Learning Forum

[Dio]

Hi,

I am doing simple smooth pipe to get Nusselt number.

Defined periodic BC at fluent meshing side, and added 1e-3 uniform inflation.

When I switch to fluent, defined periodic bc and it worked. result was higher than dittus-boelter corelation

Then wantted to decrease inflation first layer to 3.1e-6 using y+ calculator.

But that time in fluent side periodic bc does not work and got below error.

What can be the reason?

 

Thank you,

[Dio]

I tried 2D case and got similar error.

 

Thank you

[Rob]

That looks OK. You can always create a nonconformal periodic, that generally avoids problems. 

I assume you checked the periodic energy equation fudge and how Fluent calculates Nu?

[Dio]

 

Hi,

Thank you Rob. Now trying it but why I could not see faces? Just did mesh using fluent meshing, also defined periodic faces in fluent meshing.

 

Thank you

 

[Rob]

Define as "interface" & "interface_1" or "fred" & "ginger" but with the type as interface. 

[Dio]

Just got an error. (Yes I defined values in reference values to calculate Nu ✔️)

Here is how I did.

1-Changed it to interface

2-Defined like this.

Then defined boundary conditions, mass flow rate and heat flux and started analysis. Then got floating error.

What am I missing?

 

 

 

[Rob]

Nice to see the classic boundary name pairing. :)  

It's running so the basics are OK. If you review the flowfield at around 20 iterations what is going on? 

[Dio]

Yea, that is new info for me :) 

I just changed solution method from SIMPLEC to COUPLE  and it worked. No error.

It looks some error and I will play with mesh, and turbulence method now.

 

Thank you

[Rob]

Check things like area values in Fluent alongside what the experimental values & errors are.  I won't say the experiment is wrong, nor the CFD, but don't assume either is "correct" either.