Q
I would like to define a problem with static pressures at the inlet and outlet, but the Help Manual suggests that this does not appear to a sensible thing to do.

A
The advice in the manual is sound; setting static pressures at the inlet and outlet may well give convergence problems. You may however find that it is possible to converge such a set-up, in which case there is nothing physically wrong with the model.

If you can specify the inlet condition in terms of a total pressure (i.e. static pressure + rho.U^2/2) then convergence behaviour is likely to be much better.

A useful trick may be to use a total pressure inlet, but set the inlet total pressure using CFX Expression Language (CEL) to be your required static pressure plus rho.U^2/2. Please note that this definition of total pressure is valid only for fluids of constant density, e.g. liquids and low-speed gas flows.

Below is the CCL/CEL that is needed to do this:

LIBRARY:
CEL:
EXPRESSIONS:
pstat in = 5 [Pa]
vel in = massFlow()@INLET /
(Water.density * area()@INLET)
ptot in = pstat in + 0.5*Water.density*vel in^2
END
END
END

FLOW:
DOMAIN: Domain 1
BOUNDARY: INLET
Boundary Type = INLET
Location = INLET
BOUNDARY CONDITIONS:
FLOW DIRECTION:
Option = Normal to Boundary Condition
ENDFLOW REGIME:
Option = Subsonic
END
MASS AND MOMENTUM:
Option = Total Pressure
Relative Pressure = ptot in
END
TURBULENCE:
Option = Medium Intensity and Eddy Viscosity Ratio
END
END
END
END
END

In this case pstat in is the required inlet static pressure, which the user should specify. vel in is the inlet velocity, calculated as part of the solution from the current mass flow at the inlet, the inlet area and the density. Finally, the inlet total pressure, ptot in, is calculated from the sum of the specified inlet static pressure an


Q
I would like to define a problem with static pressures at the inlet and outlet, but the Help Manual suggests that this does not appear to a sensible thing to do.

A
The advice in the manual is sound; setting static pressures at the inlet and outlet may well give convergence problems. You may however find that it is possible to converge such a set-up, in which case there is nothing physically wrong with the model.

If you can specify the inlet condition in terms of a total pressure (i.e. static pressure + rho.U^2/2) then convergence behaviour is likely to be much better.

A useful trick may be to use a total pressure inlet, but set the inlet total pressure using CFX Expression Language (CEL) to be your required static pressure plus rho.U^2/2. Please note that this definition of total pressure is valid only for fluids of constant density, e.g. liquids and low-speed gas flows.

Below is the CCL/CEL that is needed to do this:

LIBRARY:
CEL:
EXPRESSIONS:
pstat in = 5 [Pa]
vel in = massFlow()@INLET /
(Water.density * area()@INLET)
ptot in = pstat in + 0.5*Water.density*vel in^2
END
END
END

FLOW:
DOMAIN: Domain 1
BOUNDARY: INLET
Boundary Type = INLET
Location = INLET
BOUNDARY CONDITIONS:
FLOW DIRECTION:
Option = Normal to Boundary Condition
ENDFLOW REGIME:
Option = Subsonic
END
MASS AND MOMENTUM:
Option = Total Pressure
Relative Pressure = ptot in
END
TURBULENCE:
Option = Medium Intensity and Eddy Viscosity Ratio
END
END
END
END
END

In this case pstat in is the required inlet static pressure, which the user should specify. vel in is the inlet velocity, calculated as part of the solution from the current mass flow at the inlet, the inlet area and the density. Finally, the inlet total pressure, ptot in, is calculated from the sum of the specified inlet static pressure and the dynamic pressure, rho.U^2/2.
**** Entered By: dsclarke @ 06/01/2006 16:58 ****





Show Form
No comments yet. Be the first to add a comment!