作者：MITSUO SUGAWARA;TETSUNORI IKEBE;KAZUTOSHI IWASHITA;
作者單位：Technical Research Laboratory, Krosaki Harima Corp. 1-1 Higashihama-machi, Yahatanishi-ku, Kitakyushu 806-8586, Japan
刊名：Taikabutsu Overseas
ISSN：0285-0028
出版年：2002-01-05
卷：22
期：3
起頁：173
止頁：178
分類號：TQ175
語種：英文
關鍵詞：
內容簡介Optimization of refractory formulations has been done mostly by empirical (trial and error) methods. Because there are so many factors involved in determining the optimum refractory formulation. It is important to use a procedure that is more efficient than trial and error. This paper discusses an attempt to apply a neural network method for determining the optimum composition for a dry mix formulation for the hot repair of converter linings. The main factors affecting fluidity and fast hardening (sintering) at elevated temperatures were included as important variables. Results of the field evaluation of a trial product were compared with the model established by the neural network. The physical properties predicted by the model agreed very well with the field results for the developed product.
所需耐材幣：0
