Hi, CFX users

I am trying to set boundary conditions of 'fully developed flow' at pipe inlet without preliminary calculation. How can I set the BC's. Thanks.

Hi,

you can set a mass flow boundary condition at the inlet with negative mass flux. The procedure is explained in the CFX4.x solver manual.

Thank you for your response. But, I still could not set a mass flow boundary condition at the inlet. I read the CFX4.2 solver manual carefully and tried to set a flow boundary condition with very simple one block pipe. Then I got a error message. I am a CFX beginner. So I do not know what is my mistake. Would you please explain the my mistake ?

>>CFX4
:>OPTIONS

THREE DIMENSIONS

BODY FITTED GRID

CARTESIAN COORDINATES

TURBULENT FLOW

ISOTHERMAL FLOW

INCOMPRESSIBLE FLOW

STEADY STATE
:>MODEL TOPOLOGY
:>CREATE PATCH

PATCH NAME 'ENTRANCE'

BLOCK NAME 'SOLID 1'

PATCH TYPE 'MASS FLOW BOUNDARY'

LOW K
:>CREATE PATCH

PATCH NAME 'OUTLET'

BLOCK NAME 'SOLID 1'

PATCH TYPE 'PRESSURE BOUNDARY'

HIGH K
:>MODEL DATA
:>PHYSICAL PROPERTIES

>>FLUID PARAMETERS

VISCOSITY 1.0000E-03

DENSITY 9.9800E+02
:>SOLVER DATA
:>PROGRAM CONTROL

MAXIMUM NUMBER OF ITERATIONS 200

MASS SOURCE TOLERANCE 1.0000E-06
:>MODEL BOUNDARY CONDITIONS
:>MASS FLOW BOUNDARIES

>>FLUX

MASS FLOW SPECIFIED

FLUXES -1.000000E+01
:>STOP

CFX-4.2 started on Mon Dec 18 17:32:03 KST 2000

This job was run on : kngr The command line was: /CFX4.2/bin/runsolve4 -rel 4.2 -dir /TH2/th/hjukim/CFX4/o ri/con/geo/F1/tmp1/.cfx4_tmpdir -enviro 08 -c m08.fc -geom m01.geo -dynam

>>CFX4

>>OPTIONS

THREE DIMENSIONS

BODY FITTED GRID

CARTESIAN COORDINATES

TURBULENT FLOW

ISOTHERMAL FLOW

INCOMPRESSIBLE FLOW

STEADY STATE
:>MODEL TOPOLOGY

>>CREATE PATCH

PATCH NAME 'ENTRANCE'

BLOCK NAME 'SOLID 1'

PATCH TYPE 'MASS FLOW BOUNDARY'

LOW K

>>CREATE PATCH

PATCH NAME 'OUTLET'

BLOCK NAME 'SOLID 1'

PATCH TYPE 'PRESSURE BOUNDARY'

HIGH K
:>MODEL DATA

>>PHYSICAL PROPERTIES

>>FLUID PARAMETERS

VISCOSITY 1.0000E-03

DENSITY 9.9800E+02
:>SOLVER DATA

>>PROGRAM CONTROL

MAXIMUM NUMBER OF ITERATIONS 200

MASS SOURCE TOLERANCE 1.0000E-06
:>MODEL BOUNDARY CONDITIONS

PERMANENT WORKSPACE ALLOCATED

TYPE TOTAL SET TOTAL USED REAL 2716160 814827 INTEGER 1358080 484175 CHARACTER 1049 371

>>MASS FLOW BOUNDARIES

>>FLUX

LINEAR SOLVER - BLOCK DEPENDENCY INFORMATION:

PERCENTAGE OF WELL ORDERED BLOCKS 100.0000 ORDERING USED

1

####> ERROR. THE >>STOP COMMAND HAS NOT BEEN READ THE COMMANDS MAY BE IN THE WRONG ORDER OR INCORRECTLY TYPED PLEASE CORRECT AND RE-SUBMIT THE JOB THE CFX-4 SOLVER HAS BEEN STOPPED BY SUBROUTINE OUTOPT

CFX-4.2 finished on Mon Dec 18 17:32:17 KST 2000

Hi,

if you are using the 4.2 version, first of all there is a bug, or a spelling mistake in the manual whatever you want. The keyword is "FLUXES" and not "FLUX" (this was fixed in the 4.3 version):
:>MASS FLOW BOUNDARIES

>>FLUXES

MASS FLOW SPECIFIED

FLUXES -1.0E+1 1.0E+1

Secondly you have to define group numbers for the inlet and outlet:
:>MODEL TOPOLOGY
:>MODIFY PATCH

OLD PATCH NAME 'ENTRANCE'

NEW PATCH NAME 'ENTRANCE'

NEW PATCH TYPE 'MASS FLOW BOUNDARY'

NEW PATCH GROUP NUMBER 1

OLD PATCH NAME 'OUTLET'

NEW PATCH NAME 'OUTLET'

NEW PATCH TYPE 'MASS FLOW BOUNDARY'

NEW PATCH GROUP NUMBER 2

In the boundary condition set up the first group refers to -1.0 , the second to 1.0 and so on if you have more.

Hope this helps Fabian

I'm mot quite sure anymore but it is possible to have only a MASS FLOW BOUNDARY and a PRESSURE BOUNDARY ? This combination looks a little strange to me.

Marat

For this question,if the inlet is a mass flux, the converenge is difficult.this is also not the mean of :fully developed flow inlet.; so it may be good to use inlet BC. then you only need keep the momentum as a constant. that is Ok

[CFD_beginner12]

Hi all,
Is there any updates in newer versions of Ansys to solve this issue ?

[ghorrocks]

Those original posts date from over 20 years ago, and back then the current version of CFX was CFX-4. It was a completely different solver and has little in common with the current solver. So this post is not relevant to the current solver.

There are several ways of getting a better approximation of fully developed flow at a boundary condition. The easiest is to use a pressure boundary.

[CFD_beginner12]

Originally Posted by ghorrocks

Those original posts date from over 20 years ago, and back then the current version of CFX was CFX-4. It was a completely different solver and has little in common with the current solver. So this post is not relevant to the current solver.

There are several ways of getting a better approximation of fully developed flow at a boundary condition. The easiest is to use a pressure boundary.

Thank you for explaining , "20 years !!!" so sorry but the search engine bring me here and some other results, the date of post was the last thing in my interest, I ll try to search again about pressure boundary.