13th International Conference on Fracture June 16–21, 2013, Beijing, China -7- Several observations can be made from the comparison in Fig. 3: • For the LCF and LCDKF cycles, both models give predictions in good agreement with experimental data, Bordet model being less conservative than the Beremin model. • For LCIKF cycles, both models underestimate the toughness after warm prestressing, but predictions from Bordet model are closer to the experimental data. • For all transient loadings, the scatter after warm prestressing is larger for Bordet model than for Beremin model. These results tend to indicate that, after warm prestressing, the simplified version of the model proposed by Bordet gives more accurate predictions than the Beremin model which is more conservative. To better understand and assess the physical hypothesis used in the models – and thus why do they lead to slightly different results - the evolution of mechanical quantities such as plastic zone size and stress/strain fields during the WPS cycles are investigated in more details in the following section. 3.2 Evolution of local parameters The basic assumption for the WPS effect to happen is that the size of the plastic zone at reloading is much smaller than the one at preloading. In other words, this means that the material keeps the memory of the prestressing at high temperature. Conversely, if the size of the plastic zone when reloading is much larger than the one at preloading, no WPS effect is expected. The evolution of the size of the active plastic zone as a function of the temperature is shown in Fig. 4 for LCIKF cycles. The size is normalized by the square of the thickness B2, which confirms the plane strain assumption. For all the transient, the size is roughly constant up to a temperature of about -75°C, and then decreases drastically. In the temperature range where the size of the plastic zone is large, the path is under the Master Curve, thus no cleavage is expected. When the paths goes through the Master Curve and above, plasticity is almost not active - except on a very small area close to the crack tip - which explains roughly why cleavage fracture was not observed for the LCIKF cycles. Note that the evolution of the active plastic zone size could have been guessed without F.E. simulations: under small scale yielding assumption plastic zone extent scales as (K/σY) 2 and K increases linearly with temperature in LCIKF experiments, while σY increases exponentially. Beremin and Bordet models both account for active plasticity, by integrating Weibull stress on the active plastic zone for the former, by considering nucleation of microcracks that depends on plastic strain for the latter so that the models will not predict cleavage fracture during the studied transients.
RkJQdWJsaXNoZXIy MjM0NDE=