Thanks for the answer. When I say "simple algorithm didn't work," I mean it has resulted in like previous analysis. It has not converged, fluctuations still continue like picture that i attached below.
I follow your suggestion and checked my poor quality meshes. It is clear until 0.16 orthogonal quality(attached picture). Also this is my skewness measure:
           name    id cells (quality > 0.90)  maximum quality cell count
------------------------- -------- ---------------------- ---------------- ----------
     Overall Summary   none            0     0.831676   8694808
Total Number of Cell Zones : 16
[Quality Measure : Skewness]
Â
You are navigating away from the AIS Discovery experience