Validating and verifying CFD simulations

Shawnk · February 1, 2018, 22:22

Hi,
I am doing a CFD simulation of a flow system in a very unusual geometry. This means, not much is understood about the hydrodynamics of the system and there is limited (if any) experimental data available to validate the CFD simulation.

Besides experimental data, what other ways can one proof that a CF result is valid? Also, if you know any, kindly suggest papers or literature you know that have done CFD simulations of not-too-common systems and have validated them with minimal experimental data but also with some more substantial proof.

Thanks

piu58 · February 2, 2018, 00:31

AT least you may perform simulation with the same physics, but simpler geometry (for which you have reference data).

If you use the same principles in simulating your case, the deviation should be in the same range. There is always an additional problem of course, the problem with meshing complicated geometries. To overcome this, try different meshing strategies and prove, that the mesh doesn't introduce a large amount of deviation.

Shawnk · February 2, 2018, 01:34

Quote:

Originally Posted by piu58

AT least you may perform simulation with the same physics, but simpler geometry (for which you have reference data).

If you use the same principles in simulating your case, the deviation should be in the same range. There is always an addition problem of course, the problem with meshing complicated geometries. To overcome this, try different meshing strategies and prove, that the mesh doesn't introduce a large amount of deviation.

Thanks for your reply. Is there a literature you would recommend ?

piu58 · February 2, 2018, 01:55

It depends on your problem which I don't know.

Well understood / measured systems are flow over a plate and past a cylinder or sphere. You find tons of literature for this. Measuring values you may obtain from the old books of Ludwig Prandtl or Hermann Schlichting.

violethill · February 2, 2018, 02:04

Can you give us a little bit more information about your geometry? And what you are trying to measure? Without a problem statement we cannot help you other than give you general suggestions which would not be much of a help. In that way, perhaps we can also help give suggestion on the experimental part.

Shawnk · February 2, 2018, 02:04

Quote:

Originally Posted by piu58

It depends on your problem which I don't know.

Well understood / measured systems are flow over a plate and past a cylinder or sphere. You find tons of literature for this. Measuring values you may obtain from the old books of Ludwig Prandtl or Hermann Schlichting.

Thanks again for your response

Shawnk · February 2, 2018, 02:20

Quote:

Originally Posted by violethill

Can you give us a little bit more information about your geometry? And what you are trying to measure? Without a problem statement we cannot help you other than give you general suggestions which would not be much of a help. In that way, perhaps we can also help give suggestion on the experimental part.

A typical example is to consider the wheel of a car. Instead of it being filled with high pressure air, it is filled with both water and air in equal proportion. When the wheel rotates, what happens to the hydrodynamic of the water inside. I have already considered torque measurement and prediction. What other ways can one validate the water hydrodynamics?

piu58 · February 2, 2018, 02:33

You have a two pahse problem, air and water. For this the easiest point to start is the ascending bubble. You don't have external forces beyond gravity there, but it is the first example I would try to solve. There is an excellent report by Hysing about this.

February 1, 2018, 22:22	Validating and verifying CFD simulations	#1
Shawnk New Member Oluwaseun Adedeji Join Date: Dec 2016 Location: Edmonton, Alberta Posts: 6 Rep Power: 9	Hi, I am doing a CFD simulation of a flow system in a very unusual geometry. This means, not much is understood about the hydrodynamics of the system and there is limited (if any) experimental data available to validate the CFD simulation. Besides experimental data, what other ways can one proof that a CF result is valid? Also, if you know any, kindly suggest papers or literature you know that have done CFD simulations of not-too-common systems and have validated them with minimal experimental data but also with some more substantial proof. Thanks

February 2, 2018, 00:31		#2
piu58 Senior Member Uwe Pilz Join Date: Feb 2017 Location: Leipzig, Germany Posts: 744 Rep Power: 15	AT least you may perform simulation with the same physics, but simpler geometry (for which you have reference data). If you use the same principles in simulating your case, the deviation should be in the same range. There is always an additional problem of course, the problem with meshing complicated geometries. To overcome this, try different meshing strategies and prove, that the mesh doesn't introduce a large amount of deviation. __________________ Uwe Pilz -- Die der Hauptbewegung überlagerte Schwankungsbewegung ist in ihren Einzelheiten so hoffnungslos kompliziert, daß ihre theoretische Berechnung aussichtslos erscheint. (Hermann Schlichting, 1950) Last edited by piu58; February 2, 2018 at 01:53.

February 2, 2018, 01:55		#4
piu58 Senior Member Uwe Pilz Join Date: Feb 2017 Location: Leipzig, Germany Posts: 744 Rep Power: 15	It depends on your problem which I don't know. Well understood / measured systems are flow over a plate and past a cylinder or sphere. You find tons of literature for this. Measuring values you may obtain from the old books of Ludwig Prandtl or Hermann Schlichting. __________________ Uwe Pilz -- Die der Hauptbewegung überlagerte Schwankungsbewegung ist in ihren Einzelheiten so hoffnungslos kompliziert, daß ihre theoretische Berechnung aussichtslos erscheint. (Hermann Schlichting, 1950)

February 2, 2018, 02:33		#8
piu58 Senior Member Uwe Pilz Join Date: Feb 2017 Location: Leipzig, Germany Posts: 744 Rep Power: 15	You have a two pahse problem, air and water. For this the easiest point to start is the ascending bubble. You don't have external forces beyond gravity there, but it is the first example I would try to solve. There is an excellent report by Hysing about this. Shawnk likes this. __________________ Uwe Pilz -- Die der Hauptbewegung überlagerte Schwankungsbewegung ist in ihren Einzelheiten so hoffnungslos kompliziert, daß ihre theoretische Berechnung aussichtslos erscheint. (Hermann Schlichting, 1950)

February 2, 2018, 02:04		#5
violethill New Member Violet Join Date: Jun 2016 Posts: 8 Rep Power: 9	Can you give us a little bit more information about your geometry? And what you are trying to measure? Without a problem statement we cannot help you other than give you general suggestions which would not be much of a help. In that way, perhaps we can also help give suggestion on the experimental part.