Coal Composition, Coal Plasticity, And Coke Strength

Results of linear correlations of two to four variables by the least-squares method are presented. The variables include 63 parameters representing coal composition, coal plasticity, and coke strength. The composition parameters are on chemical analyses; the plasticity parameters, on three plastometer and dilatometer tests and four miscellaneous tests; and the coke-strength parameters, on three ASTM indices. Seventy-two to eighty-four plastic coals were used in the plasticity and carbonization tests, which together form the basis of this report; 43 nonplastic or weakly plastic coals were used in the coal-band studies. Plots of relations between volatile matter, carbon, and oxygen contents of 1,860 U.S. coals and the 84 coals used are shown. Similar plots between volatile matter and hydrogen or atomic H/C of the 84 coals are also included. Significances of regression coefficient signs for coal composition parameters in the significant correlations found are discussed. Based on these significances, the following relations are shown: (a) coke strength versus average aromaticity, or mineral matter content, of the parent coal; (b) characteristic temperatures of the plastometer and dilatometers, the free-swelling index, or the Gray-King type versus average hydroaromaticity of the coal; and (c) nontemperature parameters of the plastometer and dilatometers versus the physical properties of a certain group of compounds from thermal decomposition of coal. Similarly, significances between the positions of coals in the coal band and the aforementioned relations are also discussed. A structural theory of coal plasticity is proposed. Applications of the correlation results are discussed. Composition parameters, such as carbon, atomic H/C, and calorific value may be used as a basis for selecting coking coals for stronger coke. Plasticity parameters, such as the end temperatures of the plastometer and dilatometers and the free-swelling index, may be used for the same purpose. Based on the best correlations found, coke-strength indices may be estimated to within 4 to 8 units of the observed values; the end temperatures of the plastometer and dilatometer tests, 5° to 10° C.; and the free-swelling index, within 1 unit. Accuracies of all these estimated values are examined, based on the experimental errors of the data involved. Finally, estimations of plasticity parameters from one test with those from others are shown.