Dear Ruiliang,
As described in this forum topic: Learizing Baseline 5MW Wind Turbine with FAST, rigid-body modes show up in MBC3 as a pair of zero-valued (or near-zero-valued) frequencies with +/- inf damping (i.e., eigenvalues with real values only). That is, each rigid-body mode will introduce an additional mode beyond the number of enabled DOFs.
When the rotor is spinning, MBC3 changes the description from individual blade DOFs to rotor DOFs (collective, cosine, sine), as described in my May 16, 2011 post in the forum topic found here: Eigenanalysis FAST. When the rotor is spinning, each rotor mode is likely to have a different natural frequency, which are also different from the corresponding mode of the isolated, nonspinning blade. I suggest that you first linearize a mode with zero rotational speed to understand how the system behaves when the rotor is not spinning; the mode shapes are easier to interpret in this case.
Please see the following forum topic where I helped a user interpret the mode shapes of a spinning rotor–the results are nearly identical to yours: Linearization in FAST with unbalanced Rotor.
Best regards,