Automated 3D Piping Routing Synthesis Considering Seismic Resilience: Part I— Mathematical Formulation and Layout Optimization Framework

Abstract

Automated piping routing methods face dual challenges in high-dimensional space planning and structural safety verification. Traditional workflows treat seismic qualification as a retrospective step, leading to inefficient design loops and costly backtracking. This study (Part I) presents a hybrid optimization framework that integrates deterministic A* pathfinding with a Multi-Objective Genetic Algorithm (MOGA) to automate seismic-aware 3D piping routing synthesis. To bypass the prohibitive computational cost of full-scale Finite Element Method (FEM) simulations during iterative optimization, a lightweight linear-response spectrum modal analysis is used as a surrogate physical indicator to rank candidate layouts. A stochastic midway point strategy is introduced during initial population generation to ensure topological diversity and bypass local minima. The framework was validated using a 32 x 60 x 20 m congested design space derived from the Onagawa Nuclear Power Plant blueprints with carbon-steel piping (STPT410). The optimization loop evaluated 10,000 unique candidates and converged in 46 minutes and 19 seconds, reducing the aggregate evaluation index by 93.2%. The results successfully characterized the non-linear trade-offs between material length, bend frequency, and dynamic seismic stress. Notably, the framework systematically identified configurations in which expanding pipe length and adding bends to lower structural stiffness successfully shifted the system's natural frequencies away from the resonance peak, thereby reducing peak seismic stress by up to 86.1%. This framework substantiates a transition from reactive verification to proactive, physics-informed layout synthesis.