Studies Of The Variability Of Feldspathic Sands Refined From The Arkansas River

During the early 1960's, a method was developed for refining sand from the Arkansas River to produce a feldspathic product that was proposed for use in making colorless glass. Because of the need for chemical uniformity, a cooperating glass manufacturer specified certain ranges of percentages for each chemical constituent of the sand. These specifications were met by fulfilling the single requirement that alumina vary by no more than 20.1 percent from the specified value. This translated to a total variation of only about 1 percent in the feldspar content of the mixed quartz-feldspar products. Because the feldspar was lighter than the quartz, the average density of refined sand varied depending on composition. Densities, determined to five significant figures, of several samples were correlated with their chemical analyses, and the correlation could then be used with density determinations to more reliably indicate compositions than by chemical analyses only. The effects of several variables on the composition of refined sands were studied. There was a regular variation in the proportion of feldspar with screen size. A series of samples with their natural size distributions were more nearly uniform in composition than the same sands after screening and recombination. The screening and recombination process was done so that the sands had identical size distributions. Variations in the refining showed little effect on composition. A pile of flow-back sand from a construction sand washer was found to be severely segregated. However, samples taken from many points in the river near the sand plant and from the flow-back discharge pipe over several weeks very nearly met the uniformity specification. If care was taken to avoid segregation either before or during processing, it should be possible to meet the uniformity specifications for a feldspathic glass sand.