Scholarlink

<p>This paper presents a hybrid statistical and response surface regression modelling framework for predicting and minimizing burr formation in the sheet metal blanking process of stainless steel (SS316), copper (C11000), and aluminium (AA1100). Building upon the limitations of prior material-focused investigations, this study introduces a generalized and flexible modelling framework capable of accommodating diverse materials with varying mechanical characteristics. The experimental design employed a Taguchi orthogonal array to study sheet thickness and punch–die clearance at three levels. Blanking trials were performed on a 10-ton hydraulic press under dry cutting, with burr height measured using high-precision instruments for reliability. Two prediction frameworks were established: (i) a linear regression model employing forward stepwise selection, and (ii) a non-linear quadratic model based on Response Surface Methodology (RSM). Results indicate that a clearance of 5% and a thickness of 1.5&#xa0;mm consistently produced minimum burr heights across all materials. Non-linear RSM models demonstrated superior accuracy (R<sup>2</sup> &gt; 95%) and significantly reduced error metrics, particularly for steel where RMSE decreased by 49% compared with the linear model (R<sup>2</sup> = 82.25%). For copper and aluminium, linear regression achieved sufficient accuracy due to their relatively predictable material response. The originality of this work stems from its multi-material comparative evaluation integrated with combined regression–RSM modelling, an approach rarely reported in blanking studies. The proposed framework establishes reliable process optimization trends and demonstrates strong potential for integration with digital-twin-enabled intelligent manufacturing under the Industry 4.0 and 5.0 paradigms.</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

Taguchi-based experimental design and response surface modelling for burr prediction in multi-material sheet metal blanking process

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年1月8日

查看全文
  • Swapnil Bhoir,
  • Munna Verma,
  • Manoj Dongare

摘要

This paper presents a hybrid statistical and response surface regression modelling framework for predicting and minimizing burr formation in the sheet metal blanking process of stainless steel (SS316), copper (C11000), and aluminium (AA1100). Building upon the limitations of prior material-focused investigations, this study introduces a generalized and flexible modelling framework capable of accommodating diverse materials with varying mechanical characteristics. The experimental design employed a Taguchi orthogonal array to study sheet thickness and punch–die clearance at three levels. Blanking trials were performed on a 10-ton hydraulic press under dry cutting, with burr height measured using high-precision instruments for reliability. Two prediction frameworks were established: (i) a linear regression model employing forward stepwise selection, and (ii) a non-linear quadratic model based on Response Surface Methodology (RSM). Results indicate that a clearance of 5% and a thickness of 1.5 mm consistently produced minimum burr heights across all materials. Non-linear RSM models demonstrated superior accuracy (R2 > 95%) and significantly reduced error metrics, particularly for steel where RMSE decreased by 49% compared with the linear model (R2 = 82.25%). For copper and aluminium, linear regression achieved sufficient accuracy due to their relatively predictable material response. The originality of this work stems from its multi-material comparative evaluation integrated with combined regression–RSM modelling, an approach rarely reported in blanking studies. The proposed framework establishes reliable process optimization trends and demonstrates strong potential for integration with digital-twin-enabled intelligent manufacturing under the Industry 4.0 and 5.0 paradigms.