(1. 中南大學(xué) 材料科學(xué)與工程學(xué)院，長(zhǎng)沙 410083；
2. 中南大學(xué) 教育部有色金屬材料科學(xué)與工程重點(diǎn)實(shí)驗(yàn)室，長(zhǎng)沙 410083；
3. 金杯電工股份有限公司，長(zhǎng)沙 410083)
摘 要: 通過(guò)維氏硬度測(cè)量和透射電鏡(TEM)觀察研究冷拉拔對(duì)Al-Zr-(RE)合金組織與性能的影響。結(jié)果表明：鋁鋯合金中Sc、Er的添加可以有效細(xì)化晶粒，改善第二相的析出，且析出的彌散Al3(Sc, Zr)相能夠抑制再結(jié)晶，釘扎在冷拉拔過(guò)程中產(chǎn)生的位錯(cuò)阻礙位錯(cuò)運(yùn)動(dòng)，提高材料的硬度。在實(shí)測(cè)得到的維氏硬度值的基礎(chǔ)上，采用誤差反向傳播(BP)算法訓(xùn)練人工神經(jīng)網(wǎng)絡(luò)，建立以變形量和稀土元素添加量為輸入?yún)?shù)和維氏硬度為目標(biāo)函數(shù)的網(wǎng)絡(luò)。網(wǎng)絡(luò)訓(xùn)練值與實(shí)驗(yàn)值較吻合，相關(guān)系數(shù)R達(dá)到0.992 1，用建立的網(wǎng)絡(luò)進(jìn)行仿真，仿真的相關(guān)系數(shù)為0.979 3，證明了網(wǎng)絡(luò)的可靠性與良好的泛化推廣能力。
關(guān)鍵字: Al-Zr-(RE)合金；冷拉拔；人工神經(jīng)網(wǎng)絡(luò)；維氏硬度；顯微組織
（1. School of Materials Science and Engineering, Central South University, Changsha 410083, China;
2. Key Laboratory of Nonferrous Metal Materials Science and Engineering, Ministry of Education,
Central South University, Changsha 410083, China;
3. Gold Cup Electric Apparatus Co., Ltd., Changsha 410083, China）
Abstract:The effects of cold drawing on the microstructure and properties of Al-Zr-(RE) alloys were studied by the Vickers hardness measurement and transmission electron microscope (TEM) observation. The results show that the elements Sc and Er have the ability of refining grains and promote the precipitation of the Al3(Sc, Zr) particles. This dispersed precipitates can pin the dislocations forming during cold drawing and hinder the movement of dislocations, thereby improve the hardness of alloys. By measuring the Vickers hardness of different alloys under different deformations, an artificial neural network (ANN) based on the error back propagation is built to find the relationship of them. The results of ANN model have a good agreement with the experimental values. The correlation coefficient of observed values and training ones is 0.992 1, and the correlation coefficient of observed values and simulation ones is 0.979 3, showing a good generalization ability and outreach capacity.
Key words: Al-Zr-(RE) alloys; cold drawing; artificial neural network; Vickers hardness; microstructure
