SHRINKAGE IN DRYING.
The amount that a clay will shrink in drying is expressed in per cents of the unit length or volume. In the first instance the shrinkage would be designated as linear shrinkage, and in the second, as volume or cubical shrinkage.
In Table I, page ??, will be found the percentages of both linear and volume shrinkage for several shale clays as determined by direct meas urement. It will be noted that the variation in linear shrinkage in 60 bricks of each clay is far in excess of reasonable limits. When the linear shrinkage varies from 32 to 133 per cent from the average, the data must be wholly unreliable. In presenting this data it is felt that the failure to produce more consistent results lies in part in the shortness of the shrinkage distance, and in part in carelessness of the operator. In marking the freshly made bricks a stencil devised by J. F. Krehbiel was used, so that initially the shrinkage lines were marked upon the brick with accuracy. In measuring the decrease in length of the shrinkage line after the bricks were dried, a vernier shrinkage scale was used that read accurately to the third place. The large variations in the results were therefore a surprise to the operator.
The volume shrinkage varied within fairly reasonable limits, but even here the variations are quite large considering the size of the bricks used.
It is felt that if in one case the variation could be only 0.5 per cent there ought not to be any excuse for a variation of 33.8 per cent in another or an average on all samples of 11 per cent.
Inasmuch as the volume shrinkage data proved to be the more accur ate of the two they were used as a basis on which to calculate* the linear shrinkages as shown in the following table: The linear shrinkage which probably is the more correct for that sample is underscored. In cases where there is not an underscored linear shrinkage, there is no possible way to judge which one is the most correct. In case the calculated practically agrees with the measured linear shrink age, both are underscored.
If the volume and linear shrinkages had been correctly measured, there would have been no discrepancy between the calculated and determined linear data. If any importance at all is to be attached to shrinkage data it is evident that extreme care should be exercised in their determination. When possible, the measured linear should be checked by calculation from the volume shrinkage and vice versa.