Scholarlink

Injection molding is a complex process where effective mold design is essential to avoid defects like incomplete parts, flash, sink marks, weld lines, air bubbles, warping, and shrinkage. Proper mold design helps minimize these defects, but it is challenging due to the number of variables across stages such as plasticization, filling, packing, and cooling. This study focuses on optimizing the cooling phase of injection molding using numerical simulation tools (Moldex3D) to evaluate the impact of design variables on process performance. Due to the complexity and multi-objective nature of the problem, Artificial Intelligence (AI) techniques, including data mining, Artificial Neural Networks (ANN), and Multi-Objective Evolutionary Algorithms (MOEAs), were used to explore solutions effectively. The optimization aimed to improve the thermal efficiency of conformal cooling systems, critical for ensuring part quality and minimizing cycle time. A case study with a cylindrical part and multiple cooling systems initially defined 34 objectives, including temperature gradients, cycle time, and defects. These objectives were reduced to four using Principal Component Analysis (PCA) to facilitate optimization. The results showed significant improvements in temperature uniformity and defect reduction, confirming the effectiveness of integrating AI-based optimization techniques with numerical modeling for advanced cooling system design in injection molds.

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

Methodology for Designing Injection Molds: Data Mining and Multi-objective Optimization

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • António Gaspar-Cunha,
  • João Melo,
  • Tomás Marques,
  • António Pontes

摘要

Injection molding is a complex process where effective mold design is essential to avoid defects like incomplete parts, flash, sink marks, weld lines, air bubbles, warping, and shrinkage. Proper mold design helps minimize these defects, but it is challenging due to the number of variables across stages such as plasticization, filling, packing, and cooling. This study focuses on optimizing the cooling phase of injection molding using numerical simulation tools (Moldex3D) to evaluate the impact of design variables on process performance. Due to the complexity and multi-objective nature of the problem, Artificial Intelligence (AI) techniques, including data mining, Artificial Neural Networks (ANN), and Multi-Objective Evolutionary Algorithms (MOEAs), were used to explore solutions effectively. The optimization aimed to improve the thermal efficiency of conformal cooling systems, critical for ensuring part quality and minimizing cycle time. A case study with a cylindrical part and multiple cooling systems initially defined 34 objectives, including temperature gradients, cycle time, and defects. These objectives were reduced to four using Principal Component Analysis (PCA) to facilitate optimization. The results showed significant improvements in temperature uniformity and defect reduction, confirming the effectiveness of integrating AI-based optimization techniques with numerical modeling for advanced cooling system design in injection molds.