Simple Linear Regression (SLR) Model for Re-Bending Behavior of a Non-Crimp Dry Thick Carbon Fiber Fabrics

Abstract

During manufacturing process, fabric such as the non-crimp dry thick fabric (NCF) is bent and re-bent many times until the fabric takes the desirable shape in the mold. Understanding the behavior of dry NCFs requires conducting experimental bending tests which is expensive due to the cost of the carbon fabrics. This paper aimed to model the bending behavior of NCF using simple Linear Regression Model in which the bending moment force is used as a response to displacement. The data modeled were obtained from testing three samples each sample was tested three consecutive times. A total of nine simple linear regression models were created. These nine models showed strong correlation between bending force and extension. The results also showed that the fabric becomes more flexible when it is subjected to re-bending process. Good performance of these models was confirmed using cross-validation method indicating that all presented models in this study were able to predict the bending behavior of non-crimp dry thick fabric.