Damage detection of a typical historic masonry tower using control charts

Control charts have become a common method for damage detection within vibration-based structural health monitoring. Successful applications have been implemented including for historical buildings preservation. Nonetheless, their performance is affected by several algorithm parameters that should be properly set and more research is needed to optimise such tools. To this end, the present work discusses the preliminary results of a project that focuses on the development of a cost-effective damage detection strategy based on control charts for masonry towers in seismic prone areas. Here, a prototype of historic tower is defined based on an extensive documentary search, a methodology to simulate realistic long-term monitoring data is developed and, finally, different parameters that may affect the damage detection performance are analysed and compared against small-extent damage scenarios. The importance of normalising the features to account for seasonal fluctuations and setting a correct threshold for detection clearly emerged and drives the future scope of the work.