PSI - Issue 33

A. Kansy et al. / Procedia Structural Integrity 33 (2021) 1173–1180 A. Kansy, M. Kaynak, C. Bleicher, H. Kaufmann/ Structural Integrity Procedia 00 (2019) 000–000

1174

2

1. Introduction According to Kloos [1], the size effect on the fatigue strength can be classified by the technological, statistical, geometrical and surface size effect. However, statistical and geometrical size effects cannot be considered independently of each other. The statistical size effect is based on the fact that the probability of occurrence of a weak point leading to a crack is greater in the interior of a large specimen than in smaller specimens. The geometrical size effect takes into account the micro-support effect due to an occurring stress gradient so that the material can locally withstand higher stresses [2]. Fatigue assessment for a component is usually performed on the basis of experimental tests from small-scale specimens. In order to transfer the fatigue strength of the specimen tests to the component, it is necessary to take into account the statistical and geometrical size effect. The concept of the highly stressed volume has become established for determining the statistical and geometrical size effect [3]. The smaller the highly stressed volume, the higher usually the fatigue strength. A detailed study on the statistical size effect of grey cast iron was carried out by Buthmann [4]. Continuous cast material of the grey cast iron grade EN-GJL-250 with the same chemical composition and similar cooling conditions was used for specimen removal. Specimens with diameters of 4, 12 und 20 mm and notch factors K t of 1, 3, 5 were tested under axial and torsional loading. In addition, data from other studies were included in the evaluation. The local stress amplitude was plotted against the highly stressed volume. The evaluation by the three-parameter Weibull distribution proved to be a good way to describe the dependence of the fatigue strength on the highly stressed volume. For the present research work, the grey cast iron grade EN-GJL-300 was investigated concerning its statistical and additionally geometrical size effect and evaluated via the concept of the highly stressed volume HSV 90% . 2. Material and experimental setup 2.1. Material Data Y-shaped cast samples and a component of EN-GJL-300 were used for this work. The component used was an industrial gearbox housing with dimensions of approximately 1400 mm x 700 mm x 500 mm and a weight of 1000 kg. Images of the castings are shown in Figure 1.

Figure 1. Cast samples (left) and component (right) with marking of the specimen removal

Nomenclature K t

notch factor [-]

 a,n nominal stress amplitude [MPa]  a,n,lim nominal stress amplitude at 1ꞏ10

HSV 90% highly stressed volume [mm³] number of cycles to failure [-]

7 cycles [MPa]

N f N k

local stress amplitude [MPa] probability of survival [%]

 a,l

cycles at knee point [-]

P S

N lim

1ꞏ10

7 cycles [-]

T

scatter band [-]

σ