RF.T.ATION OF VOLUME SHRINKAGE TO WATER IN EXCESS OF THAT REQUIRED TO FILL THE PORES.
It would seem that if the volume had been determined at regular in tervals as the bricks lost their mechanical water by evaporation, the per centage up to the time that the brick reached its maximum shrinkage, would stand in closer relation to the volume shrinkage than does the total mechanical water and volume shrinkage. It is not known what value such a test would have, but it would probably be considerably more than is the determination of total mechanical water alone.
In Table XI is shown the percentage by weight of water that would be required to fill the pores of bricks made from Iowa clays and that which is in excess of the "pore water." These clays were ground until they would pass through a 40-mesh sieve,' then wetted with water and thoroughly wedged. Grinding the clay until it would pass a 40-mesh sieve would reduce the size of the larger grains, and to some extent break down bunches of grains by force that would not have been affected by the water used in wedging. The data is of interest on this account in connection with the problem of shrinkage.
In making Table XI Byers' and Williams figures for porosity,* specific gravityt and water of plasticity/ were taken, and the data calculated as follows: If porosity, or volume of pore space, is 29.77 per cent in a unit volume there would be 0.2977 parts by volume of pore space, and 1.0000-0.2977 or 0.7023 volumes of clay. On the assumption that the pore space is filled with water and the specific gravity of the clay is 2.3 1, there would be 0.2977X1.00=0.2977 parts by weight of water, and 0.7023X2.34=1.6434 parts by weight of clay, or pressed as per cents-15.3 and 84.6 per cent respectively of water and clay. This 15.3 per cent of water then is the amount of water by weight that would be required to fill the pore spaces in a brick that would weigh 100 at the time when all the particles have become fixed or arranged in the exact position that they will maintain during the remainder of the drying period. This, it is assumed, would give the weight of water that
remains in the pores of the bricks at the time the clay has reached its maximum air shrinkage. This amount of water subtracted from the amount required to develop plasticity would give, if the foregoing as sumption is correct, the amount, the amount of water required to lubri cate the particles sufficiently to cause a state of mobility which we have learned to designate as plasticity.
Fig. 10 shows that there is some reciprocal relation between the amount of water in excess of that required to fill the pores of a dried brick, (as given in Table XI) and the volume shrinkage.
In Table XII are shown the calculations on the Illinois clays, designed to bring out the same facts given in Table XI. In this table, however, the amount of hygroscopic water is given in each case so that it can be reckoned in as part of the mechanical water, if so desired. It must be borne in mind, however, that the amount of water calculated as being in excess of that required for filling the pores does not in any way include the hygroscopic water. The hygroscopic water is not added in with the water of plasticity, because there is some doubt as to just where and how the clay retains that water on drying. It is supposed to be held either in or between the grains, and does not greatly exceed the amount (on ac count of the natural humidity of the air) that the powdered clay would retain as moisture.
The porosity data used are those given in Table I.
In the above table there is the same indication of a reciprocal relation between the "excess water" and volume shrinkage as noted in the case of the Iowa clays and shown in Fig. 11. We have here then the promise of a means of obtaining analytically a line on the drying behavior of a clay other than volume shrinkage taken alone.