Experimental Determination of Stress Intensity Factors for Cracks in Turbine Discs

Abstract

Stage-I1 fatigue crack growth paths in firtree fixtures have been predicted using a photoelastic technique. Initiation was assumed to occur at the edge of contact on the load-bearing flanks and, subsequently, the cracks were extended in the direction of the maximum circumferential stress. After a short initial length, in which propagation was perpendicular to the contact surface, the direction of crack growth was, in a broad sense, equivalent to the hoop direction in the disc. Stress intensity factors were found for the majority of the crack path using data taken from the isochromatic fringe patterns. It was concluded that the most catastrophic failure was likely to occur from the innermost land of the firtree and that the width of the firtree had little influence on the stress intensity factors.

Crack paths under mixed mode loading

“…Fracture mechanics studies using transmission photoelasticity require fine slits to be introduced into epoxy models of engineering components [16,17]. Several methods have been developed to determine K I and K II using the full field of data surrounding the slit tip [18,19].…”

“…Several methods have been developed to determine K I and K II using the full field of data surrounding the slit tip [18,19]. Nurse and Patterson [16] also developed a photoelastic method to predict the direction of crack growth using the theory that long cracks usually grow under mode I loading in direction perpendicular to maximum tangential stress. They found that when K II /K I is less than 0.7, this direction is approximately equivalent to the axis of symmetry observed in the isochromatic fringes loops and so one can predict the direction of crack growth.…”

Study of Effective Stress Intensity Factor through the CJP Model Using Full-Field Experimental Data

“…The CJP model is based on Nurse and Patterson’s approach [ 26 ] for the description of Muskhelishvili’s complex potentials [ 9 ] as Fourier series. The stress tensor components surrounding a fatigue crack according to this model are given by Equations (1)–(3) [ 1 ]: σ x = r −1/2 {−1/2 (A + 4B + 8E) cos(θ/2) − 1/2 B cos(5θ/2) − 1/2 E [ln(r) (cos(5θ/2) + 3 cos(θ/2)) + θ (sin(5θ/2)+ 3 sin(θ/2))]} + C, σ y = r −1/2 {1/2 (A − 4B − 8E) cos(θ/2) + 1/2 B cos(5θ/2) + 1/2 E [ln(r) (cos(5θ/2) − 5 cos(θ/2)) + θ (sin(5θ/2) − 5 sin(θ/2))]} − F, τ xy = r −1/2 {A sin(θ/2) + B sin(5θ/2) − E sin(θ) (ln(r) cos(3θ/2) + θ sin(3θ/2))}, where r and θ are the polar coordinates in the crack plane (see Figure 1 ), and A, B, C, E, and F are the coefficients describing the stress fields.…”

“…The CJP model is based on Nurse and Patterson's approach [26] for the description of Muskhelishvili's complex potentials [9] as Fourier series. The stress tensor components surrounding a fatigue crack according to this model are given by Equations ( 1)-( 3) [1]:…”

Experimental stress intensity factors for cracks in a thin plate with stiffeners

“…The cracks were produced artificially by introducing a short 'starter' crack at the hole and then incrementally extending the crack in a direction predicted for slow crack growth (9,10). At each increment the stress intensity factors were determined using a procedure developed by the authors (ell).…”