A little bug in the tutorial case "turbChannel"

[X. G. Zhou]

Dear community.

Maybe there was a little bug in tutorial case "turbChannel" when post-processing the streamwise turbulent heat flux $\overline{u'T'}$.

We should revise the code in subroutine userchk of file "turbChannel.usr", as 

Line 262     &           (stat_y(7*nint+i)-(stat_y(0*nint+i)*
Line 263     &                              stat_y(5*nint+i)))/qw,

Since the time-averaged values of streamwise velocity and temperature are both not zero in this case. Therefore, we need to substract the product of their time-averaged values when post-processing the streamwise turbulent heat flux. 

I tested the above code using a simple very-coarse grid case with Nek5000.  The trend and numerical value of streamwise turbulent heat flux (sub-figure b) and wall-normal turbulent heat flux (sub-figure c) are basically consistent with the results of Kawamura Lab and OpenFOAM.

I hope this post appears not to late.  Please criticize and correct. 

Best,

Student Xingguang Zhou.

Xi'an Jiaotong University

2024/12/03