Hello, Paul, Thank you for your answer for the same problem last month. But I still could not figure out this issue. (Domain:6.4/2.0/3.2, Resolution: 12*8*12 elements with 8th order, turbulent Re_tau=178.12, Pr=0.71) , piece of my *.rea file ***************************************************************************** 1.00000 DENSITY -178.12 Re_b 0.000000E+00 BETAG 0.00000 GTHETA 0.000000E+00 PGRADX 0.000000E+00 FLOWRATE 1.0 RHOCP -126.47 CONDUCT ***************************************************************************** my *usr file ***************************************************************************** subroutine userq (ix,iy,iz,ieg) include 'SIZE' include 'TOTAL' include 'NEKUSE' ie = gllel(ieg) qvol = vx(ix,iy,iz,ie)/15.771 source = 0.0 return end ***************************************************************************** Here 15.71 is averaged mean velocity over channel section. After a long time running (dimensionless time > 350 ) I compared the result , the velocity profile looks good. But the temperature is much higher than Kasagi's result. I'm wondering whether the parameters above are correct ? I attached the plot of mean velocity and mean temperature. Thank you again -- Zhicheng Wang