[hussein93]

Hey, maybe someone can explain me that. I am investigating a cascade flow in a turbine with very high pressure fluctuations. At the inlet domain I have a unsteady boundary condition.Now I am looking for a boundary condition for my outlet. I don't quite understand the following: I have a certain outlet area, i.e. the distance from the blade to the outlet. I extend this outlet range for example by 2 times. Then I start a simulation with a fixed boundary condition at the new outlet. If the flow now passes the first outlet but has not yet reached the second outlet, I extract the time-dependent condition there during the simulation and this extracted condition is then my outlet boundary condition at the desired location in the simulations.

greetings

[FMDenaro]

Originally Posted by hussein93

Hey, maybe someone can explain me that. I am investigating a cascade flow in a turbine with very high pressure fluctuations. At the inlet domain I have a unsteady boundary condition.Now I am looking for a boundary condition for my outlet. I don't quite understand the following: I have a certain outlet area, i.e. the distance from the blade to the outlet. I extend this outlet range for example by 2 times. Then I start a simulation with a fixed boundary condition at the new outlet. If the flow now passes the first outlet but has not yet reached the second outlet, I extract the time-dependent condition there during the simulation and this extracted condition is then my outlet boundary condition at the desired location in the simulations.

greetings

First of all, you have to be sure about the proper BC at the outlet. If the flow is supersonic, no conditions enter into the solution, while if is subsonic you have to fix one condition at the outlet.



But, assuming a subsonic outlet, the main issue I see in your idea is that you are superimposing a sort of spatial periodicity that is arbitrary.
Simple convective conditions allow you to get unsteady profiles at the outlet.

[LuckyTran]

To explain what is likely going on with the turbine cascade simulation:


A naive approach to doing CFD is to use a constant, time-independent fixed pressure outlet boundary condition. This particular researcher is trying to do better by using an unsteady and non-uniform pressure outlet. They obtain this time-dependent temperature profile by running another simulation that has an extended outlet condition. Moving the constant pressure BC further downstream allows them to extract the unsteady pressure profile and they then go back to apply it as the outlet BC to the original simulation.

[FMDenaro]

Originally Posted by LuckyTran

To explain what is likely going on with the turbine cascade simulation:


A naive approach to doing CFD is to use a constant, time-independent fixed pressure outlet boundary condition. This particular researcher is trying to do better by using an unsteady and non-uniform pressure outlet. They obtain this time-dependent temperature profile by running another simulation that has an extended outlet condition. Moving the constant pressure BC further downstream allows them to extract the unsteady pressure profile and they then go back to apply it as the outlet BC to the original simulation.

Since the idea requires to do a simulation with a larger domain I don't see the advantage, just use the case where the outlet is far from blade.
Furtermore, I don't understand this sentence:


"If the flow now passes the first outlet but has not yet reached the second outlet, I extract the time-dependent condition there during the simulation"


That would be true only in case of supersonic condition in that region. Otherwise, the first outlet still is coupled with the second outlet by the characteristic entering in the domain.

[LuckyTran]

My interpretation of what they are trying to do is to generate a surrogate pressure outlet BC since a turbine cascade rig is usually a single stage with with one set of vanes and one set of blades yet trying to mimic a multi-stage turbine. The argument is that the outlet of the cascade stage is in practice the inlet to another turbine stage. Applying a constant pressure would overly constrain the wakes and that is why you would rather extract something that is unsteady. However, I also don't follow what is the meaning of extracting it for only a certain flow-time and not all-time. Maybe it is related to the inlet forcing that is not being described, I don't know.