PSI - Issue 7

Ludvík Kunz et al. / Procedia Structural Integrity 7 (2017) 44–49 Ludvík Kunz / Structural Integrity Procedia 00 ( 201 7) 000–000

45

2

1. Introduction Superalloys are important structural materials for high-temperature applications in many branches of industry. Despite being studied and developed for many decades, there is a permanent interest of industrial enterprises in alloys with better mechanical and corrosion properties and reasonable economic cost. Cast nickel-base superalloys like IN 713, its low carbon variant IN 713LC or MAR-M 247 are frequently applied materials for the production of turbine blades, turbine discs or turbocharger wheels. Despite the very long time of engineering application of cast superalloys the available characteristics of their high-temperature behavior, particularly the high cycle fatigue life at high mean stresses are very little dealt with in the open literature. This is quite surprising when compared to a number of studies in the low cycle fatigue or a number of investigations of the influence of dwell periods on the high-temperature behaviour and lifetime. Turbine blades of a gas engine or turbocharger wheels are examples of components which are in service loaded at high temperatures with high tensile mean stresses and superimposed high-frequency cycling. The tensile stress is a result of centrifugal forces during the turbine operation and the high-frequency cycling comes from blade vibrations. For safe design against fatigue at the presence of mean stress several schemes have been proposed. Gerber, Goodman, Haigh or Soderberg diagrams are applied in engineering practice since the beginning of the last century, e.g. Heywood (1962), Suresh (1998). These diagrams represent constant fatigue life charts for a combination of stress amplitude and mean stress in such a way that the regions where fatigue failure will not occur, or will take place at a defined number of cycles are demarcated. The constant fatigue life diagrams express the general observation that the fatigue lifetime at a given stress amplitude decreases with increasing tensile mean stress. Some data, however, indicate that application of small stress amplitudes may result in an increase of the lifetime, see e.g. an old paper by Vitovec (1957). For the majority of diagrams it is characteristic that they are often weakly supported by experimental data in the region of high mean stresses and small cyclic stress amplitudes, i.e. in the region where the interaction of creep/high-cycle fatigue appears. Vitovec (1957) summarized the knowledge on creep/high-cycle fatigue interaction available at that time and came to the conclusion that the effect of vibrations is dependent on material, temperature and mean stress. Small high frequency stress amplitudes superimposed on the mean stress can be harmless or can result in shortening or extension of the lifetime. Sheffler (1972) presented strong reduction of the lifetime of two refractory alloys subjected at high temperatures and tensile mean stresses to very high-frequency cycling with the ratio of alternating to mean stress in the range from 0.1 to 0.65. No effect of small amplitude cycling with 100 Hz frequency at temperature 900 °C for IN 100 was reported by Lesne and Gailletaud (1987). For CMSX-4 single crystals it has been shown that vibrations superimposed on static load at 800 °C slow down the creep process, however, accelerate the fracture process, Lukáš et al. (1996). From the current knowledge, despite many decades of research of creep/fatigue interaction, it is not possible to draw reliable conclusions on the influence of small high-frequency vibrations on the lifetime. Simultaneously, the details of the mechanisms influencing the acceleration or retardation of creep and initiation of fatigue cracks under combined loading differ according to the material, temperature, loading conditions and environment. Moreover, the knowledge on the damage mechanisms is limited. The aim of this study was to experimentally determine the effect of high-frequency cycling superimposed on tensile mean stress on time to fracture of cast Ni-base superalloys IN 713LC at 800 °C and MAR-M 247 at 900 °C and to investigate the damage mechanism. 2. Material The semi-products for specimens were shaped by casting into a two-flow shell mould, Fig. 1(a). The samples for creep/fatigue testing were machined according to the drawing shown in Fig. 1(b). The casting temperature was 1360 ± 10 °C for both the alloys. The semi-products for specimens from IN 713LC alloy were not heat treated. The specimens were machined directly from the as-cast semi-products. The coarse dendritic structure in the gauge length is shown in Fig. 2(a), where the grain boundaries are highlighted by the full black lines. The average grain size determined by an intercept method was about 1.0 mm. The semi-products from MAR-M 247 were processed by hot isostatic pressing (HIP) at 1200 °C and pressure 100 MPa for 4 h. The material was subsequently solution annealed at 1200 °C for 2 h with cooling on air followed by precipitation annealing at