(N^3*Dt^5*rho)

P=Absorbed power (W)
N=Rotational speed in revolutions per second (rev/s)
Dt=Impeller diameter (m)
rho=Fluid density (kg/m3)

Absorbed power is calculated from simulation, other parameters are known.

Daniele

Thank you very much for reply Daniele. To be honest I wrote it but no aswer still also here is the thread I created
http://www.cfd-online.com/Forums/mai...rediction.html
but any way thank you very much