Automated Weld-Geometry Optimization of Tubular Joints for Fatigue Design Using Surrogate Models

As offshore wind farms move into deeper waters, jacket support structures increasingly rely on welded tubular joints whose fatigue life is governed by local stress concentrations. This paper introduces an automated, datadriven workflow with surrogate modeling to optimize weld-geometry detailing in tubular X-joints, reducing effective notch factors under manufacturing constraints. A pushbutton "design-simulate-learn-optimize" loop automatically generates Bézier spline weld seams in Rhino/Grasshopper, transfers geometry via a Python interface to ANSYS, and triggers standardized model setup, meshing, solving, and post-processing without manual intervention. Notch factors are evaluated using the Neuber/Radaj reference notch approach (recommended by IIW) to account for micro-support effects, while radial basis function optimization (RBFOpt/RBBFMOpt) surrogates guide the search efficiently. Optimized weld shapes could reduce notch factors by up to 56% and also reduce weld material. The proposed framework is generic and can be extended to other welded tubular joints and hollow-section components in offshore and civil infrastructure, enabling automated, data-driven weld-shape optimization already in the design phase and its seamless integration into parametric CAD/FEA-based digital design workflows.