Learning Forum

[Navid Goudarzi]

When we run Fluent and try using the expression for the turbulence intensity, we keep getting this error: "expression is not single-valued". The expression is correct. Any tips?

[Rob]

It may be correct but is it single valued? 

[SRP]

Hi,

You are getting this error because given expression calculates the values at all cells.

Thank you.

[Navid Goudarzi]

Thank you for the response.

It is not single-valued. It is the TI value as a function of "y" coordinates. 

The TI expression is defined on the inlet. We do not get a similar error for the inlet velocity expression. Any thoughts?

[Rob]

If you work out the values using Excel or the like what does the profile look like? 

[Rob]

Having checked it looks like you can only have a single value for Turbulent Intensity. 

[Navid Goudarzi]

Rob,

The profile follows a power law polynomial (similar to the velocity profile with a different power). To validate the CFD against experimental results, we need to insert TI values as expressions rather than single values as there is a significant change with "y" coordinates, changing from 40% to less than 5%. Is there any other way that we can apply it? 

e.g., should we change the inlet condition from "turbulent intensity and turbulent viscosity ratio" to "turbulent kinetic energy and dissipation rate" and then define expressions for them? 

[Rob]

For profiles we usually use k and epsilon/omega depending on the model(s) used. I've asked internally whether there's something we're missing for TI, but may not be able to feed much back depending on the answer: it may just be "don't". 

[Navid Goudarzi]

Thank you! We will try the K-epsilon (we still need to define an expression for it) and I look forward to hearing back from you. 

[Rob]

How were you creating the TI profile expression? 

[Navid Goudarzi]

It seems that TI might be able to be defined as a UDF rather than an expression: Enter expression for Turbulent intensity or turbulent kinetic energy (ansys.com) would you confirm? (we did not get the ANSYS reply on this trend where it says the "C" value should be positive as the equation will be defined based on the experimental reference case --- please see our equation below where the "C" value is negative).

Here is the TI profile expression: 0.2*(y/0.28 [m])**-0.3594

 

[Rob]

We're not sure why the Expression isn't working, but a UDF DEFINE_PROFILE should be fine. 

The profile must return a positive value, and the "y" position is against the inlet location in the CFD model. Not sure if Aitor is on the new Forum, but we tended to correct their replies on the very rare occasion there was anything missing: our goal is to moderate rather than answer everything but Community is still quite young.