The Application of the Computer-aided Design in BDP Flame Retardant PC/ABS Formulation

This paper introduces the study on the Application of the Computer-aided Design in BDP Flame Retardant PC/ABS Formulation.As for the fire-retardant materials such as BDP/PC/ABS,its research method is still limited to the traditional formulation design method based on the experiment which requires so large amount of repetitive experimental work that it consumes a lot of manpower and material resources and can not meet the requirements of the design of the halogen-free flame retardant materials' formulation in efficiency.In order to solve these problems,based on the previous experimental results,computer-aided design is applied.According to the theory of Perceptron Network and BP algorithm,the modes of input,output and hidden layers are determined in the network of formulation design of BDP/PC/ABS and BDP/SiO2/PC/ABS.The activation function of each layer is Sigmoid function.The molecule content of the three parts,BDP,PC/ABS and SiO2,is used as three neurons of the input layer respectively.There are three neurons in the hidden layer.It is a single-objective model with one neuron in output layer-the Limiting Oxygen Index(LOI) or the Elongation.After using nine pairs of samples,respectively as the training samples and test samples,to train and test the network,when the output is set to LOI,the forecasting error was within ± 4%;when it is set to the Elongation,forecasting error was within ± 5%,which can meet the requirements of practical recipes.