PSI - Issue 13

Ho-Wan Ryu et al. / Procedia Structural Integrity 13 (2018) 1932–1939 Author name / Structural Integrity Procedia 00 (2018) 000 – 000

1937

6

This rule is a superposition of Armstrong and Frederick (1966) nonlinear kinematic hardening rule. Nonlinear kinematic hardening rule with three back stresses has six parameters ( C 1 ,  1 , C 2 ,  2 , C 3 ,  3 ) which should be determined by a stabilized hysteresis loop. In this case, these parameters can be determined by fitting from the stabilized hysteresis loop. Total nine parameters for SA312 TP316L SS and CF8A CASS materials were determined for three different strain amplitudes with straightforward manner. The parameters are summarized in Table 2 and Table 3. In Fig. 5, the results of cyclic tensile simulation show that determined parameters can predict cyclic hysteresis loop precisely.

Table 2. Combined hardening parameters of SA312 TP316 SS.  o C 1  1 C 2  2 C 3  3

Q

b

0.4% 112 0.8% 69 1.2% 116

200824 2000 199960 2000 114783 2000

9717

103

6697 1696 1049

0 0 0

26 68

12.57

30303 287 29935 262

7.50 5.76

118

Table 3. Combined hardening parameters of CF8A CASS.  o C 1  1 C 2  2 C 3

Q

b

 3

0.4% 92 0.8% 134 1.2% 105

335210 2000 143276 2000 217389 2000

10770 68 39206 346 29298 228

6615 3064 1582

0 0 0

31 49 81

18.67 11.25

7.15

(a) (c) Fig. 5. Hysteresis loops of SA312 TP316 SS material obtained from strain-controlled loading test; (a) 0.4%, (b) 0.8% and (c) 1.2% strain amplitude. (b)

4. Results of cyclic C(T) simulations

The results of cyclic C(T) simulations were shown in Fig. 6 as load versus LLD curves. SA312 TP316 SS and load ratio R =-1.0 are described as a representative example. From the results of FE simulation, peak load points for each cycle in each tensile and compression region were connected as a line. These lines were compared with experimental load versus LLD results. Combined hardening parameters from three different strain amplitudes were used in cyclic C(T) simulations. Using  a versus LLD curves as input data, the effect of cyclic material parameters can be explained with resulting loads from simulation. Results of load versus LLD derived from each parameter set was shown in Fig. 6(a) for SA312 TP316 SS material. Among the three results, the result from 1.2% strain amplitude shows the closest to experimental data. But, it is predicted with higher load value than those of test in compression region. The size change of yield surface Q and ratchetting related parameter  3 can be modified for cyclic C(T) simulation. Parameter Q was determined from hysteresis loop, but effective strain amplitude can be different in cyclic fracture toughness test. And ratchetting parameter cannot be found from cyclic tensile tests, so it also should be determined from parametric study in cyclic