PSI - Issue 17

Waleed H. Alhazmi et al. / Procedia Structural Integrity 17 (2019) 292–299 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

296

5

the matrix and E c = 156.4 GPa if E m = zero. The measured value of the tensile strength of unnotched, i.e. smooth, specimen is equal to 2.9 GPa. This value lays between the minimum and mean values found in the technical data sheet of the supplier. The maximum elongation of CFRP is approximately equal to 1.87%, i.e. equals 110% the value found in the technical data sheet of the supplier. Based on the net stress analysis, i.e.  n = P/[(w-d H )t], there is no difference between the mechanical behavior of smooth and notched specimens, i.e. notch insensitivity. On the other hand, Huh and Hwang (1999)found a good agreement between the average stress criterion and the experimental data of circular notched quasi-isotropic, graphite/epoxy laminates. As shown in Fig. 3, the stress-strain responses were linear for the majority of the tests. The same trend was also observed by Plumtree and Sorensen (2002) in unidirectional carbon fiber-reinforced epoxy composites with v f = 60%. In all open-holed specimens, the crack initiated in the vicinity of the hole at a low very stress and then propagated parallel to the fiber direction, i.e. fiber-matrix splitting, and parallel to the applied load due to the transmission of shear forces between fiber and matrix. Although the specimen was broken into several strips at the middle, it still possessed strong damage tolerance due to the presence of bonded end tabs. The fiber matrix splitting apparently reduced the stress concentrations in the 0-degree plies near the hole and, as a consequence, improved the notch strength (Yan et al. 1999). The present experimental results support their finding and the stress concentration factor was equal one after fiber-matrix splitting. The numerical and theoretical results of SCFs in orthotropic plates containing a circular hole and subjected to tensile load are shown in Fig. 1. The numerical SCFs were calculated based on either longitudinal stress,  YY , or equivalent stress,  Eq. . The theoretical results are greater than the numerical results because Eq. 1 overestimates the value of K T  . The numerical solution is only true before any damages such as fiber-matrix splitting. Therefore, such traditional numerical and theoretical solutions are inappropriate candidate to predict SCF in the present CFRP plate. On the other hand, Tan and Kim (1990) found good agreement between the theoretical solution of SCF and experimental results of multidirectional graphite/epoxy composites.

GPa E 11 = 165 E 22 = 11 E 33 = 11 G 12 = 5.3 G 13 = 5.3 G 23 = 3.9  12 = 0.26

21

K Torth , Eq. 3 K Torth-net , Eq. 5

16

Based on

 Eq.  YY

Numerical

11

6

 13 = 0.5  23 = 0.5

Stress concentration factor

1

0

0.2

0.4

0.6

0.8

Hole diameter/specimen width, 

Fig. (1) Stress Concentration factors as a Function of  . In all fatigue tests of notched specimens, the cracks initiate early in the life of the specimens and propagate parallel to the fiber direction. The amount of fatigue damage depends on the number of cycles and the ratio of  f-max /  ult , as would be expected. There is a small degradation in the modulus of elasticity. This finding is supported by the experimental results found in Ref. (Hollaway and Leeming 1999). The review in Ref. (Hollaway and Leeming 1999) showed that, flaws in the fibres of a unidirectional CFRP specimen may cause fibres to break before matrix damage