Multi-Dimensional Regression Models for Predicting the Wall Thickness Distribution of Corrugated Pipes

Area of Science:

• Materials Science
• Manufacturing Engineering
• Computational Modeling

Background:

• Corrugated pipes offer superior stiffness, flexibility, and material efficiency compared to rigid pipes.
• Increased production rates and rising commodity prices necessitate high-quality corrugated pipes with reduced material input.
• Corrugated pipe geometry and wall thickness distribution critically influence performance, with mold block geometry being key for optimization.

Purpose of the Study:

• To develop predictive mathematical models for corrugated pipe wall thickness distribution.
• To optimize mold block geometry and structure for reduced material usage and enhanced pipe properties.
• To overcome the time-consuming and costly nature of traditional experimental design and testing methods.

Main Methods:

• Development of multi-dimensional mathematical models using symbolic regression based on genetic programming (GP).
• Transformation of the blow molding problem into a dimensionless representation for analysis.
• Conducting a screening study to identify key parameters, followed by a numerically driven parametric design study.
• Utilizing data-driven modeling with a comprehensive dataset to derive novel regression models.

Main Results:

• Successfully developed models that predict wall thickness distribution as a function of mold geometry.
• Identified significant parameters influencing wall thickness through a screening study.
• Validated model accuracy using statistical error analysis and various metrics.

Conclusions:

• The developed models enable accurate prediction of wall thickness distribution in corrugated pipes.
• These predictions facilitate structural analysis and digital mold block design.
• The approach optimizes wall thickness distribution, leading to more efficient material utilization and improved pipe quality.