Parametric dynamic interaction assessment between bells and supporting slender masonry tower

Abstract

Recent years have seen a growing interest toward implementation and testing of structural health monitoring techniques for cultural heritage structures, and many scientific papers report on the application of operational modal strategies as an effective knowledge-based tool for vulnerability reduction of masonry buildings. Focusing on historic masonry bell-towers, being such structures particularly prone to earthquake-induced damage, the most part of the studies discuss structural monitoring and vibration-based identification methods with the goal of their seismic protection. As a consequence, while there is great number of researches that investigate masonry towers behaviour under earthquake loads, only a few scientific papers discuss their structural response under service loads such as bell-loads. This issue is also of paramount importance, since in many real cases the bell-ringing has been stopped due to the dynamic interaction phenomena that are activate between the bells and the host structure. With the aim to contribute of improving the knowledge in this field, this paper focuses on a methodology for the study of the dynamic interaction between bells and slender masonry towers. The proposed methodology is divided into four phases: (i) Geometric and structural characterization of the tower and bells; (ii) Evaluation of the dynamic forces generated by the swinging bells; (iii) Experimental campaign to characterize the dynamic properties of the tower by means of operational modal analysis; (iv) Parametric finite element analysis. To illustrate the methodology, a real case of masonry bell-tower in which bell-ringing had to be stopped due to a history of strong vibrations is discussed. The paper includes a method of analysing the dynamic properties of masonry bell-towers, in which the dynamic interaction between the harmonic bell forces and the fundamental tower modes is analysed by means of a calibrated numerical model and the dynamic amplification factor.