Abstract
This work investigates the possibility of applying Fourier Transform (FT) analysis of the force signal to follow fatigue behavior of metals under oscillatory displacement-controlled tests in uniaxial tension/tension. As a first step, three different materials were selected (cold rolled steel, aluminium and brass). The FT analysis revealed a low level of nonlinearities in the force response, which was possible to measure and quantify as higher harmonics of the imposed sinusoidal deformation. Due to geometric reasons, the odd higher harmonics represent the symmetric nonlinearity while even ones are related to asymmetry, so both odd and even harmonics need to be analyzed separately. The time evolution of the higher harmonics showed that the odd higher harmonics continuously increase during the test. Criteria to better predict the mechanical fatigue and failure (life time) are then proposed based on the integral and derivative based on the time evolution the odd higher harmonics. In contrast, for tests in the high cycle fatigue regime, the even higher harmonics are mainly noise at the beginning of the test (undamaged state), but start to rise after the occurrence of a crack due to internal crack friction. Based on the analysis performed, FT analysis of the force during mechanical fatigue testing of metals is a sensitive tool used to predict failure and to improve our understanding of the dynamics involved in mechanical fatigue.
Highlights
The most common and straightforward way to investigate mechanical fatigue is to identify a correlation between the number of cycles to failure (Nf ) with the average applied stress (σ0 ) or the strain amplitude (ε0 ) for stress or strain controlled tests respectively; i.e., the so-called S-N or strain-life curves [1]
When a sample is deformed in cyclic strain controlled tension/tension (T/T), the constant static deformation is larger than the strain amplitude (ε0 ) of the dynamic deformation, to avoid compression at the strain minimum of each cycle
The decrease in the I2/1 intensity in Figure 6b, instead of the typical I2/1 increase with crack onset, can be explained by a vector compensation in a complex plane of the I2 contribution caused by the noise of the machine and the asymmetry in the force response caused by the crack initiation
Summary
Failure of a part under a repetitive load is called mechanical fatigue break-up. Even today, the most common and straightforward way to investigate mechanical fatigue is to identify a correlation between the number of cycles to failure (Nf ) with the average applied stress (σ0 ) or the strain amplitude (ε0 ) for stress or strain controlled tests respectively;. Strain-life curves report the average value of the number of cycles necessary for part failure under specific loading conditions over a certain amount of fatigue tests for the same testing conditions. They only provide information about the most probable number of cycles to failure [2], but no information about the failure process (onset of cracks). The predictive power of a measured fatigue lifetime for a single sample or application is very limited This is why a deeper understanding of the fatigue process is needed, to predict and avoid failure (accidents), and to save resources by not stopping the test/application too early due to high safety factors. Better fatigue life times are predicted, as well as the possibility to detect crack initiation and propagation with high sensitivity
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
