Many soil structures, including retaining walls and road embankments exhibit signs of distress, such as vertical and lateral movement, rotation, or rupture, during their service life. The determination of explanations for the observed defects is often straightforward where classical rotational or sliding failure mechanisms are evident. However, in situations where several potential explanations are identified, or where previous remedial work has been undertaken, the determination of the most likely explanation becomes tedious, costly, or both. A transparent diagnostic procedure is necessary to determine the cause of defects expeditiously when there is potential for loss of life, or where it is necessary to prescribe corrective action to prevent the defect recurring. More importantly, the prediction of the future performance of the soil structure requires a clear understanding of the performance of the earth structure before and after any corrective work is undertaken. In these circumstances it is important to correctly identify explanations for the observed defects and to incorporate this knowledge in the assessment of the earth structure. This paper describes an iterative procedure for the diagnosis of in-service defective earth retaining walls. The term diagnosis is used in the paper to refer to the process of determining a set of technical explanations for the occurrence of the observed defects. Using probability theory, the geotechnical engineer is in a position to make informed decisions on; the level of confidence associated with a given hypothesis being correct, whether further investigations are necessary, or whether a decision can be made to terminate the diagnosis and for corrective action to be undertaken.
Iterative, hypothesis-and-test, procedures are widely used in the design or diagnosis of engineering systems. Diagnosis is initiated by a set of observed defects, which are the unusual or undesired aspects of performance.
