Online Analysis Method of Cement Raw Materials Based on Fourier Transform Infrared Spectroscopy
Hu Rong,Liu Wen-ging,Xu Liang,Jin Ling,Yang Wei-feng,Wang Yu-hao,Hu Kai,Liu Jian-guo
DOI: https://doi.org/10.3964/j.issn.1000-0593(2020)01-0041-07
2020-01-01
Spectroscopy and spectral analysis
Abstract:Timely analysis of the key components in cement raw materials is critical to the quality control of cement products. While on-site manual sampling and sample preparation are required for current analytical methods, they lead to a problem on timeliness. To quickly and safely determine the four key components of Fe2O3, SiO2, CaO and Al2O3 in raw meal of cement samples, a quantitative analysis method based on fourier transform infrared (FTIR) spectroscopy was developedin the paper. The theoretical basis on FTIR technology to determine the composition in cement raw materials was discusse dat first. Cement raw materials are complex multicomponent mixtures of minerals and rocks, which are mainly made up with iron raw materials (such as limonite) , siliceous raw materials (such as quartz) , calcium raw materials (such as calcite) and aluminum raw materials (such as beryl) and so on. These minerals and rocks have broad characteristic bands in the visible and near-infrared spectrum with many overlapping bands, and the intensities of the characteristic bands are low. Therefore, multivariate calibration was used for quantitative analysis. The corresponding experimental system was designed and built to analyze the composition content in raw meal of cement samples subsequently. The samples were 60 ground cement raw meal samples with different contents of key ingredients, which were provided by the cement manufacturers. The compositions of samples covered four key oxides of Fe2O3, SiO2, CaO and Al2O3. The diffuse reflectance spectra of the samples were collected by the experimental platform. X-ray fluorescence (XRF) analysis was used to determine the content of each oxide component in the sample as the reference values. The quantitative analysis models of Fe2O3, SiO2, CaO and Al2O3 were next established by partial least squares method (PLS). The sample set was first divided into a calibration set and a prediction set in a 7 : 3 ratio by Kennard-Stone algorithm. The wavenumber range of 4 000 similar to 5 000 cm(-1) was selected in the PLS model with 520 spectral elements in total. A regression model was established between the spectra of 42 samples and their contents of Fe2O3, SiO2, CaO and Al2O3 in the calibration set. According to the condition of Q(h)(2)>= 0. 009 75, seven factors were selected to establish the final quantitative analysis model. In the FTIR quantitative analysis model of the four oxides of Fe2O3, SiO2, CaO and Al2O3 the correlation coefficients between the calibrated four oxide contents and the content measured by XRF analysis respectively were 98. 49%, 98. 03% 98. 18% and 99. 24%, and the root mean square error respectively were 0. 04, 0. 22, 0. 26, and 0. 08. The model had a high accuracy of calibration. The quantitative analysis models of the four components were used to predict the four component content of samples in the prediction set, and the predicted values were compared with the reference values measured by XRF analysis lastly. The quantitative analysis models of the four components for Fe2O3, SiO2, CaO and Al2O3 had a high prediction accuracy with the prediction correlation coefficients of 91. 35%, 91. 50%, 91. 57%, 94. 67% and the predicted root mean square error of 0. 08, 0. 45, 0. 54, 0. 26, respectively. The foundation of rapid quantitative analysis of components in cement raw materials based on the fourier transform infrared spectroscopy during the cement production control process was established.