Error Evaluation and Suitability Assessment of Common Reliability Methods in the Case of Shallow Foundations

A geotechnical structure's reliability index calculated using identical input parameters and assumptions can significantly vary as a function of the used method. The different approaches to solving the reliability problem could result in an error which depends on many factors. The most important error sources are the complexity of the performance function, the number of random variables, their mutual correlations, and marginal statistical distributions. A review of relevant literature in the field of reliability in geotechnical engineering revealed a lack of information on the errors of individual reliability methods for geotechnical problems and general criteria for assessing their suitability concerning the error size. The paper defines the reliability method error and proposes criteria for assessing the suitability of reliability methods in geotechnical engineering. Based on the proposed criteria, the suitability of common reliability methods was evaluated in the example of a shallow foundation, analysed according to Eurocode 7, DA 3, such that Ed=Rd. It is shown that due to the mathematically complex expression of the reliability integral, methods that are easier to use result in a larger error and are not suitable for a reliability analysis of shallow foundations. Sophisticated methods are more accurate but require specific knowledge and resources that are not often used in daily engineering practice.

Automated summary

The reliability index of a geotechnical structure can vary significantly depending on the method used, even with identical input parameters and assumptions. Different approaches to solving the reliability problem can lead to errors influenced by various factors, including complexity of the performance function, number of random variables, correlations between variables, and statistical distributions. While more sophisticated methods may be more accurate, they require specific knowledge and resources that are not commonly utilized in daily engineering practice.