Appendix 

 INTERPRETATION AND USE OF CPT DATA 



CORE PENETRATION TEST (CPT) PRACTICE 



The CPT was introduced in Europe about 50 years ago and recently 

 has gained acceptance in many countries. Its initial application was to 

 pile design as the test resembles a model pile test. However, research 

 has extended its utility to soil classification and determination of 

 relative density, friction angle, settlement on sand, and clay impressi- 

 bility. Also, methods for designing shallow foundations from CPT data 

 have been developed. Many papers on CPT test equipment and data inter- 

 pretation can be found in the proceedings of two Eurpoean Symposiums on 

 Penetration Testing reported in ESOPT (1975) and (1982). 



Soil Properties 



Soil Classification . Efforts to classify soils with the CPT were 

 first reported in Begemann (1965). His method was based on the ratio of 

 sleeve friction to cone pressure (i.e., the friction ratio). In 

 essence, he found that the friction ratio increased as median grain 

 diameter decreased. Begemann 1 s observations have been generally con- 

 firmed by others who developed soil classification charts. Common to 

 classification charts is a dependence on the cone type used and the 

 difficulty in classifying mixed (sand/silt/clay) soils. Of interest to 

 the Navy are classification charts developed for electrical friction 

 cones as this is the type of cone used in offshore investigations. 

 Martin and Douglas (1981) published such a chart (Figure 12) which is 

 perhaps the most comprehensive classification chart available. Work has 

 also been done to extend CPT soil classification to carbonate soils 

 (Beringen et al . , 1981). In Beringen's chart, cone resistance is used 

 to estimate the degree of cementation. Other parameters (e.g., grada- 

 tion and microscopic examination) are used to further classify calcar- 

 eous soils. 



Relative Density . Relative density can be estimated from CPT cone 

 pressure data but is confounded by lateral stresses, grain size, depth 

 of overburden, and other parameters. Consequently, theoretical 

 approaches to determining density have not proved as successful as 

 empirical procedures. Caution, however, is warranted. Vi 1 let and 

 Mitchell (1981) pointed out that these empirical relationships are not 

 unique but vary according to the sand being penetrated. Schmertmann 

 (1978) presented a plot of cone pressure versus vertical effective 

 stress for different relative densities (Figure 26). These curves are 



42 



