PSI - Issue 54

Mohammad Shamim Miah et al. / Procedia Structural Integrity 54 (2024) 3–10

4 2

Author name / Structural Integrity Procedia 00 (2023) 000–000 to keep them connected with modern technologies e.g. sensors. Due to many underlying advantages over manual inspection and monitoring real-time monitoring or monitoring based on sensory information is gain ing serious attention Li et al. (2002). Interestingly, sensory information based monitoring are quite reliable as long as the data acquisition system and the sensors are in perfect operation condition. If the operation condition is interrupted for whatever reason the monitoring task become difficult. To avoid the reliability on single type of sensor e.g. accelerometers or displacometer, having multiple type of sensors might be useful as more data will be available. Therefore, multi-sensor data fusion is gaining popularity due early mentioned benefits Duan et al. (2020). It is undoubtful that the modern structural health monitoring (SHM) systems can provide diverse benefits and save lives and cost. However, in order to get benefit of modern sensors based monitoring systems few thighs need to be synchronized. For instance, the basic/per-knowledge of structures, placement of sensors, types and number of sensors, proper monitoring strategy and data analysis, last but not least, both hardware and software capability. Over the last few decades, the SHM has adopted many strategies including a wide range of sensors adoption Sun et al. (2010); Enckell (2006); Miah and Lienhart (2023). Among them few can be listed, such as, radar data-based monitoring Ahlborn et a. (2013); Biondi et a. (2020); Guo et a. (2021), acceleration data-based smart monitoring Dyke (1998); Miah et a. (2017), remote sensing for monitoring Dong et al. (2009); Miah and Lienhart (2023); Ahlborn et a. (2013); Soldovieri et al. (2021), geodetic sensors data-based Ehrhart and Linehart (2017), etc. As there is no perfect strategy due to the unpredictable and uncountable issues associate to SHM, hence, it is a need to develop and update the existing methodologies. It can be noted that regardless of sensors type and strategy the uncertainties associated with SHM is unavoidable. The type of uncertainties in SHM are uncountable and they can be induced by dynamic loads and their magnitudes, measured data, measurement systems and sensors. In this context, many researchers have reported various type of uncertainties and their treatment for the SHM Jenkel et al. (2008); Hirai and Mita (2016); Lam (1998); Gokce (2012); Liu et al. (2015); Bull et al. (2021); Caspani et al. (2022); Lorenzoni et al. (2016); Putra et al. (2013); Li (2016). One step further, the uncertainties even create more problem when those data are used to develop virtual model for monitoring and prognosis Lam (1998); Gokce (2012); Miah and Lienhart (2023); Li and Ou (2011). The development of data-based modelling is getting attention due to underlying advantages, such as, data-based model does not required any information about the physics of the structure. Additionally, the data-based are so powerful tool that can assist to monitor and track the performance of any structure without physical visit to the site. Also such model has the potential to predict the foreseen behaviour, however, for predicting the future behaviour of the existing structure the accuracy of the model is crucial. Therefore, it is essential to minimize the uncertainties of the data to have an optimal SHM strategy. From the above discussion, it can be summarized that the uncertainties in SHM are not limited to any specific issue rather it comes from various sources. As a result, the overall monitoring task gets complicated due to early mentioned problems. In addition to those uncertainties, this study has focused into the induced uncertainties of the abrupt dynamic excitations and sensors location. More precisely, there were three ac celerometers on the bridge and one accelerometer has been placed under the bridge on an external support. The accelerations date have been measured and different quantities have been evaluated to understand the associated uncertainties. The rest of the paper contains problem specification, results and discussion, finally, the end part has summarized the outcome with remarks. 2. Description of the Problem A 2 m long steel bridge has been used to perform the tests. Total 4 accelerometers are placed at different locations as depicted in Fig. 1. More specifically, 3 accelerometers have been placed on the bridge and 1 sensor is placed under the bridge on an external support (detached from the bridge). The sensors are placed intentionally to see their capability of capturing signals and as well as the quality of the signals. It is expected that sensors at the middle on the bridge would get the most of the signals correctly with proper amplitudes. The applied dynamic loads are abrupt to resemble an operational dynamical loading conditions. Also extra masses have been placed to see their contribution into the dynamical properties e.g.