PSI - Issue 50

Alexander Eremin et al. / Procedia Structural Integrity 50 (2023) 73–82 Alexander Eremin / Structural Integrity Procedia 00 (2019) 000 – 000

81 9

significant values. Bottom extensometers (3- 3’ and 4 - 4’) showed tensile strains while upper (1 - 1’ and 2 - 2’) changed insignificantly. Total compressive strain according to extensometer 0- 0’ increased. 3.3. Discussion Previous tests according to ASTM D6641 Standard Test Method for Compressive Properties of Polymer Matrix Composite Materials Using a Combined Loading Compression (CLC) Test Fixture orthotropic CFRPs with a [0/90] 5S layup showed the strength of 657±81 MPa and 63.7±6.9 GPa modulus. CAI testing in the present paper showed much lower properties: 8.55 J specimen strength is 253.4 MPa and modulus is 47.1 GPa. 17.1 J and 25.65 J demonstrate the strength and modulus of 173.2 MPa and 46.26 GPa, 192.0 MPa and 45.7 GPa correspondingly. Firstly, this lower strength is partially due to the presence of impact damage. Secondly the higher strength of the orthotropic laminate versus quasi-isotropic one results in higher possibility of buckling even with the modified CAI testing device. Critical buckling stress can be calculated for elastic plate using the equation presented by Rees (2009):

2 b h  2

D

(1)

K





buckling

where h is a plate thickness, b is a plate width, K is a buckling coefficient ( K =5.5 for the present case), and D is a stiffness that is calculated as:

3 2 12(1 ) Eh  

(2)

D



where E is elasticity modulus and ν is a Poisson ratio. Critical buckling stress equals to 240.5 MPa, with a 0.37% of the failure strain assuming fully elastic behavior. In order to have a purely compressive failure the strength of the tested material should be lower than the critical buckling stress. With nearly equal values buckling is expected. However, due to geometry and clamping variations and non-regularities these values may vary. In the presented paper the first specimen failures at 253 MPa that is close to the predicted, however, the longitudinal strain is slightly higher – 0.62%. This is most likely associated with a possible non-elastic behavior that increases strains before failure. Moreover the presence of manufacturing flaws changes the stressed behavior as well. The presented specimen contains the impact damage that can be a reason of nonlinearity. It can be seen that the elastic modulus of the first specimen (as well as for 2 other specimens) is much lower: 47.1 GPa versus 63.7 GPa. This along with a DIC data confirms that buckling starts nearly from the initial of loading leading to deflection of the plate and reduction of its stiffness. The second and third specimens with larger impact damages are fractured at much lower stress that is due to the reduction of the load bearing capacity. It is important to note that regardless of the sizes of the damage the deflection due to buckling for all specimens starts increase more rapidly after nearly 0.15% of longitudinal strain. Similar modulus and the same position of this infliction point for different specimens allow concluding that the buckling behavior in the first half of the loading is mainly governed by the size of the plate and not the damage severity. Besides the general shape of the deformed surface obtained by DIC local strain distributions can be evaluated. It is shown that optical extensometers demonstrate some points where the mechanism of straining changes. Bending due to buckling occurs in such a way that upper extensometers (1- 1’ and 2 - 2’) are inherent to the reduction of compressive strains, i.e. the tensile strain due to bending compensates initial compression. Bottom extensometers (3- 3’, 4 - 4’) are less susceptible to an increase of compressiv e strains due to bending. The difference in the behavior of upper and bottom is due to small asymmetry in the clamping of the plate. The influence of partial fracture on the behavior of the plate should be separately analyzed. This is observed in the third specimen after nearly 0.5% of compressive strains. In this case partial delamination of the right part of the specimen occurs as the jump has been revealed by the right-side extensometers. The second partial fracture is much