ANALYSIS OF NATURAL FREQUENCIES FOR EARLY DAMAGE DETECTION IN CFRP SHAFTS USING RSM-BASED FREQUENCY CONTOUR PLOTS

Abstract

This study explores early damage detection in carbon fiber-reinforced polymer shafts by analyzing natural frequencies. Modern engineering components often face increased flexibility and high stress levels, leading to cracks in rotating parts, which can result in premature failures. To address this issue, modal analysis, specifically natural frequency analysis, is employed to identify deviations caused by cracks. Cracks alter stiffness and mass distribution, leading to shifts in natural frequencies. The study employs finite element models to simulate various crack locations and depths, normalizing them with respect to shaft diameter and length. A cantilever shaft configuration is utilized with refined mesh structures near the transverse crack. The analysis leverages the frequency contour method with response surface methodology to visualize how crack depth and location influence normalized natural frequencies. Results indicate that crack depth has a significant impact on natural frequencies, while crack location has a subtler effect. Combining depth and location produces the most pronounced frequency variations. The corresponding R² values for the normalized first and third natural frequencies are 95.69% and 96.32%, while for the normalized second natural frequency this value is 75.70%. The study also demonstrates the use of frequency contour curves for accurate crack detection, with the 1st and 3rd natural frequencies being reliable indicators.