Weird rotor scan troubleshooting

[xiaoruiDong]

This is a special case where a ring flip is involved. The molecule is CC1([CH]CO)COO1.

The red rotor is the rotor being scanned, rotors in other colors are those being frozen. The strategy is proposed by the latest troubleshooting module. The 4-member-ring flipped during the previous scan, looks similar to this:

The output file I got is output.out.zip

The error got is

 New curvilinear step not converged.
 Error imposing constraints
 Error termination via Lnk1e in /cm/shared/c3ddb/gaussian/09.d01/g09/l103.exe at Wed May  6 06:25:10 2020.

So it seems like ARC is over-constraining the molecule for its scan. But I cannot really tell how it is over-constrained. My guess is the ring is over-constrained but cannot tell why and how. Let me know your thoughts and I really appreciate your help. Thanks