an investigation of fatigue and fretting in a dovetail joint

of fatigue life, as is commonly found in other contact problems. T M This paper describes a more extensive experimental investigation of the problem, combined with a numerical stress analysis~ for the interpretation of the results.
Байду номын сангаас
Introduction
The fatigue life of the dovetail joints between blades and disks in gas turbines is normally assessed by means of spinning-disk tests, which are expensive and time consuming. Boddington and Ru'iz' have shown that the state of stress can be reproduced by means of simple specimens consisting of a rectangular sheet, modeling the disk, with three pairs of 'blades' fixed to the sheet back-to-back by dovetails, as illustrated in Fig. 1. The specimen is loaded under biaxial tension in a fatigue-testing machine in which two 250-kN hydraulic actuators, acting on sheet and blades respectively, are controlled to operate between any two independent loads, displacements or strains, in-phase or out-of-phase. One actuator is supported vertically through a spring and damper arrangement on a cruciform base; the other is suspended horizontally from a rectangular frame, also on springs. The machine was designed to minimize the displacement of the geometrical center of the specimen, ensuring that the loading was entirely along the two perpendicular axes, with very little bending. This was verified by testing a strain-gaged cross piece before loading the actual specimens. In a few preliminary tests conducted to commission the fatigue-testing machine, it was found that the specimen life was strongly dependent on the ratio between the amplitude of the loads imposed on sheet and blades. This is ascribed to the relative slip between blade and sheet, which increased as the dovetail opened up under increasing sheet loads, rather than to a change in the peak stress.' Fretting damage is held to be responsible for the reduction
Testing Procedure and Results
From the preliminary tests, a loading and specimengripping procedure that eliminated misalignment and resulted in a uniform biaxial loading was developed. The outer blades were thinned down to ensure that the central blade took approximately I0 percent more load than the outer blades so that failure always occurred at the central dovetail, where the stress field most closely resembles that of the true spinning disk. A total of 28 tests were conducted; the results are given in Table 1. In all cases, the
An Investigation of Fatigue and Fretting in a Dovetail Joint
by C. Ruiz, P.H.B. Boddington and K.C. Chen
ABSTRACT--Specimens designed to reproduce the state of stress of the dovetail joint between blade and disk in a typica) gas-turbine configuration have been tested in a 250-kN capacity biaxial fatigue machine. The materials tested were Ti-IMI 829 (disk and blade), Inco 901 and steel FV535 (blades only). The investigation highlights the importance of fretting at the disk/ blade interface. A design parameter that combines fretting damage and peak stress is proposed as a basis for the life assessment of the joints.
Specimens
The general arrangement of the specimens and the profile of the dovetail are given in Fig. 1. The dovetail slots were rough cut with a conventional milling cutter and finished with a 4-mm-diam carbide cutter. The finishing cutter was fixed to a specially designed slotting device that applied a rotary and plunging action to the cutter and was guided by a numerically controlled milling machine. In this way, accurate profiles with a surface finish better than 1 /an were consistently and reproducibly generated. The blades were finished by form grinding. Two strain gages were fixed to the blade shanks, to measure the in-plane loading in each blade and to detect any out-of-plane bending. The sheet was also strain gaged to, facilitate the alignment of the specimen and testing machine. The materials used for sheet and blades were: Sheet--IMI 8296 titanium alloy. Chemical composition (percent): 5.5A1 - 3.5SN - 3Zr - 1Nb 0.25Mo - 0.3Si Ti. Mechanical properties: 0.2 percent proof stress 820 MPa, tensile strength 950 MPa. Blades--IMI 829 as sheet (8 tests). INCO 901 nickel alloy. Chemical composition (percent): 0.1C - 0.4Si - 0.5Cu - 0.5Mn - 12.5Cr - 2.9Ti - 0.35A1 1Co - 5 . 7 5 M 0 0.03S - 42.5Ni - Fe. Mechanical properties: 0.2-percent proof stress 900 MPa, tensile strength 1200 MPa (9 tests), Alloy steel 12Cr Co Mo V Nb to UK specification FV535; tensile strength 1065 MPa, yield 930 MPa (1 i tests).