PSI - Issue 2_B

Tommaso Pini et al. / Procedia Structural Integrity 2 (2016) 253–260 Author name / Structural Integrity Procedia 00 (2016) 000–000

258

6

Fig. 4. Volumetric strain (a), crazing and/or cavitation (b) and shear (c) contributions to longitudinal strain vs. longitudinal strain for EI matrix

Yield stress vs. time to yield curves at different temperatures and rates are shown in Fig. 5 along with the master curves obtained shifting the data along the logarithmic time axis.

Fig. 5. Yield stress vs. time to yield isothermal curves (larger symbols) and relevant master curves (smaller symbols) at the reference temperature of 23 °C for E (solid symbols) and EI (hollow symbols) matrices Solid lines are power laws fittings.

4.3. Fracture tests

Results from double torsion tests, performed at different rates and temperatures, are shown in Fig. 6 for E and EI matrices. Master curves, obtained by shifting the data along crack speed axis, are reported also. Fig. 7 shows the shift factors obtained from small strains, yield and fracture data. They are all very similar, only the shift factors relevant to fracture of E matrix are slightly different. The dependence of fracture toughness on crack propagation speed is different for the two matrices. In the case of E matrix, an increasing trend is observed. Following Williams and Marshall’s approach as reported in (Bradley et al. (1998)), fracture toughness is equal to ( ) Ic c y G t α δ σ = (10) where δ c is the crack opening displacement, considered constant, and σ y (t α ) is the yield stress at a time comparable with that necessary for the crack to grow across the process zone. Considering the size of the process zone given by Dugdale’s crack tip model and in the hypothesis that relaxation modulus vs. time and yield stress vs. time to yield can be described by power laws with exponents m and n respectively, fracture toughness dependence on crack propagation speed follows a power law:

n m n

Ic G a ∝ 

1

+ −

(11)