CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   inlet velocity profiles (http://www.cfd-online.com/Forums/cfx/19120-inlet-velocity-profiles.html)

cfxuser August 13, 2002 14:50

inlet velocity profiles
 
hi,

I want to calculate a swirling pipe flow with a specified inlet velocity profile of the type

U_ax=f(r)

W_tan=f(r)

The profile is not an expression but stored in a file in cylindrical coordinates:

r_1 u_1 w_1

r_2 u_2 w_2 etc.

The geometry is set up so that that the x-axis is the axis of rotation. So the radius would be calculated as r=sqrt(y^2 + z^2). Now my problem is:

Is it sufficient to define a new variable "rad" with CEL so that the solver "knows" that the tabulated values are cylindrical ?

Thanks for your comments

Neale August 14, 2002 17:07

Re: inlet velocity profiles
 
You will need to define u, v and w velocity components from your radial and axial components. This requires transforming your cylindrical components into Cartesian in CEL. You could do this with 1D interpolation functions and the following:


CEL:
EXPRESSIONS:
r = sqrt(x^2+y^2)
theta = atan(y,x)
Vrad = Vradf(r)
Vaxial = Vaxialf(r)
Vtheta = Vthetaf(r)
Ucomp = Vaxial
Vcomp = Vradial*cos(theta) + r*Vtheta*sin(theta)
Wcomp = Vradial*sin(theta) - r*Vtheta*cos(theta)
END
END


where Vradf, Vaxialf, and Vthetaf are 1D interpolation functions for each of your velocity components. Looks like you might need to split your text files up. Hopefully I got my sin() and cos() correct!

cfxuser August 17, 2002 06:18

Re: inlet velocity profiles
 
Neale,

thanks a lot for your help! Unfortunately I still get the following error message:

Error processing expression: U = Ucomp

Obviously something is wrong with the boundary condition definition. As you suggested I defined the CEL expressions, declared the functions and defined the junction box routines (which I derived from the CFX documentation for user subroutines). Maybe you can see the mistake in the CCL file below ... ?

Best regards and many thanks in advance.

1. EXPRESSIONS:

(..)

rad = sqrt(z^2+y^2)

phi = atan(y,z)

Vrad = Vradf(rad)

Vaxial = Vaxialf(rad)

Vtang = Vtangf(rad)

Ucomp = Vaxial

Vcomp = Vrad*cos(phi) + rad*Vtang*sin(phi)

Wcomp = Vrad*sin(phi) - rad*Vtang*cos(phi)

(..)

2. FUNCTIONS:

(..)

FUNCTION : Vaxial

Option = User Function

Argument List = [m]

Result Units = [m s^-1]

(..)

FUNCTION : Ucomp

Option = User Function

Argument List = [m s^-1]

Result Units = [m s^-1]

(..)

3. USER ROUTINE DEFINITIONS

(..)

USER ROUTINE : Load Inlet Profiles

Option = Junction Box Routine

Junction Box Location = User Input

Calling Name = user_input_rad

Library Name = user_input_rad

Library Path = /home/cfxuser/Testcase

(..)

USER ROUTINE : Vaxial

Option = Junction Box Routine

Junction Box Location = User Input

Calling Name = vaxialf

Library Name = user_input_rad

Library Path = /home/cfxuser/Testcase

(..)

3. BOUNDARY CONDITIONS :

(..)

MASS AND MOMENTUM :

Option = Cartesian Velocity Components

U = Ucomp

V = Vcomp

W = Wcomp

END (..)

Neale August 26, 2002 14:13

Re: inlet velocity profiles
 
From this I can't tell. Is there any more output in the output file which is more descriptive than what you have?

One thing you should be able to do is set Ucomp = Vaxiaf(rad) directly, rather than through another expression variable.

I'm also not sure why you have created a user function called "Ucomp". This should not be necessary. All you should need is to define three user functions: Vradialf, Vaxialf and Vthetaf and set Ucomp, Vcomp and Wcomp accordingly. Maybe this is/was the problem?

Neale


All times are GMT -4. The time now is 03:09.