A Statistical Approach in the Analysis of Geometrical Product Specification during the Cold Tube Drawing Process

Abstract

Cold tube drawing provides higher accuracy compared to hot approaches. The process can be used to reduce the dimensions of tubes, and depending on the reduction size, the wall thickness of these may be subject to changes. In the process, any form of variability provoked by external factors is highly sensitive, given that the resulting tubes are often the final step in tube production. This paper focused on the evaluation of the influence of pre-tube factors on key variables after the drawing process, i.e., the final roundness, outer diameter, and wall thickness of the tubes. For these purposes, a factorial design with fixed factors was implemented. It was also a goal to investigate if the single-pass type of drawing would guarantee good statistical results potentially leading to significant time and financial reductions. The measurements were executed in the machine ZEISS CenterMax. The statistical analysis took place on Minitab 19. The results prove that most factors, and their interactions, significantly impacted the response variables, leading the authors to understand that a single-pass approach would not properly work under the conditions defined for the experimentation. These results also allow for reflection on the causes and necessary measures related to lubrication, technological heritage, and quality that would impact the results themselves.