PSI - Issue 41

Abdelmoumene Guedri et al. / Procedia Structural Integrity 41 (2022) 564–575 Abdelmoumene Guedri et al. / Structural Integrity Procedia 00 (2022) 000–000

574

11

y t 

100

N



 

i (4) In which t and y are respectively the experimental value and predicted value of true stress; t̅ and y� are the mean values of t and y respectively; N the number of data sets. The ANN model resulted from the previous sections is used hereafter to predict the deformation conditions, which corresponds to test points and previous training points. As indicated in the above, high correlation coefficient would indicate a good prediction capacity of the model. Similarly, AAR is used as an alternative goodness fit test. The AAR value was found to be equal to 0.31% for all data sets. This low error would indicate the high accuracy of the ANN model for both testing and training data sets. Figures 13 show predicted and experimental flow stress values along with their correlation relationships. The best-fit line corresponds to the 45 degrees line and all errors between predicted and observed flow stress values are lower the 5%. Likewise, the resulting correlation coefficients between observed and predicted values are 0.9999 and 0.9998 for the training and testing data sets, respectively. These results would confirm, and therefore, validate the learning and generalization capacity of the ANN model. i i 1 N t  i ARE % 

Fig.13.Experimental versus predicted flow stress values.

5. Conclusions The minimum of the ductility is around 900 ° C, in agreement with the previous work (Ping-Hua, 1989) on alloys of high purity. This ductility in steels containing dispersed elements at the same time as sulfur and phosphorus probably depends on the competition between the damage of the grain boundaries and the restoration of these grains under the effect of dynamic recrystallization. In the presence of sulfur and phosphorus, the level of segregation increases with the decrease of the deformation temperature in the case of the precipitation treatment, and generates a weakening of the joints and a decrease of the ductility. In the  domain, the ductility in the temperature range concerned varies directly with the ferrite content. The embrittlement mechanism in this area is based on the relationship between the presence of primary ferrite at grain boundaries and necking. The presence of a few islands of ferrite at the austenitic grain boundaries slightly lowers the ductility. This ductility decrease observed at 800 °C is attributed to the formation of a thin primary ferrite film following the precipitation treatment and causing a concentration of the deformation in the vicinity of the grain boundaries. On the other hand, at 750°C the ferrite film is wider and by a relaxation effect of the concentration of the deformation, improves the ductility. The hardening effect observed on the curves of evolution of the maximum stress as a function of temperature is due to the presence of the precipitates in the matrix and at the grain boundaries. These precipitates dispersed on the matrix obstruct the plastic deformation and thus promote an increase in the tensile strength of our material. The temperature and the rate of deformation modify the ultimate tensile strength and ductility. Both characteristics grow monotonically as the rate of deformation increases. The effect of temperature is more complex in the two-phase