Abstract
The difference in the degree of cure of the composite in an autoclave is one of the main characterization parameters of the uniformity of the degree of cure of the composite material. Therefore, it is very important to develop an effective method for predicting the difference in the curing degree of a composite autoclave to improve the uniformity of the curing degree of the composite materials. We researched five machine learning models: a fully connected neural network (FCNN) model, a deep neural network (DNN) model, a radial basis function (RBF) neural network model, a support vector regression (SVR) model and a K-nearest neighbors (KNN) model. We regarded the heating rate, holding time and holding temperature of the composite material’s two holding-stage cure profile as input parameters and established a rapid estimation model of the maximum curing degree difference at any time during the molding process. We simulated the molding process of the composite material in an autoclave to obtain the maximum difference in the curing degree as the test sample data to train five machine learning models and compared and verified the different models after the training. The results showed that the RBF neural network model had the best prediction effect among the five models and the RBF was the most suitable algorithm for this model.
Highlights
Composite materials have the advantages of high specific strength, high specific stiffness and designability of the mechanical properties of materials
Autoclave forming is one of the most common forming processes, which refers to a process method in which a single layer of pre-preg is stacked in a predetermined direction to place a composite material blank in a thermopressed tank and complete the curing process at a given temperature and pressure
In order to solve the above problems, this paper investigates different machine learning models, proposes a prediction method for the uniformity of composite materials in an autoclave forming based on machine learning and trains five models including the fully connected neural network (FCNN) model, the deep neural network (DNN) model, the radial basis function (RBF) neural network model, the support vector regression (SVR) model and the K-nearest neighbors (KNN) model
Summary
Composite materials have the advantages of high specific strength, high specific stiffness and designability of the mechanical properties of materials. Bogetti and Gillespie [2] conducted a two-dimensional (2D) cure simulation for thick thermoset composites using the finite difference method by which the transient anisotropic heat transfer equation coupled with cure kinetics was solved to predict the temperature and degree of cure distributions as a function of the autoclave temperature history. Blest [6] studied the resin flow, heat transfer models and simulations of the curing composites in an autoclave and found that the numerical simulation results were considered to be approximately valid compared with the existing test data. The innovation of this article lies in the application of traditional machine learning methods to the field of composite materials These models provide a new and effective method for the estimation of the maximum ∆α of a composite in an autoclave forming
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
