PSI - Issue 19

8

Marc J.W. Kanters et al. / Procedia Structural Integrity 19 (2019) 698–710 Marc Kanters et al. / Structural Integrity Procedia 00 (2019) 000–000

705

Figure 8: (a) SN-curves for for various microstructures, where markers represent measurements, lines model fits using the analytical models (solid lines) and the Digimat material model (dashed lines). (b) Comparison of the Digimat material model and experiments for the various microstructures for . Properly capturing the influence of anisotropy is not trivial and can lead to errors in calibration. In general, samples with 0 o orientation are used to calibrate the in-axis properties of the composite, samples with 45 o orientation the shear properties, and samples with 90 o orientation the off-axis properties. However, the shell/core type of morphology in the injection moulded plaque (see section 2.3) will cause difficulties. For example, when performing the simulations on specimens with a 0 o orientation, the analysis will show the highest stress and shortest lifetime in the highly oriented shell. A similar exercise for the 90 o orientation will reveal that the lifetime is dominated by the high stress in the highly oriented core. As a result, the lifetime is dominated by the layer with fibers oriented in the loading direction for both the 0 o and 90 o specimens, albeit in the shell and core respectively. Therefore, the use of only 0 o , 45 o and 90 o orientated specimens will not allow proper identification of the failure indicator required to capture anisotropy. Although a reasonable fit on the three input curves could be attained, it will lack any predictive capability. A solution can be found in considering other or more orientation angles and maintaining a high level of detail for local fiber orientation during calibration is a prerequisite. Note that using an average orientation for the sample will not even allow identification of these issues. With the anisotropy captured, the load ratio dependence can be validated. The dependence on , as obtained in section 4.1.3, allows definition of the mean-stress sensitivity parameter, , that Digimat uses to capture the influence of mean stress and stress amplitude, � , on lifetime via: � � � ��� � ⋅ (1) The model predictions for different R-values for samples with 0 o and 45 o orientation for different R-values are compared with measurements in Figure 9a, showing a good description of both mean-stress sensitivity and anisotropy. To show the overall consistency of the material calibration, all coupon results are presented in Figure 9b, showing a good consistency between model and experiments for a wide range of load ratios and microstructures. Note that, even though the accuracy is very satisfying, it shows that for these anisotropic materials, due to the fitting compromise the overall precision that can be expected within commercial tools is a bit worse than ± factor 3 for simple coupon geometries.