PSI - Issue 28

A.D. Evstifeev et al. / Procedia Structural Integrity 28 (2020) 2261–2266 / Structural Integrity Procedia 00 (2019) 000–000

2264

2

 For a given compute prediction errors � � � � � �� � �� � � �� � � 1 2 � �  Evaluate

� � � � ∑ � � � ���� �� � �� � � ��� and � � � � ∑ �� � � � ���� �� � �� � � � 1 2 � � 1 � ���  Put in order scalars � � � �� from smallest to biggest ones  Compute the rank ℛ� � of | � � �| in the ordering, where ℛ� � � 1 if | � � �| is the smallest one, ℛ� � � 2 if | � � �| is the second smallest and so on.  Return 1 if ℛ� � � � � , otherwise Return 0 4. Results and discussion All experimental data was treated according to methods described above. The confidence level was chosen equaled to 95% for � 100 and � � 5 . The results in table 1 show that the static strength generally grows with the increase of magnesium percentage in the alloys whereas incubation time value tends to decreasing and their dynamic strength properties is getting worse. It is clear that obtained intervals for the incubation time are relatively narrow and if its median values are supposed to be true then maximum estimation error cannot be greater then 30% with the confidence level 95%. Fig 2 and Fig 3 demonstrate results of the analysis for AMg3 and AMg4.5 alloys where red lines correspond to the analytical prediction of the incubation time approach for the boundary values of interval . All data points are in a good correspondence to theoretical curves and the incubation time criterion approved itself once again. Moreover, it can be concluded that AMg4.5 alloy has better dynamic strength properties than AMg3 since its average value of the incubation time a little bit greater. 5. Conclusions The implementation SPS-procedure to the structural temporal approach permitted to analyze experimental data on tensile tests under dynamic as well as static loads with given confidence level. The incubation time parameter responsible to dynamic strength properties can be also evaluated by the interval of admissible values under the lack of data points and weak assumptions about distribution of random noise. Table 1. Experimental and data analysis results. Material � AMg2 �12.5; 22.3� �9.1; 17.5� �14.0; 18.7� �7.6; 15.2� �6.8; 10.6� 188 236 235 320 359 AMg3 Amg4.5 AMg5 Amg6