PSI - Issue 13

Przemysław Strzelecki / Procedia Structural Integrity 13 (2018) 631 – 635 Author name / Structural Integrity Procedia 00 (2018) 000–000

632

2

for smooth elements and notched are identical. Schijve (2009) has identified an issue with the scatter of the test results. The test results for two structural components were presented, where the fatigue life of one of the components was 8.5 times higher at 50% failure probability than that of the other component. Due to higher scatter, both components showed the same fatigue life at 0.01% probability. However, at 1% probability, the difference was 2.3 times the fatigue life. Test results show that the reduction in fatigue notch factor will not always yield expected results.

Nomenclature A

elongation

the modulus of elasticity stress concentration factor number of cycles stress amplitude ultimate tensile stress the coefficient of variation percentage reduction of area yield stress

E

K t N S u S y V Z b d m S

constant term in a S-N curve equation

constant term in a S-N curve equation in the location parameter

scale coefficient in a S-N curve equation

scale coefficient in a S-N curve equation in the location parameter

n r

radius of the specimen notch the scale parameter 10 m ⋅ log( S )+ b gamma function the location parameter 10 n ⋅ log( S )+ d

α v the shape parameter β ( S ) Γ () ζ (S)

standard deviation of the 3-parameter Weibull distribution for 10 expected value of the 3-parameter Weibull distribution for 10 5 cycles 5 cycles

σ x μ x

The purpose of the study was to show how the scatter of fatigue life changes according to the stress concentration factor. The scatter of fatigue life were analysed by Schijve (2005), Klemenc and Fajdiga (2012) and Gope (2012). It was showed that the scatter depends on number of cycles and the 3-parameter Weibull distribution better fit to experimental data. Additionally, normal distribution can be used only for 5 % probability, see ASTME-739-91 (2006). To estimate the S-N curve, the 3-parameter Weibull distribution was used, expressed as follows Weibull (1949): ���� � � � ��� � � � � � � ���� � � ��� � � � exp �� � �� � ����� � � ��� � � � � � � (1) The maximum likelihood method is used to estimate parameters for equation (1) and was described in Strzelecki and Tomaszewski (2016) and Sarkani et al. (2007). 2. Experimental tests To analyse the scatter of the fatigue life for different the stress concentration factor was carried out the fatigue tests for AW 6063 T6 aluminium alloy. The specimens were made from a 10 mm drawn rod. Fig. 1 and Table 2 shows the geometry of the specimens used in the fatigue tests. The mechanical properties of the material were determined experimentally in a static tensile test in accordance to PN-EN ISO 6892-1:2016 (2016). Table 1 shows the average values for 20 specimens of the tested material. The fatigue tests were performed using the rotating bending test station, verified in Strzelecki and Sempruch (2012). The load frequency used was 50 Hz. Fig. 2 shows the fatigue test results. Out of the 30 tests were used for each geometry of specimen. The number of tested specimens was chosen according ISO-12107 (2012) for reliability pursuits.