Uncertainties in Evaluation of Strength of Marine Sand

The strength of a marine sand is investigated by means of cone penetration tests and plate load tests, and by measurement of the resistance against penetration by the steel skirts underneath a gravity platform. Several methods are used to interpret the test results and to evaluate the angle of internal friction, φ\N’. Probability methods are used to analyze the uncertainty about φ\N’ due to spatial variation in the data and model errors in the equations used for interpretation. It is found that different methods of interpretation give different values of the internal friction. The correct value of φ\N’ cannot be determined with certainty because the errors in the interpretive models are not well known. Hence, subjective opinion is introduced as a means of exercising the engineer’s judgment. Bayesian updating is used to combine the results obtained by the different tests. The methods illustrate how the various uncertainties may be combined to arrive at the most likely value of the angle of internal friction and the uncertainties about that value.