(The job used as a basis for above table is given as a season's work of six miles of 18-foot pavement.)

PART II

Adobe and Clay Subgrades

4. SUBGRADES

4.1

The heavy adobe and clay subsoils of California are one of the most prolific sources of road failure today in this State.

Failures on the State System

The most notable of the earlier failures on the State Highway System, partly due to heavy adobe subsoils, were those sections on the coast highway north of Los Angeles in what is known as the Calabasas and Conejo Grade areas. These sections consisted of a lean 1:22:5 concrete, four inches in thickness, the slab being laid directly upon the adobe subgrade. It has been stated that a sand cushion between the adobe and concrete was used in at least one instance as an experiment.

These road sections, completed in 1914 and 1915, showed signs of early distress. The characteristic longitudinal cracks appeared in the slabs and these gradually multiplied and widened until many of the slabs were mere irregular segments of broken concrete with openings varying from two inches to four inches in width.

Efforts to repair the slabs consisted of breaking out and widening the long lines of parallel jagged breaks and filling with new concrete. The results were unsatisfactory and the sections, while not immediately disintegrating under traffic, were unsightly and rough and considered as initial failure areas. Subsequently, initial failures of more or less severity occurred in other sections of the State, due to the same cause.

Sections of the State roads laid on the adobe soils in Alameda, Sonoma, Solano and Contra Costa Counties followed in the wake of the Calabasas failure. The majority of these sections showed signs of distress and failure soon after they were opened up to traffic. By numerous replacements and expensive maintenance, the failure areas have been kept from complete dissolution, but, in general, although the average age of the pavements in question is about 4% years, many of the sections are in a condition of premature decay, with but a short span of life before the necessary reconstruction.

Among the later failures, from the same cause and in the face of all those preceding, are sections in Yolo, San Luis Obispo, Santa Barbara, Colusa, Glenn, and again in Alameda and Sonoma Counties. These are among the most spectacular failures in the State. Some of these sections were completed as late as 1919, approximately five years after the initial failures in the Calabasas and Conejo Grade areas.

Such is the brief but destructive history of the adobe and heavy clay subsoils on the State Highway system. Of the two soils adobe predominates to a marked degree, and due to its properties of swelling or expanding when absorbing moisture, followed by shrinking or contracting as the moisture evaporates, it is one of the most treacherous and difficult of subsoils to handle in road construction. Adobe apparently predominates in California over other states. of the western slope. Its ravages in this State have strewn both the State Highway and many of the individual county road systems with partial or complete road failures.

In the tabulation under chapter 5.02 hereinafter, the high percentage of failures on adobe subgrades is shown. While it is an open question in some instances as to whether poor concrete coupled with heavy traffic or adobe subgrade is the principal cause of the failure, there can be no doubt under any condition that adobe has been a contributing and probably the MAJOR cause in every instance. The summary is valuable in that it gives a general idea of the distressed concrete areas on adobe and heavy clay subsoils irrespective of traffic conditions, weakness of design or integrity of the construction work.

It is of interest to note that the sum of $2,730,600.00 has been expended by the California Highway Commission in the construction of four-inch concrete roads on bad adobe and similar subgrades, where partial and complete failure has already taken place. During the long period over which this expenditure was taking place, no comprehensive investigation, research or experimentation was being made to determine a solution for this particular problem, although early failures must have been of grave concern.

Some scattered and minor attempts at "adobe-proof" construction were attempted in a number of sections, but these lacked the initiative and support of concerted organization action and are therefore not classed as conclusive or comprehensive. They were too limited in extent and entirely lacked comparative methods for guidance, and are therefore of relative value only. Reconstruction of portions of adobe failure areas has already commenced and more must follow each year, and many miles of the State Highway system in adobe sections remain to be built. Failure to carry out adequate investigation and experimentation will prove as costly in the future as it has in the past, and it is apparent that any expenditure in this research will prove both wise and profitable and no doubt lead to a practical and probably economical solution of the problem.

An examination of adobe subsoils and distressed pavements on the State Highway system leads to certain conclusions with respect to the peculiarities of this type of soil, its action under varying moisture content, and the probable cause of rupture of road slabs viewed from theoretical considerations and practical field observations.

One of the most noticeable features of adobe subsoils is the almost universal tendency to crack a concrete sidewalk or road slab longitudinally. This often occurs along the center line of pavement or within the middle third area. Theoretically, the center longitudinal cracking might be explained as follows:

Case I: A concrete road slab rests on a well rolled and compacted semi-dry adobe subgrade. During the rainy season the side ditches of a low fill section flow full. Horizontal capillarity causes gradual movement of moisture from the side ditches toward the center of fill. At the same time vertical capillarity results in rapid upward movement in heavy soils, moistening the entire subgrade uniformly under the concrete road slab. This uniform moisture content causes uniform swelling of the adobe subsoil. Consequently it is fair to assume that uniform pressure is exerted on the bottom of the concrete slab over the entire width of span.

If we assume reactions due to the weight of the concrete slab plus the frictional resistance due to earth or rock shoulders and consider a uniform load on the slab due to uniform adobe soil pressure, considering simple beam action, the point of maximum moment would be at the center of the slab while the maximum sheer would be at the two reactions. In this case, the bending moment would govern and the slab would rupture at the center of the span, which is just what occurs in practice in a large percentage of initial adobe failures. The weakness of the above theory is found in assumed uniform moisture content and uniform pressure, but more particularly in possible reactions due to slab weight and friction of shoulders. It is, of course, apparent that if reactions are considered as due to the weight of slab alone, the uniform pressure of the adobe soil would probably lift the entire segment of concrete, causing transverse rather than longitudinal cracks.

Case II: In connection with the theory of moisture movement in soils, recent Government tests (Public Roads, Vol. 3, No. 28) have demonstrated somewhat conclusively that vertical capillarity is, in general, more rapid than horizontal capillarity in heavy soils. Under such a condition Case No. 1 fails. Due to the more rapid action of vertical over horizontal capillary water, unequal moisture content of subgrade under slab would result, and in consequence the subgrade pressure on the slab at the sides would occur in advance of the middle third section. If this hypothesis is true, the force exerted by the earth pressure on short sections three or four feet from each edge of pavement in advance of the middle section, would