PSI - Issue 68

Andreas J. Brunner et al. / Procedia Structural Integrity 68 (2025) 1266–1272 Brunner et al. / Structural Integrity Procedia 00 (2025) 000–000

1270

5

debonding, fiber breaks), see Sause et al. (2012). Relevant mechanisms are matrix cracks and, possibly, fiber-matrix debonding. If interlaminar delamination resistance is similar to fiber-matrix adhesion toughness, both mechanisms may occur simultaneously or change back and forth during delamination propagation. SEM imaging of the fracture surfaces might then yield information on the relative contributions of the two mechanisms.

(a)

(b)

Fig. 1. (a) AE signal attenuation curve determined from lead-pencil breaks applied on the GF-EP1 laminate at increasing distance from the sensor. (b) Scanning confocal microscope image of a GFRP fracture surface after a Mode I test

Table 4. Fracture surface increase from roughness measurements for GF-EP2 under different loading modes Loading Mode Remarks

Average fracture area increase (%)

Standard deviation ( % )

Mode I tensile opening Mode I tensile opening Fixed Ratio Mixed Mode I/II

180 182 180

25

Light microscope (10 measurements)

-

Single measurement with confocal microscope

25

Light microscope (10 measurements)

Mixed Mode I/II bending

185

18

Light microscope (10 measurements)

Mode II in-plane shear

184 182

13 21

Light microscope (10 measurements) Light microscope (10 measurements) only

Average

Fractography with SEM as well as surface roughness measurements with a light microscope indicate that effective fracture surfaces are larger than those calculated from delamination length times specimen width. Areas around 590 x 400 µm 2 for each measurement were inspected with a light microscope (type Keyence VHX-7000/ VHX-7100). Averages of about 180% increase in fracture surface (Table 4) are from ten different spots. Inspection of a Mode I fracture surface with a scanning confocal microscope (type Keyence VK-X1000/VK-X1100) on an area around 260 x 100 µm 2 at higher resolution confirmed the values (Figure 1b). SEM images of fracture surfaces from AS4/PEEK laminates after Mode I tests, see Jar et al. (1991) or Gao et al. (2001) indicate significant roughness, even though not quantified. Hence, surface area corrections increasing the estimates will apply for the AS4/PEEK as well, however,

quantification is not possible. 3.3. Estimates with corrections

In an attempt to improve the "simple" estimates, the fracture area and the number of AE signals have been corrected. On one hand, only signals with sources located in the process zone are counted, and, on the other, the estimates shown in Table 5 for GF-EP2 are increased by the specific corrected area value for each loading mode. As in Table 3, the corrected estimates roughly agree within one standard deviation. Again, the largest difference is observed between the two Mixed Mode I/II test set-ups. For GF-EP2, the average estimate (around 380 micrometer) amounts to roughly one