Hi, I'm starting to use CFX and I have one question:what is the best turbolence model to resolve a problem with very small way (0,5mm) and high pressure drop (50 Bar)? And you have some suggestion for the mesh?

thank you

Hi,

There are many issues to consider:

Do you need to resolve the boundary layer?

Do you need to resolve the large scale turbulent structures?

Steady state or transient?

Does the flow have a significant swirling component?

What Reynolds number is the flow?

Glenn Horrocks

Hi,

the most important information that i need to resolve are the bondary layer so what is the best wall approach?

the objects that i'm studing are small in order of 10 mm of diameter and not long but in some case the way of flow is very small (0,5-1mm),

the flow is steady state,

liquid is oil for oleodinamics and i think there isn't a significant swirling component.

sorry for my english and thank you for your help

Hi,

Of course you need to model the effects of the boundary layer, but do you need to resolve it right down to the viscous sublayer or is putting the first node in the log layer region OK by using a wall function?

What is the Reynolds number of the flow?

Most likely the SST turbulence model will do, it is usually a good one for general purpose stuff.

Glenn Horrocks

Hi,

the Re number is:3.1965E+05 and i have a problem with a big pressure drop in a small way. i have to study forces (pressures) that are on the wall near pressure drop but the results show me values of pressure in the nearest cell before pressure drop like: -1.20e+007 that are physics impossible. i need cavitation opition? i have used K-e equation with a boundary condition of INLET (pressure) OUTLET (Pressure). Is K-e the right equation?

Thanks

Hi,

Assuming your Reynolds number is based on flow velocity and an important flow diameter then it looks like you are in the mid turbulent region. For this region either k-epsilon or SST should work pretty well.

The flow will cavitate if the pressure you report occurs. Adding a cavitation model will complicate the simulation considerably. What is the physical volume of the low pressure region? If it is small you may be able to ignore it and not use a cavitation model. If the cavitation volume is significant then you have no option but to model it.

Regards, Glenn

Hi,

the volume with cavitation problem is small, but this wery low pressure is on the wall that i have to study and i'm worried that the forces results are not comparable between several configuration that study (every configuration have the same mesh and boundary configuration but small geometrical difference).

Thanks a lot.