A PROBABILISTIC APPROACH TO THE CORRECTION OF SOIL STRENGTH

A method is outlined which presents a means of correcting measured soil strengths for sample disturbance which involves the establishment of the frequency distribution of the disturbance and the application of this to the frequency distribution of the observed strengths. The method could be employed to correct for a particular sampling technique, for a drilling crew of for a whole testing laboratory adequate input data is available. One difficulty in the application of this method is the determination of the required input. The frequency distribution of the measured strength can be obtained directly from the laboratory test results but the degree of disturbance produced by a particular drilling crew, by a particular method of sample handling or by a particular method of sample preparation and testing is difficult to assess. Although the actual frequency distribution of the corrected strength could not be determined for the specific strength data cited, the computed values of the mean and variance of the corrected strength agreed closely with values computed from the observed data. This approach was applied to the correction of UU triaxial test results, but it is also directly applicable to unconfined compression test results. Although the actual method of correction will differ somewhat for other types of strength test, the general concepts employed in this approach are valid for any time of test data, provided that the method of correction and the required input can be specified. The paper discusses the concepts of perfect sample, sound sample and disturbance. Details are given of frequency distributions and correction of measured strengths. An example of strength correction is presented.

Media Info

  • Features: Figures; References; Tables;
  • Pagination: p. 358-370

Subject/Index Terms

Filing Info

  • Accession Number: 00263044
  • Record Type: Publication
  • Report/Paper Numbers: Proceeding
  • Files: TRIS
  • Created Date: Nov 12 1974 12:00AM