[leasken]

Hello, I am trying to define the hydraulic diameter for a channel flow by using an expression in CFX-Pre but cannot figure out how to properly implement it. Tried with

maxVal(y)@inlet - minVal(y)@inlet

but it resulted in the following error message:

ERROR
Attempt to evaluate the CEL callback function 'maxVal'.
Although it is valid to create an expression that
uses such a function, the preprocessor does not support
its evaluation.

I want to use a general expression and utilize the domain in CFX-Pre somehow. Any suggestions?

[Lance]

Originally Posted by leasken

the preprocessor does not support its evaluation.

Have you tried running the simulation anyway? Even though cfx-pre cannot evaluate the expression, the solver might be able to do it.

[leasken]

Originally Posted by Lance

Have you tried running the simulation anyway? Even though cfx-pre cannot evaluate the expression, the solver might be able to do it.

I cannot start the solver with this expression, CFX-Pre tells me that

The function 'maxVal' referenced by parameter 'Normal Speed' in object '/FLOW:Flow Analysis 1/DOMAIN: Default Domain/BOUNDARY:inlet/BOUNDARY CONDITIONS/MASS AND MOMENTUM' has a disallowed argument, 'y'. The solver does not support the use of this variable in this context.

[honcho62]

I think your problem is that you're requesting the maximum value of a singular scalar. So it's kinda pointless in the sense that y is both its min and max value. You're better off defining a line along the vertical edge and then saying maxValue(velocity u)@line for instance.

[ghorrocks]

honcho: I do not think that is correct. The original expression is asking for the max and min values of the y nodal locations on the boundary patch labelled inlet. This sounds valid to me. I am confused about the problems Martin is reporting with his expression as I have used very similar expressions to this (max/min values of xyz nodal values on boundary patches) before and have not had any problems with it.

[leasken]

Originally Posted by ghorrocks

honcho: I do not think that is correct. The original expression is asking for the max and min values of the y nodal locations on the boundary patch labelled inlet. This sounds valid to me. I am confused about the problems Martin is reporting with his expression as I have used very similar expressions to this (max/min values of xyz nodal values on boundary patches) before and have not had any problems with it.

Yes, I have also used a similar expression for calculating the length of an Iso Clip in CFD-Post. Thought that I could use the same here, but nope..

[mrshives]

I've had a similar, frustrating problem. It's a hack solution... but you could try creating an additional variable called "myY" (set to equal "y" using an expression) then use "myY" in your minVal and maxVal functions. ... sometimes cfx just behaves badly....

Also don't worry about the pre-processor not supporting the function... that's just a warning, the real question is whether the solver supports that function (and I think the answer is yes)

[leasken]

Originally Posted by mrshives

I've had a similar, frustrating problem. It's a hack solution... but you could try creating an additional variable called "myY" (set to equal "y" using an expression) then use "myY" in your minVal and maxVal functions. ... sometimes cfx just behaves badly....

Also don't worry about the pre-processor not supporting the function... that's just a warning, the real question is whether the solver supports that function (and I think the answer is yes)

No progress.. The warning issued stops CFX-Pre from creating a .def-file and no simulation can then be run.

I tried a bit with your hack solution, but not sure on how to define the "myY"-expression, any suggestions? Now I tried creating the expression "myY" and defining it as y, and also something like y@inlet, but nothing..

[leasken]

So I tried to solve this problem using parameters. I created an expression "Dh" and assigned a value to it. I ran a simulation and in CFD-post I selected "Use as Workbench Input Parameter". In the Parameter Set I changed the definition of Dh from the value "0.005 [m]" to an expression using two other input parameters, so basically

P1 H 5
P2 W 5
P3 Dh 2*(P1*P2)/(P16+P18)

But then there's no unit assigned to P3 and Workbench complains about it. P1 and P2 are parameters from my CAD-file, where I have constructed H and W as expressions in the CAD software, imported the geometry to Workbench and enabled the parameters under CAD-parameters in the Mesh module. When P1 and P2 are enabled and appear in my Parameter Set, there is no unit assigned to them so P3 is dimensionless. When I just try to add [mm] after P1 and P2 in P3:

P3 Dh 2*(P1[mm]*P2[mm])/(P16[mm]+P18[mm])

Workbench says "Name 'P1_Q' is not defined." in both the Value and Unit box for P3. Rather frustrating. What am I missing or doing wrong?

[leasken]

Originally Posted by leasken

So I tried to solve this problem using parameters. I created an expression "Dh" and assigned a value to it. I ran a simulation and in CFD-post I selected "Use as Workbench Input Parameter". In the Parameter Set I changed the definition of Dh from the value "0.005 [m]" to an expression using two other input parameters, so basically

P1 H 5
P2 W 5
P3 Dh 2*(P1*P2)/(P16+P18)

But then there's no unit assigned to P3 and Workbench complains about it. P1 and P2 are parameters from my CAD-file, where I have constructed H and W as expressions in the CAD software, imported the geometry to Workbench and enabled the parameters under CAD-parameters in the Mesh module. When P1 and P2 are enabled and appear in my Parameter Set, there is no unit assigned to them so P3 is dimensionless. When I just try to add [mm] after P1 and P2 in P3:

P3 Dh 2*(P1[mm]*P2[mm])/(P16[mm]+P18[mm])

Workbench says "Name 'P1_Q' is not defined." in both the Value and Unit box for P3. Rather frustrating. What am I missing or doing wrong?

Fixed it... Multiplied by 1 [mm].

P3 Dh (2*(P16*P18)/(P16+P18))*1[mm].