 |
 |
Troops: I reviewed many graphic presentations, some in color, and associated discussions, etc., for Lab 10 report, but all except two are FATALLY flawed! The second objective clearly states, "Adjust the diffuser length to ensure that the outflow BC specification (of vanishing normal derivative) is appropriate for each Re." This lab objective emphasizes the lecture-covered fundamental CFD adage, "You get out what you put in, especially on the boundaries!"
Not moving the outflow BC further downstream (extend solution domain) generates a trash solution for Re = 600, and compromises somewhat the Re=500 solution. The end result is you dutifully report the L/s error at Re=600 was relatively smaller. You have been instructed throughout your BS degree program to plot your data and look at it for consistency. Only some reports have this graphic – if done you would clearly see how your data differ from that in text Figure CMn.10.
Others conjectured about experimental data inaccuracy to explain the differences. It’s obvious you did not read the text which clearly states the cause for gradual disagreement between CFD and experiment is that the step wall flowfield is becoming fully 3D at the larger Re, hence a 2D solution is no longer addressing the real problem.
Finally, there is nothing
“turbulent” about these flows at such low Re! Benefit from this learning
experience! The archive will
be opened shortly.
AJB