A PRELIMINARY STUDY OF CERTAIN CALIFORNIA ADOBE SOILS

Adobe is a term applied to certain California soils, which, when wet are very plastic and when dry are excessively cracked, the individual clods being hard and tough. Figure 1 shows a characteristic soil of this type in its dry state.

Adobe soils, because of their large volume changes under various moisture conditions, form poor subgrades for permanent highways. The present study was undertaken to secure data on these volume changes or amounts of expansion and contraction-as they are fre

quently called. The following table gives the county from which the various soils were received.

No information on adobe soils could be found in engineering publications available in the university library. It was necessary therefore, to proceed slowly and try out ideas and schemes before preparing a comprehensive program of investigation.

Since data on volume changes was the principal information desired, that work was given most attention. It was begun with actual volumetric determinations which were soon abandoned for linear shrinkage tests from which volumes could be computed. The linear shrinkage experiments were very satisfactory because it was possible to show moisture-volume relations for a range of values and also check determinations made by actual volume measurements. These tests were therefore made on all soils sent to the laboratory. The Butte County soil was then used to determine the effects of lime and sand on volume changes. When this work was well under way the moisture content, specific weights and sieve analyses were made for each soil. A few specimens were then prepared so that an idea of the tensile and compressive strengths of adobe could be obtained. The last experiments made were probably the most interesting because one served to substantiate, by actual expansion tests, the expansions computed from lincar shrinkage tests, while the other showed that confined adobe soil exerted a very high pressure when it absorbed water.

Expansion Actual.

METHODS OF EXPERIMENTATION

A cylindric specimen was cut, in a lathe, from a lump of soil, No. 12, from Ventura County. The apparatus is shown in Figure 2. The specimen of soil was fitted into the porous (Alundum) thimble and the metal plunger placed on the soil. The thimble was then put into a beaker of water and the raising of the plunger measured.

Expansion-Computed from linear shrinkage tests.

Tests that could be quite easily made and yet yield considerable information were required. The linear shrinkage experiments satisfied these requirements. Strings of mud about 12-inch by 12-inch by 30 inches were made up of a representative sample of each soil as received. Each sample was mixed to approximately the same consistency which required about 30%, by weight, of water. The exact amount of water used for each soil is shown on Plates III to IX inclusive. The molded specimens were placed in grooved boards and measurements between two gauge marks on each end of the string (the name given to these specimens) immediately begun. Three strings were made up of each soil; two were measured while pieces of the third were dried to determine moisture content corresponding to every measurement of length. The boards and strings of soil are in the background of Figure 3 together with four small pans with moisture specimens.

Expansion-Computed from volumetric tests.

Two sets of volume determinations were made using the soils after they had been passed through a 100-mesh sieve. In the first set the soil was mixed with about 30% water and molded by hand into cylindric specimens whose volume was approximately 2.0 cu. in. These specimens are in the left of Figure 6. Each specimen was weighed in air and again in kerosene. These original measurements together with similar measurements taken after the specimens had been dried in air for about a week and later in an oven, maintained at 100° C., to constant weight were used to determine the change in volume.

The second set of volume determinations were made with more plastic mud. The soils were mixed with about 60% water and poured into tin cups whose volume was about 23

Figure 3. Volume changes due to drying thin mud of adobe soil. Cup at right end refilled with water to show original appearance of specimens. Linear shrinkage tests in background.

cu. in. The exact volume of each cup had been previously determined. The specimens were weighed and set aside to dry. After being oven dried they were again weighed and the decrease in volume of the specimen determined by weighing the kerosene required to fill the cups after the specimens had become saturated. The cup experiments are in the foreground of Figure 3.

Expansion-Effect of lime and sand.

The experiments to determine the effect of lime and sand on the volume changes in adobe soils were carried out on the sample from Butte County. The specimens were the linearshrinkage strings previously described. Additions of 10, 20, and 40% sand and 1 and 5% lime were used. The tests were similar to those already described.

Expansive Force in confined adobe soil as affected by absorbed water.

The difficulty of confining wet adobe under pressure was clearly recognized before this experiment was begun. An apparatus which had been previously used to remove water (the soil solution) from very fine-grained soils-all passing a 200-mesh sieve-was set up as in Figure 5. It is shown in cross-section in Figure 4.

Powdered soil, Butte County adobe soil, was passed through a 100-mesh sieve and put into the cylinder. A pressure of 6,000 lbs. was applied. This load was required to compact the soil to the density of the clods as found. The load was decreased to 32 lbs. per sq. in. and the water was then poured into the funnel. It saturated the sand which completely surrounded the soil. A one-foot head of water was maintained on the specimen throughout the test. The increase in soil pressure or expansive force is shown in Plate XIII.

Plunger ground to fit cylinder..

Sand.

Adobe soil.......

Screen....

Water level...

Not to scale

Figure 4. Cross-section of apparatus shown in Figure 5.