Paper Number
PO102 

Session
Poster Session

Title
Fatigue analysis via Fourier transform of the stress

Presentation Date and Time
October 17, 2018 (Wednesday) 6:30

Track / Room
Poster Session / Woodway II/III

Authors (Click on name to view author profile)

1. Hirschberg, Valerian (Université Laval, Department of Chemical Engineering and CERMA)
2. Wilhelm, Manfred (Karlsruhe Institute of Technology, Institute for Chemical Technology and Polymer Chemistry)
3. Rodrigue, Denis (Université Laval, Department of Chemical Engineering and CERMA)

Author and Affiliation Lines (in printed abstract book)
Valerian Hirschberg¹, Manfred Wilhelm², and Denis Rodrigue^1
1Department of Chemical Engineering and CERMA, Université Laval, Quebec City, Quebec G1V 0A6, Canada; ²Institute for Chemical Technology and Polymer Chemistry, Karlsruhe Institute of Technology, Karlsruhe 76128, Germany

Speaker / Presenter 
Hirschberg, Valerian

Text of Abstract
In this work, the stress response of mechanical testing on notched rectangular solid polystyrene (PS), polymethylmethacrylate (PMMA) and styrene-acrylonitrile (SAN) samples in oscillatory shear was analyzed via Fourier transform (FT) to determine fingerprints of continuous fatigue. The tests were performed at room temperature using a torsion geometry and were filmed to visualize changes. Large strain amplitudes were applied so the stress response was nonlinear and higher harmonics were detected via FT. The idea is to analyze the time evolution of the stress via linear (storage (G’) and loss (G”) moduli) and nonlinear parameters (higher harmonics), their derivatives and integrals. These parameters were used to better understand fatigue, to detect and describe specific events such as crack initiation and propagation, to predict the fatigue lifetime and to develop failure criteria. The linear parameters (G’, G’’) were found to decrease monotonically, while the I_3/1 intensity (relative amplitude of the third harmonic to the fundamental one) steadily increased until failure. These three parameters were found to change linearly with the cycle number, shortly after the beginning of the test until failure onset. The fatigue lifetime was found to follow a power-law function of the rates of change of G’, G’’ and I_3/1 in this regime. For undamaged samples, the nonlinear parameter I_2/1 is within the noise level, but its intensity increased when macroscopic cracks were created. Additionally, to better understand the origin of failure, the integral of the nonlinearity (Q = I_3/1/?₀²) until failure was analyzed as a function of the fatigue lifetime. A power-law function between the integral of Q and the fatigue lifetime was found. The time evolution of G’, G’’, I_3/1, I_2/1 and the integral of Q as a function of the fatigue lifetime, are proposed as new, highly sensitive, criteria to predict failure and detect the onset of macroscopic cracks.