A Multi-parameter Optimization Framework for Casting Quality: Beyond Riser Geometry

Casting defects such as shrinkage porosity, hot tearing, and misruns continue to impose significant economic burdens on foundry industries worldwide, with rejection rates for complex castings sometimes exceeding 15%. While riser geometry has traditionally been the primary lever for defect control, emerging evidence suggests that a holistic, multi-parameter approach encompassing pouring temperature, mold permeability, alloy composition, and solidification rate can substantially outperform single-variable optimization strategies. This study introduces a comprehensive Multi-Parameter Optimization Framework (MPOF) for casting quality control that integrates Taguchi design-of-experiments, finite-element solidification modeling, and response surface methodology to simultaneously optimize nine independent process variables. A total of 81 experimental runs were conducted using an aluminum–silicon alloy (A356) in green-sand molds, with porosity index, surface roughness Ra, and Vickers hardness as quality metrics. Results indicate that pouring temperature (contribution ratio: 28.3%) and mold permeability (21.7%) exerted the greatest influence on porosity formation, while alloy silicon content (19.4%) most significantly affected hardness distribution. Optimal parameter combinations achieved a porosity index reduction of 63.5% relative to baseline conditions and improved ultimate tensile strength by 18.2%. The MPOF revealed complex interaction effects between cooling rate and gating system design that had previously been masked in single-variable studies. Validation trials confirmed model predictions with a mean absolute error below 4.1%. This framework provides foundry engineers with a systematic, data-driven methodology for quality optimization that transcends the conventional focus on riser geometry alone, offering applicability to diverse alloy systems and mold configurations.