High-strength steels (HSS) are increasingly used in advanced engineering applications due to their exceptional mechanical properties. However, accurately predicting stress fields and localized yielding in welded joints—especially at sharp geometric features such as weld toes and notches—remains a significant challenge, particularly in the inelastic region. This thesis addresses this challenge by developing and validating an analytical framework based on the Strain Energy Density (SED) method for modeling localized structural behavior in HSS welded joints.

The foundational model extends a linear elastic formulation by den Besten to account for post-yield behavior. Three analytical approaches—Neuber’s rule, Stowell’s generalized approach, and Glinka’s SED method—were critically assessed for the suitability of extending den Besten's equation, with Glinka’s method selected for its ability to capture localized plasticity while maintaining computational efficiency. The model assumes localized plasticity surrounded by elastic behavior, and its predictions were verified against two bounding solutions: Irwin’s ideal elastic–perfectly plastic lower bound and an extended linear elastic upper bound. Results consistently fell within these bounds, supporting the model’s physical accuracy.

Numerical validation was also carried out using finite element analysis (FEA) in ANSYS on a Partially Penetrated Double-Sided (PPDS) T-joint configuration. Material modeling incorporated true stress–strain data, realistic weld geometry, and boundary conditions representative of actual welded joints. The simulation results showed strong agreement with the analytical predictions, particularly in the region of localized plasticity near the weld toe where the error was limited to up to about 3%.

This study confirms that the SED-based analytical method offers a practical and accurate approach for assessing welded joints with sharp notches—regions prone to stress singularities. The framework is computationally efficient, physically grounded, and broadly applicable across welded structures. Academically, it supports future research into refining boundary definitions, material models, and mesh optimization techniques. Practically, it enables safer and more efficient designs in HSS applications by providing engineers with reliable tools for evaluating post-yield structural behavior.