[zese]

Dear friends

I am simulating multiphase flow inside a swirl injector (water-air) by VOF model. at the inlet I have to specify pressure inlet for which I have the value based on experiments. however, after simulation is converged the mass flow at the inlet is different from measurements. what do you think the problem might be?

BR

[andrew-c]

Without knowing more details about the experiment and your model, it'll be very hard to say. Some questions that might help:

- How have you judged convergence of your model?

- How similar is your geometry to the experimental case?

- What simplifications have you made in the model with respect to the experiment (this is probably the most important question) - i.e., are energy/heat transfer on? Is gravity on? Are material properties constant, or variable? Are you modeling surface tension effects?

- Are there any other parameters you can compare with the experiment besides the inlet mass flow (outlet mass flow, temperature, velocity profile, etc.)? These may help diagnose the problem.

- Are you running transient or steady state?

- What flow model(s) are you using?

If you can supply more information such as the above, it may help others to help you.

[zese]

Originally Posted by andrew-c

Without knowing more details about the experiment and your model, it'll be very hard to say. Some questions that might help:

- How have you judged convergence of your model?

- How similar is your geometry to the experimental case?

- What simplifications have you made in the model with respect to the experiment (this is probably the most important question) - i.e., are energy/heat transfer on? Is gravity on? Are material properties constant, or variable? Are you modeling surface tension effects?

- Are there any other parameters you can compare with the experiment besides the inlet mass flow (outlet mass flow, temperature, velocity profile, etc.)? These may help diagnose the problem.

- Are you running transient or steady state?

- What flow model(s) are you using?

If you can supply more information such as the above, it may help others to help you.

Thank You for your answer.

I'll answer your questions first then describe my model.
- How have you judged convergence of your model?
yes I checked convergence.

- How similar is your geometry to the experimental case?

it is the exact solid model of my experimental device.

- What simplifications have you made in the model with respect to the experiment (this is probably the most important question) - i.e., are energy/heat transfer on? NO
Is gravity on? YES
Are material properties constant, or variable? Constant, water
Are you modeling surface tension effects? YES, constant 0.73 water-air

- Are there any other parameters you can compare with the experiment besides the inlet mass flow (outlet mass flow, temperature, velocity profile, etc.)? These may help diagnose the problem. Actually, No just Mass flow and the spray angle. angle matched but mass flow no

- Are you running transient or steady state? Steady state, Implicit

- What flow model(s) are you using? VOF


The schematic of experimental setup are shown as the picture attached

I picked the number shown in the pressure gage as the total gage pressure at the inlet.

the simulation geometry is shown in the picture 2
Since I didnt want to model spray and just wanted to see inside the injector and the primary breakup I just used VOF model.
I used Implicit, steady state, sharp/dispersed model, phases (water and air).
Turbulence RNG k-epsilon
the Boundary conditions are as follow

Inlets: total pressure guage 300000 pa and water VOF 1
outlet: pressure outlet pressure guage 0
walls: no-slip

model: coupled and coupled with VOF
for VOF compressive

I stopped simulation when all residuals remain constant through the simulation after 10000 iterations.

massflow @inlet 0.139
massflow EXperimental 0.16

even simulationg by transient it did not change very much.

[andrew-c]

Originally Posted by zese

Thank You for your answer.

I'll answer your questions first then describe my model.
- How have you judged convergence of your model?
yes I checked convergence.

Can you clarify how you determined your model is converged? This may be relevant.

Is temperature constant in your experiment? If not, it may be good to try simulating energy. And are your material properties consistent with the temperatures of your experiment?

Other than this, I could suggest trying a different turbulence model to see if this changes anything.