( 1. 中南大學(xué) 粉末冶金國(guó)家重點(diǎn)實(shí)驗(yàn)室，長(zhǎng)沙 410083;
2. 中南大學(xué) 機(jī)電工程學(xué)院， 長(zhǎng)沙 410083)
摘 要: 針對(duì)喂料粘度模型參數(shù)求解和現(xiàn)有模流分析軟件無(wú)擬合功能的問(wèn)題， 引入Cross-WLF七參數(shù)模型對(duì)MIM中非牛頓流體流動(dòng)過(guò)程進(jìn)行了研究， 提出了自適應(yīng)快速遺傳算法擬合該模型參數(shù)， 開發(fā)了粘度模型參數(shù)擬合求解器， 得到了W-Ni-Fe高密度粉末喂料和316L 不銹鋼喂料粘度模型的7 個(gè)參數(shù)， 擬合結(jié)果的復(fù)合相關(guān)系數(shù)分別達(dá)到0.998 489和0.998 200。 研究結(jié)果為高密度類零件和不銹鋼類的質(zhì)量預(yù)測(cè)、 模具和工藝參數(shù)優(yōu)化設(shè)計(jì)提供了必須的材料數(shù)據(jù)。
關(guān)鍵字: 金屬粉末注射成形； 遺傳算法； 參數(shù)擬合； 粘度模型
（ 1. State Key Laboratory of Powder Metallurgy,
Central South University, Changsha 410083, China;
2. School of Mechanical and Electrical Engineering,
Central South University, Changsha 410083, China）
Abstract: For solution of feedstock viscosity model parameters and shortage of regress function in current molding analysis software, the flow process of Non-Newtonian feedstock fluid was firstly studied by using Cross-WLF Seven-Parameter model and a Quick Self-adaptive Genetic Algorithm was built and the viscosity model parameter fitting solver was realized. The seven parameters of Cross viscosity model of W-Ni-Fe and 316L stainless MIM feedstock were fitted, the multicorrelation coefficients (R) of fitting results are 0.998 489 and 0.998 200, which means its accuracy is very high. Then the necessary material data was obtained for quality forecast of mold core part made of stainless steel and heavy alloy tungsten ball, and for optimal design of mold and process parameters by this way. It establishes the foundation of material database of MIM.
Key words: metal powder injection molding; genetic algorithm; parameters fit; viscosity model
