PSI - Issue 48

Gašper Fašun et al. / Procedia Structural Integrity 48 (2023) 19–26 Fašun et al/ Structural Integrity Procedia 00 (2023) 000–000

22 4

between 13 mm and 27 mm for material A and 13 mm and 19 mm for material B. The largest observed inclusion during the examination of the whole surface area of the test specimen was diA = 21.9 mm for material A and diB = 16.9 mm for material B.

(b)

(a)

Fig. 2. Microstructure of materials A (a) and B (b). Prior to the metallographic analysis, the hardness HRc was measured on the samples, which returned values of 41 HRc for material A and 38 HRc for material B. On the entire cross-section the micro-hardness HV0,5 was measured, given average values of 415  15 kgf/mm2 for material A and 388  15 kgf/mm2 for material B. Tensile test samples were manufacturing in accordance with ISO 6892-1 standard [9]. The cut-out location of the samples is marked with the number 1 on the barrel cross-section in the Fig. 1. Two tensile tests were performing for each of the materials, given an average ultimate tensile strength of   = 1303 MPa and   = 1152 MPa. Experimental measurements of the fracture toughness in terms of the parameter   ܥ were perform with the testing machine Instron 1255, and three CT samples with side grooves for both materials. They were manufactured and tested in accordance with the ASTM E399 standard [10] and cut from the location that is marked with the number 3 on the barrel cross section in the Fig. 1. The measured average values (of three measurements) are   ܥ  = 151.4 ܲ   and   ܥ  = 151.1 ܲ   . No significant difference can be observed for both materials. The measurement of the fatigue crack growth properties was performed on bending SENB samples with a four point bending fatigue device RUMUL. The geometry of the test samples and the experimental procedure were performed in accordance with the ASTM E647 standard [11]. The samples were cut from the location marked with the number 2 on the cross section of the barrel shown in Fig. 1. The loading ratio  was equal to 0.1 throughout the experiments, and the ambient temperature fluctuated between 23°C and 25°C. In order to account for the long crack growth rate da/dN the following modified Paris-Erdogan expression is used [12]: ܰ  = ܥ ∆− ∆ ℎ  (1) where C and m are constants dependent on the material, the environment and the load-ratio . The results were plotted on a log-log graph, and fitting the data allowed the estimation of the material parameters C and m. For the fitting process the threshold for long crack is needed, DKthR, which is estimated by using the following expression proposed by Chapetti for steels and R = 0.1 [16]: ∆ ℎ =− 0.0021   + 8.4 (2) where   is the ultimate tensile strength of the material and  K thR has units of ܲ   . Eq. (2) returns values of ∆  ℎ = 5.66 ܲ   for the material A, and ∆   ℎ = 5.98 ܲ   for the material B. Then, fitting Eq. 1 to the experimental results the calculated parameters for the materials A and B are m = 1.76 and C = 6.864 10 -11 m/cycle , and m =1.72 and C =6.925 10 -11 m/cycle , respectively. Fatigue limit was measured by using the thermographic method [14]. Cylindrical samples were machined with length  = 160 mm and the smallest radius  ݅ = 30 mm, which were cut from the barrel from the location on the cross section in the Fig. 1 marked with the number 4. Fatigue tests were carried out by using an INSTRON 1255 servo-hydraulic fatigue testing machine at 12 hz, with a stress ratio R = 0.1. Temperatures increases  T for each