Automated structural dynamic modelling using model-free health monitoring results

Abstract

Structural health monitoring (SHM) methods provide damage metrics and localisation, but not a means of answering subsequent questions concerning immediate or long-term damage mitigation, risk, or safety in re-occupancy. Models based on the SHM results would provide a means to test these issues, but typically require extensive human input, which is not available immediately after an event to enhance and optimise immediate decision-making. This work presents a simple, readily automated modelling approach to translate SHM results from the proven hysteresis loop analysis (HLA) method into foundation models for immediate use. Experimental data from a 3-storey structure tested at the E-Defense facility in Japan are used to assess model performance. The model’s ability to capture the essential dynamics is assessed by comparing peak dynamic displacement and cross correlation coefficient (Rcoeff). For all 6 events, 3 storeys, and 2 directions, median (5-95% Range) of peak displacement error was 0.82 (0.17, 4.09) mm, and average Rcoeff = 0.82, all of which were significantly improved if the worst event was excluded. Overall, accurate nonlinear, time-varying baseline models were created using data from SHM damage identification and localisation methods using relatively quite simple model structures. The method is readily automated via algorithm, and the models were suitable for initial investigation and analysis on safety, damage mitigation, and thus re-occupancy. Such models could take SHM from being a tool for damage identification and extend it into further decision-making, creating far greater utility for engineers and owners, which could further spur impetus for investment in monitoring.

Experimental verification for a mechanics‐based digital cloning method predicting structural performance and damage

“…Thus, despite the differences, the predicted and measured IDR values deliver similar post‐earthquake evaluations for damage level in all storeys of the test structure under Events 2−3. Hence, the digital twin model can potentially be utilised to assess structural collapse and its financial risks by the well‐proven Incremental Dynamic Analysis (IDA) methods using IDR values as one of the main evaluation measures 42,48,49 …”

“…Hence, the digital twin model can potentially be utilised to assess structural collapse and its financial risks by the wellproven Incremental Dynamic Analysis (IDA) methods using IDR values as one of the main evaluation measures. 42,48,49 Tables 3 and 4 compares the peak values of predicted and measured inter-storey restoring force, in both negative and positive directions, and their differences and error percentages for Events 2−3, respectively. The results show average errors of 9% and 8% for Events 2−3.…”

“…Therefore, HLA can capture and tracks changes in physical parameters of the monitored structure from these linear segments and breakpoint based on force-deformation hysteresis loop shapes. [35][36][37] HLA's SHM performance has been validated for several simulated 2,37-39 and shake-table tested 38,[40][41][42] structures covering a wide range of structural nonlinear behaviour and damage mechanisms. Recently published papers 36,43 verified the accurate SHM performance of HLA against a realworld instrumented structure accumulatively damaged due to multiple seismic events.…”

Structural performance and damage prediction using a novel digital cloning technique

“…59 Thus, this digital twin can potentially be employed to assess structural collapse and associated financial risks by IDA using IDR values as one of the main evaluation measures. 30,37,38 Tables 3 and 4 compare the peak values of measured and predicted inter-story restoring forces, both in positive and negative directions, and their differences and errors for Events 2 and 3, respectively. While the results show average errors of 6% and 4% for Events 2 and 3, the maximum absolute difference is 0:15310 7 N, which is only ;2% of the total weight of the BNZ building's Unit 3.…”

“…To address these deficiencies, the model-free, mechanics-based hysteresis loop analysis (HLA)-SHM method was developed and validated for numerical 3,[25][26][27] and experimental 25,[28][29][30] structures covering a wide range of nonlinear dynamic behaviors and damage mechanisms. HLA is a rapid, real-time SHM method with an explicit design needing no prior training process, while returning coherent SHM results immediately after earthquakes.…”