Predicting and Optimizing the physical-mechanical properties of epoxy/rubber/nano CaCO3 system using Taguchi approach, ANN and ANFIS methods

In this paper recycled tire rubber powder, epoxy resin, glass fiber mat and CaCO3 nano- powder are mixed and molded at different conditions. The mold temperature and other parameters are calculated and conducted at three different levels of A,B,C using taguchi method which yielded a L9(34) array. Thereafter, tensile strength, impact and failure tests were measured using ASTM standard test method. We predicted the physical-mechanical properties, modulus, impact and also the pressure strength of the samples using artificial neural network and ANFIS methods. Comparing the obtained results with experimental ones showed the least root mean square errors (RMSE) and the best regression (R2). Therefore, the predicted results obtained through our method are very well adopted with the experimental data

کلیدواژه ها:

recycled tire rubber ، composite ، physical-mechanical tests ، Caco3 nanopowder ، Taguchi ، ANN ، ANFIS ، optimizing and predicting