Joint Sizing Optimizer is an interactive tool that streamlines structural steel joint design—now in v2.0 as a Streamlit port of the revised PyQt5 desktop app. Users enter a maximum axial force and receive optimized dimensions (thickness, width, length, gap) for four joint types—Imposter, Axe, U, and Plate—alongside visual references and calculated Design Safety Factors (bolt shear, bearing, net-section rupture, block shear) aligned with CSA S-16. The experience is fast, guided, and defensible, with optional advanced bounds for deeper control.

Under the hood, the optimizer uses two well-established constrained solvers. The trust-region constrained approach builds a local model of the objective and explores within a safely sized region it trusts; that region expands or shrinks based on how well predictions match reality, and it respects nonlinear constraints throughout—great for tight, competing limits. Sequential Least Squares Programming solves a sequence of quadratic subproblems making steady, gradient-guided progress toward KKT-optimal solutions with bounds and inequalities. In practice, this pairing provides robust convergence without exposing users to algorithmic complexity. Version 2.0 is in active development with iterative optimization, refined constraint tuning, and upcoming support for custom joint configurations.