PSI - Issue 2_B

Yidu Di et al. / Procedia Structural Integrity 2 (2016) 632–639 Author name / Structural Integrity Procedia 00 (2016) 000 – 000

636

5

    

3 2

p

ε

d

d

,

0





 

 

d



(8)

f

p

inc

0,

0

where, d f and d p are parameters that controls the evolution of the damage value, and p ε

is the magnitude of the

3 2

plastic strain. When the damage D reaches a critical value D crit , the correspond FEM element is deleted and the material loses its capacity of carrying stress.

3. Material and experimental set-up

The material investigated in this work is a bainite steel S690Q. The samples are manufactured in two geometries: smooth round bar (RB) and notched round bar (NRB), as shown in Fig. 1. The cyclic tests were performed on a servo-hydraulic tension-compression machine. During the cyclic tests, a strain gauge with a measurement length of 10 mm was attached to the samples and two load amplitudes, +/-0.2 mm and +/-0.4 mm, were applied. The tests were performed in quasi-static condition with the loading rate of 2 mm/min.

Fig. 1. Samples geometry used for cyclic tests (a) Round bar sample (b) Notched round bar samples

4. Parameter calibration and failure prediction The proposed damage-coupled model is implemented into a VUMAT subroutine. All the simulations in this work were performed under the environment ABAQUS/Explicit. To reduce the consumption of computational resources, only the part with strain localization is modeled. And symmetric FE models were built for smooth round bar and notched round bar samples, as shown in Fig. 2. At the critical area, where the strain localizes, the mesh size is set as 0.15 mm×0.15 mm×0.15 mm to eliminate possible mesh size effect.