Scholarlink

<p>Fused Deposition Modeling (FDM) offers transformative advantages in manufacturing yet consistently faces a significant challenge: simultaneously optimizing for mechanical performance, surface quality, productivity, and cost. Traditional optimization methods often address isolated objectives and lack physical insight into the obtained solutions, limiting understanding of trade-offs and parameter interactions crucial for robust industrial performance. To address these limitations, this work develops a physics-informed, multi-objective optimization framework that integrates both computational and experimental analyses. The present study develops a hybrid optimization framework that integrates modeling, evolutionary optimization method, and multi criteria assessment. It systematically investigates key FDM process parameters through a Taguchi experimental design, modeling critical performance indicators (tensile strength, surface roughness, total cost, and build time) via symbolic regression. The framework employs the Non-dominated Sorting Genetic Algorithm II (NSGA-II) to generate Pareto-optimal solutions, which are subsequently ranked using the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS). The optimized parameter sets were experimentally verified under different operating scenarios, and the resulting trends were physically justified based on process mechanisms such as interlayer bonding. The framework successfully developed highly accurate predictive models where the lowest was at 89.63% for surface roughness and highest at 98.89% for build time, validating its utility as a robust decision-support tool by reliably capturing complex relationships between process parameters and multiple performance indicators. The predicted optimal solutions showed good agreement with the experimental results, with deviations generally below 15% for cost, tensile strength, and build time, and between 10% and 17% for surface roughness. It not only identified the dominant process parameters but also effectively demonstrated how multi-objective optimization can achieve a balanced trade-off among often-conflicting objectives, establishing a scalable and physically grounded approach for diverse industrial needs. It should be noted that the approach, as conducted, can also be applied to other additive manufacturing processes and materials.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Physics-informed multi-objective optimization of fused deposition modeling

  • 期刊论文

  • 发表日期：

    2026年3月2日

查看全文
  • Charisse Chan,
  • Madeleine Bonn,
  • Binyan Xu,
  • Hussien Hegab

摘要

Fused Deposition Modeling (FDM) offers transformative advantages in manufacturing yet consistently faces a significant challenge: simultaneously optimizing for mechanical performance, surface quality, productivity, and cost. Traditional optimization methods often address isolated objectives and lack physical insight into the obtained solutions, limiting understanding of trade-offs and parameter interactions crucial for robust industrial performance. To address these limitations, this work develops a physics-informed, multi-objective optimization framework that integrates both computational and experimental analyses. The present study develops a hybrid optimization framework that integrates modeling, evolutionary optimization method, and multi criteria assessment. It systematically investigates key FDM process parameters through a Taguchi experimental design, modeling critical performance indicators (tensile strength, surface roughness, total cost, and build time) via symbolic regression. The framework employs the Non-dominated Sorting Genetic Algorithm II (NSGA-II) to generate Pareto-optimal solutions, which are subsequently ranked using the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS). The optimized parameter sets were experimentally verified under different operating scenarios, and the resulting trends were physically justified based on process mechanisms such as interlayer bonding. The framework successfully developed highly accurate predictive models where the lowest was at 89.63% for surface roughness and highest at 98.89% for build time, validating its utility as a robust decision-support tool by reliably capturing complex relationships between process parameters and multiple performance indicators. The predicted optimal solutions showed good agreement with the experimental results, with deviations generally below 15% for cost, tensile strength, and build time, and between 10% and 17% for surface roughness. It not only identified the dominant process parameters but also effectively demonstrated how multi-objective optimization can achieve a balanced trade-off among often-conflicting objectives, establishing a scalable and physically grounded approach for diverse industrial needs. It should be noted that the approach, as conducted, can also be applied to other additive manufacturing processes and materials.