Gaussian-filtered Horizontal Motion (GHM) plots of non-synchronous ambient microtremors for the identification of flexural and torsional modes of a building

It is often assumed that, in order to identify flexural and torsional vibration modes of a building, it is necessary to record synchronous data from a series of sensors deployed at different points. In the present paper, we present a simple and straightforward methodology to unambiguously identify flexural and torsional modes through the analysis of non-synchronous data collected by a single sensor placed in succession at different points of the structure. This is accomplished by recording few minutes of ambient microtremor data by means of a 3-component geophone placed at different points of the same floor. Amplitude spectra are computed for determining the vibration frequencies. Successively, in order to identify the type of motion, we apply a series of narrow Gaussian filters centered at the previously-identified frequencies. By plotting the horizontal motion for each considered point, we are then able to simply and unambiguously determine whether the motion of a given frequency refers to a flexural or torsional mode. If, for a given frequency, the motion at two (or more) points has the same direction and similar amplitude, that frequency represents a flexural mode, while in case the directions and the amplitude are different, elementary considerations indicate that this is predominantly torsional. The methodology is first introduced by considering a case study where synchronous microtremor data are also recorded. In a second case study, the method is applied to non-synchronous microtremor data collected at a 25-storey building and results are compared with the numerical simulations performed by means of the Finite Element Method (FEM). •Determination of flexural and torsional modes through non-synchronous data.•Ambient vibrations acquired by a single sensor at several points.•Flexural and torsional modes identified by means of simple Gaussian filters.•Comparison of measured Eigenmodes and modal analysis via FEM.