known, however, that the friction angle measured in plane strain tests will be appreci- 

 ably greater than that determined from triaxial tests on cohesionless materials. Gen- 

 erally, the difference in values depends upon density, and will be found to decrease with 

 decreasing density. Therefore, it can be concluded that 35° is a reasonable lower limit 

 for the value of the friction angle for the materials investigated in this study. (As 

 will be noted from subsequent discussions of the stability analyses, it appears that a 

 value of 38° for the friction angle more nearly complies with the conditions for most 

 of the failures.) 



A complete series of tests that would give an accurate determination of the criti- 

 cal void ratio-confining pressure relationship was not conducted; however, volume change 

 measurements from both direct shear and triaxial tests indicated that, for the rela- 

 tively shallow depths to the failure planes in most cases, the critical void ratio is 

 in the range of from 0.55 to 0.65. The significance of this is discussed more fully in 

 the conclusions of the failure investigations. 



— 6 — 3 



Permeabilities in the range from 10 to 10 cm. /sec. were measured in tests 

 wherein sample size, maximum grain size, and void ratio (density) were varied. Few of 

 these tests adequately simulated the loose conditions and grain-size distributions 

 actually existing in the field, and thus the measured permeabilities are probably lower 

 limits for the actual field conditions. Nevertheless, these materials are capable of 

 good water tightness and due to compaction and rapid disintegration they are undoubtedly 

 continually decreasing in permeability both in road fills and in the natural state. 

 This presumed reduction in permeability over a period of time is important in several 

 ways. First, infiltration capacity is reduced, thus tending to increase overland flow 

 and scour, or erosion; in the absence of vegetation, the damage is many times greater. 

 Second, culverts and other facilities designed to handle surface water based upon rea- 

 sonably permeable fills at the time of construction may become inadequate in later years 

 as the quantities of surface water increase. Third, for a given quantity of subsurface 

 flow, one or both of two factors must increase in response to a loss in permeability. 

 Either the cross-sectional area of flow must increase, hence the phreatic surface must 

 rise closer to the ground surface, or else the hydraulic gradient must increase, thereby 

 creating higher seepage forces. A further danger from an increase in hydraulic gradient 

 is that of piping, or the washing away of soil at a point of emergence. When this 

 occurs, for example at the toe of a fill slope, the piping action causes removal of 

 soil which is necessary for support of the remaining material upslope, thus leading to 

 possible mass failure. 



In connection with the prognosis of this investigation, a comparison of the in- 

 place densities (table 2) and the results of the moisture-density test results (fig. 7) 

 reveal the significant disparity between measured field conditions and the attainable 

 conditions . 



STABILITY ANALYSIS 



Neither the field and laboratory procedures nor the types of analyses performed in 

 this study were very complex or unique. The ordinary method of slices, assuming circu- 

 lar failure surfaces, was the predominant procedure followed in most of the analyses. 

 A variation of this procedure was programmed for a small digital computer and used for 

 a more thorough search for the critical failure arc for some of the fill slopes; slices 

 of differential thickness were assumed in order to derive a closed-form solution for the 

 safety factor. The infinite slope stability theory, as developed by Taylor, 2 was used 

 to analyze the natural slope failure, and was also applied to some of the shallow 



2 Donald W. Taylor. Fundamentals of soil mechanics. New York: John Wiley $ Sons, 

 Inc. , 700 p. , illus . 1948. 



12 



