Multi-Objective Cost and Quality Optimization of CFD Simulations Exploring the Pareto Frontier

By integrating Computational Fluid Dynamics (CFD) with Design of Experiments (DOE) techniques applied to hydraulics, it is possible to optimize systems by exploring the solution space and identifying critical performance factors. Multi-objective analysis, combined with the Normal Boundary Intersection (NBI) method, proves essential in identifying optimal solutions that balance performance goals. The study demonstrates that increasing the number of replicas improved mesh quality from 90.060% to 91.492%, although it also raised simulation costs, traditionally treated subjectively, highlighting the trade-off between objective functions. The analysis of the Pareto Frontier, with equal weights, resulted in 84.473% quality and a cost of R$ 175.07 for three replicas. It is concluded that the methodology was effective in identifying Pareto-optimal solutions, providing more robust decision-making support.