PSI - Issue 38

S. Spanke et al. / Procedia Structural Integrity 38 (2022) 220–229 Author name / Structural Integrity Procedia 00 (2021) 000±000

229

10

Table 2. Stresses of specimen and RVE system

Model

Stress across the fiber

Scaling (reference is the specimen)

S 22 [MPa]

Specimen

49.14 83.53 79.80 82.94

1

Matrix

1.70 1.62 1.69

Fiber

Interface

In previous tests, a static strength of 82.42 MPa was determined on pure resin test specimens for the Rim 935 matrix system. For the 90° oriented specimens with a fiber volume fraction of 68%, a strength of 52.92 MPa was determined. Both tests took place on the 4-point bending test. The strength di ff erence between the pure resin specimens and the 90° oriented CFRP specimens is caused by notch stress e ff ects of the fibers. When comparing the stress S 22 = 83.53 MPa of the matrix (see Table 2) with the strength value of the pure resin specimens 82.42 MPa, there is a di ff erence of 3.28%.

4. Conclusions

The investigation with RVE systems with symmetrical and random fiber distributions has shown clear di ff erences. Table 1 clearly shows that the stress paths for the arbitrary fiber distributions are significantly more pronounced than for symmetrical RVE systems. Only for load Step 3 in the fiber direction, there was almost no stress di ff erence between symmetric and arbitrary systems. The numerical determination of the e ff ective homogenized material parameters of the homogenized layer with symmetric and random fiber distributions was validated with tests on a 4-point bending test. The numerical deter mination of the sti ff ness E 22 with the symmetric RVE models showed slightly lower moduli on the random fiber distributions. For the numerically determined E 11 moduli in the fiber direction, the di ff erence between the symmetric and the arbitrary fiber distribution RVE models was much smaller. The scattering of the emodules of the tests could also be shown in the simulation (see Figure 7). All in all, it can be stated that the calculation procedure for the numerical determination of the e ff ective homogenized material parameters provides valid results. In the analysis of highly stressed areas of components with RVE systems, the notch stress e ff ect of the fibers under transverse fiber stress of the unidirectional layer was simulated. At 3.28%, the matrix stress at failure of the CFRP specimen is very close to the strength value of the pure resin specimens. In the mapping of the matrix stress for highly loaded areas of CFRP components with RVE systems, a high potential is seen to simplify the fatigue strength assessment workflow for strength. Currently, the orthotropic strength space of the unidirectional layer must be set up with direction-dependent WoÈhler lines. Furthermore, these WoÈhler lines are only valid for the investigated fiber volume fraction. The idea is to evaluate the stress of CFRP components at the micromechanical level of matrix stress. The amplitude of the matrix stress is then evaluated with the WoÈhler line for the resin system. The advantage is that only one WoÈhler line is needed for the resin system and this can be transferred to other fiber volume fractions.

References

Barbero, E. J., 2008. Finite element analysis of composite materials. CRC Press, New York. Ge, W., Wang, L., Sun, Y., and Liu, X., 2019. An e ffi cient method to generate random distribution of fibers in continuous fiber reinforced composites, in ª Polymer Composites º. Hill, R., 1963. Elastic properties of reinforced solids: Some theoretical principles, in ª Journal of Mechanics and Pysics of Solids º. Pergamon Press Ltd., Great Britain, pp 357-372. Melro, A. R., Camanho, P. P., and Pinho, S. T., 2008 Generation of random distribution of fibers in long-fiber reinforced composite, in ª Composites Science and Technology º. pp 2092-2102.