Study On The Formula Prediction And Performance Of Glass Fiber Reinforced Sealing Gasket By Beater-addition Process

Glass fiber have a low price and excellent performance.However,in the field of non-asbestos sealing material by the beater-addition,there are not many reports on the research of glass fiber.In order to study the properties of hybrid glass fiber gasket materials,glass fiber was introduced in the gasket formula in this paper.It is expected that this research can be used to develop a non-asbestos sealing product with good performance and low price.The main work and achievements are as follows:(1)The silane coupling agent,PEO and concentrated sulfuric acid have been used for surface treatment of glass fiber,and worked in this paper.(2)In order to study the performance of these gaskets,the compressive property and recovery,tensile strength,density and ignition loss were experimented.Some of the prepared samples had been able to meet the technical requirements of these properties.And the multiple linear regression was used to study the correlations between the performance of gaskets and the formulation constituents.The conclusions can be obtained that glass fiber and QY composite cotton have a great influence on the performance of the gasket.(3)Under the same process conditions,the conventional properties,high temperature resistance and the price ratio of glass fiber gasket and mineral cotton fiber gasket were analyzed and compared.The results show that in the limiting temperature range,the glass fiber gasket has good heat resistance.But there is still backward in the compressive property and recovery,tensile strength,density.Its price is relatively low and the glass fiber has a certain application prospect in the field of sealing material.(4)The multi-objective optimization function was adopted to optimize the gasket material formula,and the optimized results,basically,achieved the desired target value.At the same time,the comparison between the target value and the measured value of the formula was obtained by using the optimization function.Although there are some differences in individual performance values,the comprehensive properties is better,and the optimized formula has a certain reference value.(5)These experimental data were used as the sample,and the BP neural network model of gasket was designed in this paper.Through the prediction of the network,when the number of neurons in the hidden layer is eight,the predicted values are close to the actual values.And the results can be accepted by taking into account of a few of training samples.