Q-criterion calculation for cell-based .pvtu files

Dear all,
I am attempting to calculate the Q-criterion for my cell-based data stored in a .pvtu file along with its associated .vtu files. After converting the cell-based data to point data, I used the ‘Gradient’ filter on the velocity vector to compute the Q-criterion. However, the resulting values are globally zero, which appears to be caused by a mismatch between the point data and the corresponding coordinates.
Any insights or suggestions would be greatly appreciated!
Best regards,
Y.X.Wang


The case is flow over a 3D cylinder, and the above is the U distribution

This is the result of Q-criterion calculation