A heat transfer process determines the thermal cycles in a work piece during the friction stir welding(FSW) and affects its microstructure and mechanical properties.A good understanding of the heat transfer process in the work piece is helpful for predicting the residual stress in the welding work piece and the hardness in the weld zone.A three-dimensional heat transfer model based on the tool shape and size for FSW is established in this paper. With the Ansys finite element software and the temperature data at the limited number of measurement points,the temperature distribution and change of 6 mm thick purple copper friction stir welding were analyzed with finite elements.Heat transfer at the bottom surface of the work piece and the temperature-dependent thermal conductivity of the purple copper were considered during the calculation.The calculation results are in good agreement with the experimental results.

Many results were obtained by plenty of quadrature-experiments on copperplate friction stir weld,which makes up of the training data of the ANN.The welding parameters(rotation speed,translation speed and pressure)were selected as the input and the tensile strength or the excellent of the joint (1represents a good joint and0represents a bad one)was selected as the output of the network.The network trained by the result of the quadrature-experiments with aberrated sample eliminated was used to estimate the quality of the joint.The friction stir welding ANN and user interface were established by MATLAB.It can be used to optimize the friction stir welding technical parameters.

搅拌摩擦焊(FSW)是近几年发展起来的新型摩擦焊接技术.铝合金搅拌摩擦焊接已成为国内外研究的热点,但对铜及其合金搅拌摩擦焊接的研究尚不多见.通过大量试验分析,优化了铜板搅拌摩擦焊接工艺.研究表明:工业纯铜具有良好的搅拌摩擦焊接性能,优化工艺可获得强度超过母材的搅拌摩擦焊接接头.