PSI - Issue 14

Sanjeev M Kavale et al. / Procedia Structural Integrity 14 (2019) 584–596 Sanjeev M. Kavale, Krishnaraja G Kodancha, Nagaraj Ekbote / Structural Integrity Procedia 00 (2018) 000–000

586

3

T 11 -Stress

T 11 T 33

T 33 -Stress SENB Single Edge Notch Bend CT Compact Tension 2D 2 Dimensional 3D 3 Dimnesional FE Finite Element P Applied Load σ Applied Stress S

Span (S=4W for SENB specimen)

Stress Triaxiality Factor Young’s Modulus Strain in z-direction

h E

ε 33

J -Integral

J

2. Finite Element Analysis In this work initially IF steel is considered for the analyses. The Young’s Modulus of the material considered is 197000 MPa, yield strength is 155 MPa and Poisson’s ratio being 0.30. The dimensions of the SENB and CT specimen are taken as width ( W ) = 25.4mm (1 inch), height ( H ) = 2W = 50.8mm and the crack length a = 12.7mm or a/W = 0.50. The geometric dimensions of the specimen are taken according to ASTM E1820-17A. A series of 2D and 3D FE analysis have been conducted to study the variation of Stress Intensity Factor on SENB and CT Specimens using ABAQUS 6.14. Guian et al. (2016) have concluded that the J integral method and interaction integral given by ABAQUS provide consistent magnitudes of Stress Intensity Factor when compared with the ASTM standards and thus are the most appropriate methods to determine accurate K values for both 2D and 3D. Computations were carried out considering only one half of the specimen geometry due to symmetry. The 3D analysis domain is discretized using 20 node quadratic brick finite elements (C3D20R) using reduced integration. This kind of element is used by Kim and Son (2004) and Kudari and Kodancha (2014), for 3D analyses.

Fig. 1 (A) SENB specimen and (B) CT specimen

The model is generated by taking the origin at the crack tip and the fine mesh is used in the region around the crack front to achieve better results. The meshed model for SENB and CT specimen along with the boundary conditions are shown in the Fig. 1. The number of elements in the mesh ranged from 6980 to 7800 and number of nodes ranged from 31603 to 35225 for various specimen thickness considered. The symmetrical boundary conditions have been imposed along the ligament of the specimen due to half symmetry of the model. For SENB specimen, the boundary condition is applied where the specimen sits on the roller while actual testing. The