Optimal sensor placement for the modal identification of an innovative timber structure

Abstract

This paper aims at determining an optimal grid of acceleration sensors for the accurate estimation of modal data from an innovative timber-hybrid structure. As the knowledge on the full-scale behaviour of such structures is limited thus far, an extensive identification campaign on an innovative timber building, the ETH House of Natural Resources, is currently being carried out at ETH Zürich. In conjunction with this campaign an optimal placement for modal identification sensors is undertaken, in order to extract the maximum possible information from a minimal number of sensors. The entire structure is modelled in SAP2000 and the modal information (frequencies and mode shapes) is extracted from the software. The obtained mode shapes are then utilized as the input to the sensor placement problem. Three widely accepted sensor placement methods are implemented and assessed, namely, the effective independence method (EFI), the driving point residue EFI method (EFI-DPR) and the maximum kinetic energy method (MKE). The optimization algorithms identify the most relevant degrees of freedom, which should be monitored during the testing campaign.