This study aims to investigate the behavior of alkali activated mortar,which is made of naturally available magnesium silicate as source material.For magnesium silicate,ultrafine natural steatite powder(UFNSP)is used ...This study aims to investigate the behavior of alkali activated mortar,which is made of naturally available magnesium silicate as source material.For magnesium silicate,ultrafine natural steatite powder(UFNSP)is used as the primary source of binder,and the activation is initiated through the alkali liquid which is proportioned in various combinations of silicate to hydroxide ratio(Na_(2)SiO_(3)/Na OH)ratio,and this ratio in this study varies from 1 to 3.The UFNSP is calcined at two difierent temperatures,700 and 1000℃.The mortar mix is proportioned as 1:3 between powder and the fine aggregate,and the mortar is prepared with hydroxide molarity(M)of 10 M.The mortar is cured for 48 hours at 60℃and the compressive strength was studied.All the mix were studied for its microstructural behavior along with compressive strength.The mix proportion of the mortar,and the results obtained through microstructural characterization were combinedly formed as input for artificial neural network(ANN)predictive modelling.The model is designed to predict the compressive strength,which is trained through Bayesian regularization algorithm with varying hidden neurons of 7 to 10.This experimental and predictive study shows that the strength is influenced by both Na_(2)SiO_(3)/Na OH ratio and calcination process.And the ANN is influenced by mainly calcination temperature and uncorrelation occurs in selected samples of 1000℃calcined UFNSP mix.展开更多
文摘This study aims to investigate the behavior of alkali activated mortar,which is made of naturally available magnesium silicate as source material.For magnesium silicate,ultrafine natural steatite powder(UFNSP)is used as the primary source of binder,and the activation is initiated through the alkali liquid which is proportioned in various combinations of silicate to hydroxide ratio(Na_(2)SiO_(3)/Na OH)ratio,and this ratio in this study varies from 1 to 3.The UFNSP is calcined at two difierent temperatures,700 and 1000℃.The mortar mix is proportioned as 1:3 between powder and the fine aggregate,and the mortar is prepared with hydroxide molarity(M)of 10 M.The mortar is cured for 48 hours at 60℃and the compressive strength was studied.All the mix were studied for its microstructural behavior along with compressive strength.The mix proportion of the mortar,and the results obtained through microstructural characterization were combinedly formed as input for artificial neural network(ANN)predictive modelling.The model is designed to predict the compressive strength,which is trained through Bayesian regularization algorithm with varying hidden neurons of 7 to 10.This experimental and predictive study shows that the strength is influenced by both Na_(2)SiO_(3)/Na OH ratio and calcination process.And the ANN is influenced by mainly calcination temperature and uncorrelation occurs in selected samples of 1000℃calcined UFNSP mix.
