Ch. 5] SAMPLING AND TESTING 99 



Sampling and Testing 



The several fundamental mechanical properties of soils that have 

 just been described serve as a point of departure for much of soil 

 mechanics theory. In practice there are many ways of applying soil 

 mechanics to quantitative soil problems. Each set of geological con- 

 ditions, as well as each type of construction, demands separate con- 

 sideration and a different treatment. In general, however, soil in- 

 vestigations follow a pattern of sampling— > testing —> analysis. The 

 prediction of soil behavior that is the result of this procedure is there- 

 fore dependent on (1) the representativeness of the samples, (2) the 

 pertinence of the testing, and (3) the pertinence of the mechanical 

 theory used in the analysis. 



Soil samples are of two types: (1) undisturbed samples, in which the 

 soil is removed as an integral lump, thus preserving its intergranular 

 relationships; and (2) disturbed samples, in which no attempt is made 

 to preserve the structure of the soil. Much thought has gone into the 

 design of undisturbed samplers for use in boreholes. A large variety of 

 types have been built, most of them based on a removable sampling 

 tube with a minimum of side friction and with some type of valve 

 arrangement to prevent loss of the sample on withdrawal. 



Soil testing is of two basic types: field testing and sample testing. 

 Field testing is the measuring of a soil property directly in the field 

 without isolating a sample. The driving of penetration cones, test 

 piles, and other resistance devices into the soil to determine strength 

 or bearing capacity is an example of field tests. Results from such 

 tests are generally empirically applied. Field testing is much more 

 common in the countries of northern Europe, where widespread uniform 

 soft Quaternary deposits prevail, than in the United States. 



Sample testing can be divided into three categories: (1) classifica- 

 tion tests for index properties; (2) empirical properties tests — direct 

 application; (3) basic properties tests — indirect application. The 

 second and third types are simulative tests. 



The common classification tests employed in the United States 

 include such procedures as mechanical analysis for grain-size dis- 

 tribution and liquid and plastic limit, to mention only a few. The 

 principal value of classification tests is for correlation and record. The 

 accumulation of index data, such as plastic index or grain-size dis- 

 tribution, when tied in with observations of soil performance, con- 

 tribute to the building up of an empirical understanding of soil proper- 

 ties. In fact, much soil mechanics specification for foundation design 

 is done without the aid of the more complicated soil tests and is based 



