PSI - Issue 57

Pierrick Lepitre et al. / Procedia Structural Integrity 57 (2024) 395–403 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

401

7

Figure 4 - Self-heating curve of six bare 300M steel coupons, R = -1, with the identified self-heating model

By considering a threshold between the two regimes of 1 % (the dissipation of the second regime is equalto 1 % of the total dissipation), σ AE = 895 MPa is determined. With equation (7), the relative standard deviation is estimated to 4.5 %. So, the estimated fatigue limit by self-heating for 1 % and 99 % probability of rupture are 775 MPa and 965 MPa. D = 890 MPa and the fatigue limit for 1 % and 99 % probability of rupture are 740 MPa and 1 040 MPa. Results have been summarized in Table 2. With a difference of 5 MPa on the mean conventional fatigue limit and a relative difference of less than 10 % on the 1 % and 99 % probability of rupture fatigue limits, self-heating tests and the self-heating model show an excellent correlation with conventionalfatigue data on bare 300M.To obtain the six self-heating curves, only 5 days of testing and post-processing were required. That can be compared to the several months needed to realize a staircase (number of coupons > 25 to have a good estimation of the standard deviation). Moreover, as the repeatability of the self-heating curve has been validated on six test coupons, a great first estimation could be obtained with a single self heating test, within one day. With conventionalfatigue tests, the determined values for the fatigue limit at 2 x 10 6 cycles is σ

Table 2 – Self-heating model parameters for 300M bare steel at R = -1 Parameters Unity

Self-heating test Conventional fatigue test Relative difference (%)

σ D

MPa MPa

895

890

0.6 %

P f, 1/99%

[775; 1 015]

[740; 1 040]

[2.5; 4.7] %

3.3. Load ratio effect

On F50 bare 300M test coupon,with the same testing and post processing protocol, anotherload ratio was tested, R = -0.25. The self-heating curve is plotted in Figure 5 with the self-heating model for R = -1. The rupture occurs during the loading block of σ a = 900 MPa and every loading block exhibits a stabilized temperature after a 5τ eq solicitation. Three observations can be made:

• the first dissipation regime is not affected by the loading ratio, • the higher the load ratio the earlier the transition between the first and the second regime, • the load ratio seems to affect slightly the slope of the second regime.

To compute the self-heating estimation, we used the relation given by Louge (2019)