150 philbrick. FOUNDATION PROBLEMS [Ch. 8 



the physical properties of which may be indicated only within ap- 

 proximate limits by the physical properties of representative samples 

 and which therefore must be considered as generally indeterminate 

 because of the effect of geologic flaws. Regardless of the approach 

 used, the investigation of the foundation should be continued from 

 the planning stage through the construction stage when the founda- 

 tion rock is exposed and the design assumptions can be re-examined. 



Experience and general practice include all the past knowledge of 

 foundations derived from the construction of similar structures on 

 similar foundations. And who is to say that one foundation is the 

 same as another? The only ones who can speak with authority are 

 those who had intimate knowledge of the foundation upon which the 

 past structure was built and who are as familiar with the proposed 

 foundation. And therein lies the weakness of foundation design. One 

 never knows as much about the foundation while it is being consid- 

 ered during design as one does about the same foundation during 

 construction. However, within limits and depending upon the type 

 and magnitude of the structure, past comparable experience and the 

 generally considered "good practice" of the engineering profession 

 are not to be discarded as an inapplicable approach, because they 

 represent time-tested data. The reliability of this method is a func- 

 tion of the similarity of the foundations and the structures built 

 thereon. It can be used safely only when all the data obtainable on 

 the proposed foundation indicate values equal to, or higher than, 

 those of the compared foundation. This is the common method used 

 in the design of low dams on insoluble rocks, and it seems to be the 

 method employed in fixing allowable loads on certain rocks in city 

 building codes. It usually results in setting ultraconservative values 

 with apparent ultimate safety factors as high as 20. 



The second method is based on the assumption of an idealized 

 homogeneous mass, the properties of which can be determined from 

 extrapolation of the properties of representative samples. This is a 

 typical mathematical approach to a problem, and it is beloved by the 

 stress analyst for it yields a clean answer. But there isn't a clean 

 answer. In the first place, sedimentary rocks are not homogeneous 

 and isotropic. In the second place, who is to say that the samples are 

 representative of the mass? Certainly the geologist will not say so; 

 for he has studied the surface geology, observed the action of the core 

 drill, logged the cores, noted the behavior of the injected water and 

 grout during the pressure testing and test grouting, and descended the 

 exploration shafts and pits to investigate the character and structure 

 of the rock in place. He has seen the heterogeneity of the bedrock 



