Fourier transform rheology as a tool to determine the fatigue behavior of polymers
|Advisor:||Rodrigue, Denis; Wilhelm, Manfred|
|Abstract:||This thesis proposes a new framework to analyse, quantify and predict the mechanical fatigue of amorphous polymer using a method based on the decomposition of the stress response via Fourier transform. In particular, fatigue tests were performed under strain controlled torsion and tension/tension deformation and the time data of the strain, torque and force were recorded and decomposed into linear and nonlinear contributions via higher harmonics. In particular, three concepts have been developed to quantitatively determine the time behavior of the samples. Firstly, the generation of macroscopic cracks was found to correlate with sudden increases in the I2/1 intensity. Secondly, an on-line method to predict the fatigue lifetime was developed, based on the rate of change of I3/1 with respect to the cycle number N (dI3/1/dN) before the onset of failure. This prediction was found to be more precise than Wöhler curves predictions since the correlations have on average much lower standard deviations (30 vs. 60%). Thirdly, a fatigue criterion solely based on mechanical nonlinearity was developed: the cumulative nonlinearity Qf. This parameter correlates the integral of the nonlinearity Q (Q = I3/1/yo²) until failure with the number of cycles to failure Nf. The standard deviation of the Qf vs. Nf correlation was found to be less than 30%, indicating that Qf is a more precise fatigue criterion than commonly used ones such as the cumulative dissipated energy density or the cumulative stress (±50%). Finally, these three concepts were successfully applied on different conditions (type of deformation, range of frequency, deformation amplitude) and polymers such as polystyrene (PS), polymethylmethacrylate (PMMA), styrene acrylonitrile (SAN) and polytertbuthylmethacrylate (PtBMA).|
|Document Type:||Thèse de doctorat|
|Open Access Date:||1 May 2019|
|Collection:||Thèses et mémoires|
All documents in CorpusUL are protected by Copyright Act of Canada.