PSI - Issue 39

Riccardo Cappello et al. / Procedia Structural Integrity 39 (2022) 179–193 Author name / Structural Integrity Procedia 00 (2019) 000–000

183

5

3. Experimental setup 3.1. Experimental setup and tested specimens

The testing machine used in this work is a servo-hydraulic MTS 810, equipped with a 100 kN load cell. Two different Single Edge Notched Tension (SENT) specimens made of AISI 304L stainless steel have been tested, to activate a Mode I fatigue crack propagation from the notch. The two samples have identical geometry, and the dimensions are reported in Figure 2. Both samples have been tested applying a sinusoidal load having a frequency of 15 Hz and ranging from 0 to 9 kN, with a loading ratio R = 0. Sample #2 differed from Sample#1 for having been subject to an overload after about 2 mm of crack growth, which produced an extended plastic area producing compression stresses ahead of the crack tip, which slowed down the successive growth. Both samples have been tested until the crack length reached approximately half of the specimen width. One face of the sample has been sprayed with matt black paint to enhance and uniform emissivity and monitored by means of a cooled sensor IR-camera FLIR X6540sc, with a set integration time of 759 µ s. Frames were acquired over time windows (TW) of 30 s, and at a frame rate of 200 Hz. The IR camera was positioned such to obtain a spatial resolution is 0.158 mm/pixel.

90°

thickness = 3,7

8

W = 39,48

L = 150

Figure 2 – Sketch and dimensions of the tested specimens.

A Nikon D5100 reflex camera equipped with a macro lens NIKKOR Micro 105 mm has been used to capture high resolution images from the opposite face of the sample and optically track the physical position of the crack tip. 4. Results & Discussion 4.1. Sample #1 4.1.1. Thermoelastic first harmonic signal The FH thermoelastic amplitude maps are obtained via the LSF filtering approach presented earlier in section 2.1. Figure 3 shows the isopachic maps at the 5 th (137280 cycles), the 21 st (368250 cycles) and the 36 th (503445 cycles) acquisition. The isopachics reproduce a typical cardioid-like shape at the crack tip as this advances. The FH amplitude data are employed for the fitting of the Williams’ stress solution and for the evaluation of the crack tip, as described earlier. The results of the ODLSF procedure are the contour plots showed in Figure 4. Two main aspects are pointed out: (i) the values of maximum and minimum radii describing the fitting collection area change as the crack advances, since the extent of the area ruled by the singularity increases and (ii) the coefficient of correlation is always very high (i.e. R 2 >0.97), and there is a very good superposition between the experimental data and the Williams’ model. The ODLSF procedure allowed a determination of the SIF change for each acquisition taken, which will be used together with the evaluation of the crack tip to characterize the parameters of a Paris’ law (see section 4.1.4).