PSI - Issue 42

Martin Matušů et al. / Procedia Structural Integrity 42 (2022) 102 – 109 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

105

4

Typically, the evolution of temperature on the surface of a specimen representing the self-heating (S-H) effect has three different phases; see Fig. 3. The first phase can be described by a rapid increase in temperature at the rate R 0 , which occurs in the beginning of the fatigue test. The second phase is usually represented by a stabilized temperature ∆ T s . Heat is uniformly transferred to the surrounding environment of the specimen. The stabilized temperature T s is used as a parameter for estimating the fatigue life performance, see Fargione (2001), Amiri and Khonsari (2010) or Meneghetti (2007), as well as for fatigue limit estimation, see La Rosa and Risitano (2000), Luong (1998) or Procházka and D ž ugan (2017). The third phase of the S-H effect is represented by the higher rate in temperature increase R c that leads to failure; see Fig. 3a. The temperature evolution in these three phases is dependent on the load amplitude; see Fig. 3b.

Fig. 3. an) Scheme of the S-H effect depicting three phases of temperature evolution during cyclic loading to failure; b) S-H effect with an illustration of the limiting energy and of the unique evolution of temperature originating from different load amplitudes.

In the cases presented in this article, the temperature evolution differs in the second phase of the S-H effect. It cannot be simply described by  T s , see Fig. 4, because the temperature continues to increase in the second phase. The temperature growth rate is designated as R s . A linear function is used to model the curvature in the second phase within

this paper. Fig. 4. a) Scheme of an idealized S-H effect of 42CrMo4+QT with displayed phases and observed parameters throughout the measurement. b) The manifested S-H effect of a real specimen from the A03 series with 550MPa stress amplitude and the failure 219 101cycles. 3.2. Fatigue limit estimation The definition of the conventional fatigue limit  FL is that it is a limit stress that the structure can withstand without failure in repetition for any number of cycles. The expected correlation between the S-H effect and  FL for certain materials and testing conditions is well described in La Rosa and Risitano (2000), Luong (1998) or Douellou et al. (2019), who describe that the  T s -  relationship for amplitudes below  FL significantly differs from the trend observed at stress levels above  FL . The differences between the thermal response of the material at specific stress amplitudes is often projected into a bilinear evolution of  T s -  function. The position of the intersection of both linear segments is assumed to be related to  FL , see Fig. 5a, where the stress amplitude is normalized by  FL to make