PSI - Issue 17

Shahnawaz Ahmad et al. / Procedia Structural Integrity 17 (2019) 758–765 Shahnawaz Ahmad/ Structural Integrity Procedia 00 (2019) 000 – 000

763

6

been performed for various values of initial crack size i a , and the value of F is 1.12 (surface elliptical crack). The final crack length is evaluated by the onset of unstable fracture. Using Equation (11), by substituting c K in place of K , we get ; (14) The material fatigue parameters used for the assessment of life were generated in the laboratory from specimens extracted from actual aero-engine components and tensile properties are taken from earlier published work [6] Table 3Properties of Bladed disc material Cyclic properties Values Modulus of Elasticity, E 160 GPa Paris constants C 3.5e-10 m 3.83 Fracture Toughness, c K 112.4 MPa√m Cyclic yield strength 676 MPa Cyclic strength coefficient, ' K 1179 MPa Cyclic strain hardening exponent, ' n 0.08 Fatigue strength coefficient, ' f  2139 MPa Fatigue strength exponent, b -0.17 Fatigue ductility coefficient, ' f ò 1.75 Fatigue ductility exponent, c -0.52 Table 4Effect of initial flaw size on bladed disc life Initial crack size (mm) N i (operating cycles) N p (operating cycles) N t (operating cycles) N t (hours) 0.10 3379 30 3409 10438 0.20 2122 15 2137 6451 0.30 1266 10 1276 3905 0.40 876 8 884 2706 0.45 786 7 793 2427 3. LDV Test This results from the FE modal analysis of healthy and cracked blades were validated by performing sweep test on the Laser Doppler Vibrometer setup. Blades with different crack depths,[Figure 12-14], are investigated. Electrodynamic shaker frequency is swept linearly from 10 Hz to 3000 Hz in 2.25 seconds. The value of 2 / p  for this test, is[14], 0.00014. For this value of 2 / p  peak occurs at 1.024 p  = , inferring that the recorded value of natural frequency using the sweep test is 1.024 times actual natural frequency. The FFT of the acceleration response is shown in Figure 15 for the healthy blade and in Figure 16 and Figure 17 for blades with 1.5 mm and 2.5 mm deep cracks, respectively. The measured and corrected values along with those obtained from FE simulation are listed in Table 5. c c K FS a  = (13)   2 f c max a a K FS = = c 

Figure 12 Healthy Blade

Figure 13 Blade with 1.5 mm deep crack

Figure 14 Blade with 2.5 mm deep crack

4. Modal Damping Estimation The methodology developed for the estimation of modal damping is shown Figure 18. This is implemented on a healthy blade and one with 0.5 mm deep crack shown in [Figure 19-21]. The Table 6 shows the modal damping ratio for healthy and cracked blade for second mode. The results shows a 5.2% change in the modal damping ratio due to the presence of crack.