Systematic selection of field response measurements for excavation back analysis

In a geotechnical excavation, back analyses are routinely performed using the measured field responses to derive the material parameter values for the different soil layers present at the site. For the purpose of back analyses, the engineers will usually make use of a portion of the large volumes of field data collected, in order to keep the computational effort to a manageable level. However, excavation back analyses using different selected sets of field response measurements may not yield the same knowledge of material parameter values. Therefore, measurements need to be carefully selected to obtain the best estimates of material parameter values. Currently, the selection of measurements is largely based on engineering heuristics; no method has been proposed to systematically quantify the expected knowledge of the parameter values that field response measurements could provide. In this paper, a hierarchical algorithm based on a joint-entropy objective function is proposed to systematically evaluate the knowledge gained from wall deflections measured by eight inclinometers at an excavation site. The algorithm ranks the inclinometers based on the expected knowledge yield of the parameter values. Back analysis using actual field response measurements is then carried out to corroborate the ranking. The rankings obtained from the back analysis results and the predictions of the hierarchical algorithm are very similar, which suggests that the latter method can aid in the judicious selection of field response measurements in order to obtain useful knowledge of material parameter values. Since the application of the hierarchical algorithm does not entail the use of actual measurements, such predictions can be made at the early stages of a project, even before the commencement of site activities.

Automated summary

The paper proposes a hierarchical algorithm based on a joint-entropy objective function to systematically evaluate the knowledge gained from wall deflections measured by inclinometers at an excavation site. By corroborating the algorithm's predictions with back analysis results, it is shown that the method can aid in the judicious selection of field response measurements to obtain useful knowledge of material parameter values. The hierarchical algorithm allows for predictions to be made at the early stages of a project, even before the commencement of site activities, without the need for actual measurements.