Flow Result Accuracy
 

I was performing turbulent flow modeling with Fluent on an untapered rectangular wing with an aspect ratio of 6.6, at an Re of approximately 1e6. The airfoil is NACA0015. I used the S-A turbulence model with low-Re correction.

I performed validation using a VLM solver (AVL) for lift, moment and induced drag, and the fully turbulent flat plate solution for skin friction drag. As can be seen from the plots, the CL and CD agree pretty well. How come the CM is more than 3 times off?

Cm is always tricky, and the hardest, to compare.

One of the better metrics is to calculate the error in the center of pressure or aerodynamic center, in your case, dxac = (cm1-cm2)/cl = (3.25e-2-1.0e-2)/1.0 = 2.25e-2. So your aerodynamic center is off by 2.25 percent of the chord. (Assuming your reference length is the chord length) This is to be expected.