A new approximation to modulation effect analysis based on empirical mode decomposition method
Paper in proceedings, 2018
'Modulation effect' which represents the amplification or attenuation of the SS (small-scale) structures near the wall by the outer LS (large-scale) structures is one of the two commonly accepted influence of the outer large-scale energy containing motions on the near wall turbulent fluctuations. Mode decomposition based on filtering suggested by Mathis et al.  is a widely used method for the analysis of the modulation effect. In the present study a new approximation is proposed based on the empirical mode decomposition (EMD) to investigate the aforementioned amplitude modulation effect. Both Mathis et al.  and suggested methods are being used and results are compared for two-point and single-point analysis. It has been shown that the LS and SS signals that are decomposed by filtering and EMD follow identical path. Although the similarities of the signals, the correlation coefficient, R, differ for the two methods. We observed that the suggested method in the present work exhibits a slightly higher correlation coefficient compared to the method based on filtering for the two-point analysis. However for the one-point analysis existing method seems far away to give a rational correlation coefficient value, which is too low compared to the suggested method. Therefore we believe that the suggested method which decomposes the signal by EMD instead of filtering adresses many recent studies that questions the reliability of the calculation of the correlation coefficient, R, by existing method. We believe that the variation of the correlation coefficient, R, for the identical signals extends the discussion of the correlation coefficient calculations to the very first process, that is obtaining LS and SS from the original signal.