Selecting the Probability Distribution of Cone Tip Resistance Using Moment Ratio Diagram for Soil in Nasiriyah

• Published in: Geotechnical and geological engineering 2019-06, Vol.37 (3), p.1703-1728
• Main Author: Shakir, Ressol R.
• Format: Article
• Language: English
• Subjects: Civil Engineering; Clay; Clay soils; Computer simulation; Cone penetration tests; Data points; Density stratification; Distribution; Earth and Environmental Science; Earth Sciences; Gasoline; Geotechnical Engineering & Applied Earth Sciences; Hydrogeology; Original Paper; Penetration resistance; Probability distribution; Probability theory; Rayleigh distribution; Refineries; Reliability analysis; Sand; Service stations; Soil; Soil properties; Statistical analysis; Statistical methods; Stratification; Terrestrial Pollution; Thickness; Waste Management/Waste Technology; Weibull distribution
• ISSN: 0960-3182; 1573-1529
• DOI: 10.1007/s10706-018-0716-3

Selecting suitable probability distributions (PDs) to describe cone tip resistance measurements (q c ) obtained by a cone penetration test (CPT) is considered a crucial requirement to get a good solution for geotechnical problems solved by simulating the engineering properties of soil as a random field or for use in reliability-based design. This paper presents a statistical analysis of seven PDs proposed to model q c obtained through performing CPT for soil in Nasiriyah during the construction of a new refinery petrol station. Preliminary testing for suitability of the suggested distributions has used the method of moment ratio diagram (MRD) based on the Pearson system. It was found that the soil stratification has a large effect on the distance between every two points on MRD. The type of probability distribution was also affected, and changed, by increasing the number of data points for q c included in the analysis. Logistic and Weibull distributions are considered the best PDs that represent the q c of the first layer having thickness 12 m of clay soil, followed by the other distributions, while the logistic and normal distributions were considered the best PDs among the seven suggested distributions for the second layer of 8 m silty sand and clayey sand. All the suggested distribution can represent the given q c data approximately except the Rayleigh distribution.