PSI - Issue 71

Sudarshan Solanki et al. / Procedia Structural Integrity 71 (2025) 95–102

98

A two-dimensional axisymmetric model has been used for analysis of proposed sample design having solid cylindrical geometry with uniform gauge cross-section. Mesh convergence study has been carried out using h refinement method by taking 8 node quadrilateral elements. The smallest element sizes for coarse and fine meshes are 106 and 53 microns, respectively. The percentage difference between coarse and fine mesh is less than 0.1%. The strain amplitude along the axis of cylinder for coarse and fine meshes are typically compared in Fig. 3(a). This figure shows that mesh convergence has been achieved. The remote loading of pure axial and pure torsion has been applied on axisymmetric model with three different material regimes. The remote displacements (axial displacement for pure axial case and angular displacement for pure torsional case) are applied in such a way that the maximum peak equivalent strain amplitude at any location within the DMW joint is achieved as 1%, 0.75%, 0.5%, and 0.35%. Typical FE model representing different material regimes across DMW is shown in Fig. 3(b). 5. Results and Discussions 5.1. Strain gradient quantification The effects of various peak equivalent strain amplitudes of 1%, 0.75%, 0.5%, and 0.35% were investigated under both pure axial and pure torsion loading conditions. In both cases, strain amplitudes were analysed along axial and radial paths, taking into consideration different sections of the material such as the SS base, weld and LAS base regions.

(a)

(b) (c) Fig. 4. Variation of equivalent strain amplitude along (a) loading direction for at r = 0 & r = R with 1% peak equivalent strain amplitude; (b) loading direction at r = 0 for different peak equivalent strain amplitudes and (c) radial direction with peak strain amplitude of 1%.