摘要
针对金川矿区附近瓮福化工厂排放出的大量磷石膏废弃物,开展以磷石膏为主要激发剂材料的早强型充填胶凝材料研究。首先开展早强充填胶凝材料激发剂配方的正交试验,并以试验数据作为学习样本,建立神经网络模型进行训练,由此获得了胶结充填体强度与激发剂配方关系的隐含知识。在此基础上,进行不同配比的充填体强度预测;然后,利用预测结果建立激发剂配方的优化模型并求解,最后获得了早强胶凝材料最优配比:生石灰磷石膏、NaOH、芒硝和矿渣微粉的掺量分别为5%、30%、3%、2.5%和59.5%。相应3d、7d和28d胶结充填体强度达到2.05 MPa、3.42 MPa和7.87 MPa,满足金川矿下向分层充填法采矿对充填胶凝材料的早强要求。与矿用32.5R早强型水泥胶凝材料相比,开发的磷石膏基早强充填胶凝材料成本低,替代水泥用于金川充填法采矿,可降低充填胶凝材料50%以上的成本。
The research on filling cementitious material based on phosphogypsum with high early-age strength was carried out for a large amount of phosphogypsum waste discharge of WengfuChemical Plant near Jinehuan Mine. Jinchuan Mine utilized rod milling sand as aggregate and cement as cementitious material to conduct the underhand cut-and-fill mining. The high cost of fill- ing have a direct impact on economic benefit of mine and the in- ternational competitiveness of enterprises. The orthogonal exper- iment of activator formula of filling cementitious material with high early-age strength was conducted based on the experimental data as learning samples to build a neural network model for training, whereby obtaining implicit knowledge of the relation- ship between filling body strength and activator formula. The strength of cementitious filling body of different activator formula was predicted on the basis of the implicit knowledge. After that, a activator formula optimization model was built using the predic- tion results for its solution and the optimal proportion of filling cementitious material with high early-age strength was finally ob- tained ： proportion of quicklime, phosphogypsum, NaOH, mirabi- lite and slab were 5% .30% .3% .2.5% and 59.5% and the cor- responding compressive strengths for 3 days, 7 days and 28 days were 2.05 MPa, 3.42 MP and 7. 87 MPa respectively, meeting the requirement of filling mining with cementitious material. Compared with 32.5R cement, the cost of filling cementitious ma- terial based on phosphogypsum with high early-age strength is low. It can reduce more than 50~ of the cost of cementitious ma- terial that using cementitious material based on phosphogypsum with high early-age strength to replace cement.
出处
《化工矿物与加工》
CAS
北大核心
2016年第1期31-35,共5页
Industrial Minerals & Processing
关键词
金川镍矿
下向分层充填采矿
磷石膏
早强胶凝材料
正交试验
神经网络
Jinchuan Nickel Mine
underhand cut-and-fill min- ing
phosphogypsum
cementitious material with high early-age strength
orthogonal experiment
neural network
