PSI - Issue 57

Lucas Carneiro Araujo et al. / Procedia Structural Integrity 57 (2024) 144–151 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

148

5

3. Experimental campaign

3.1. Material and specimens

The experimental campaign was conducted with fatigue specimens made of AISI 4140 (42CrMo4) steel, oil quenched, and tempered at 600 °C. The material was removed from crankshafts of stationary generators that failed due to fatigue during operation. Table 1 provides the mechanical properties of the AISI 4140 steel, which were determined through a single tensile test conducted at a displacement rate of 0.5mm/min. The Vickers hardness was measured by averaging ten readings from three different samples, with a load of 100 kgf. The specimens used in the experiment were designed following ASTM E466-15 standards and had a circular cross-section, with 10 mm in diameter. Two types of cylindrical specimens were utilized: (i) smooth specimens without artificially introduced surface defects, and (ii) specimens containing a machined micro surface hole with a diameter and depth of 550 μm . Table 1. Mechanical properties of AISI 4140 steel. Young’s modulus (GPa) 0.2%-offset yield stress (MPa) Tensile strength (MPa) Elongation (%) Vickers hardness (kgf/mm2) 202 647 932 20 320 Fatigue tests were performed under force control in accordance with ASTM E466-15 in an axial (100 kN capacity) and an axial-torsional (100 kN e 1100 Nm) servo-hydraulic testing machines. The tests were conducted in room temperature and the frequencies kept between 5 and 15 Hz, depending on the load level. All tests were carried out with totally alternating sine waves, that is, with a loading ratio R of -1. Fatigue tests were conducted to examine various ratios between shear stress amplitudes ( ) and normal stress amplitudes ( ). The ratios considered were 0, 0.5, 1, 2, and ∞, representing traction -compression, combined loads, and pure torsion. Both in-phase and 90º out-of-phase tests were performed, for the combined loads. The objective of these tests was to observe at least one failure and one run-out for each type of loading condition. A run-out was defined as the point where the specimen did not fail within a specified number of cycles, which was set at 2.10^6 cycles. A complete rupture of the specimen was used as the failure criterion. If a specimen failed, the stress level was reduced for the subsequent test. Conversely, if the run-out was reached, the next test would be performed at an incrementally higher stress level using a new specimen. In other words, no specimen was reused during the evaluation of the proposed multiaxial model. 4. Results and discussion 4.1. Fatigue strength from √ parameter In the work of Machado (Machado et al., 2020), were conducted an analysis of inclusions and used the extreme statistics method to compute the value o √ of 145 µm. This value was used in the calculation of the fatigue limit, taking into account the natural defects of the material, as per Eq. 4. For the calculation of the fatigue limit considering the presence of the micro hole, as per Eq. 3, it was only necessary to calculate the projected area of the hole in the direction of the maximum principal stress in tension. Due to the geometry of the manufactured hole, the projected area forms a square with sides equal to the diameter of the hole, resulting in a computed value of √ of 550 µm. The results obtained for the fatigue strength using the √ parameter model, considering the material inclusions and the presence of the micro-hole, are presented in Tab. 2. 3.2. Fatigue tests