Miguel Carvalho

Towards a Smart Campus: Building-User Learning Interaction for Energy Efficiency, the Lisbon Case Study

The Smart Campus is a European project based in four pilots located in Helsinki (Metropolia University), Lisbon (Instituto Superior Tecnico), Lulea (Technology University) and Milan (Politecnico di Milano), having engaged 76,000 users since August 2012 during 33 months. The main objective of this project was to demonstrate the central role of Information and Communication Technology (ICT) based services that act upon Energy Management Systems and control heating, ventilation and air conditioning (HVAC) and Lighting on achieving energy efficiency through dynamically negotiating building environmental conditions with the users on the University pilots. This paper presents the energy saving results and the best practices obtained in the Lisbon Pilot where ICT equipment was installed in a Library, a lecture Amphitheater and a set of offices. Overall savings were almost 40 % in some test locations. It was observed that energy savings were different between test locations due to particularities such as space typology, occupation and utilization patterns, equipment installed, HVAC and lighting control systems and users interaction.

Identification of the Unconstrained Modes of 3D Axisymmetric Structures from Measurements under Constraining Support Conditions

An inverse method for extracting the unconstrained modal parameters of 3D axisymmetric structures from measurements performed under constrained configurations is proposed. The work was originally motivated by the need of knowing the tuning of large historical carillon bells which now lie on reinforcement fixtures, implying a quite different response of the bells from their original suspended state. In this paper, we extend to three-dimensional bodies the identification technique recently developed by the authors for a case of study consisting on a simple discrete mass-spring circular ring. Considering a modal model of a free cylinder as a first approximation of a bell, we present the identification strategy and then illustrate the efficiency of the technique by providing identification results for the original system modal frequencies from the constrained system. By operating directly on the constrained transfer functions measured at the constraint locations, the technique is not prone to modal identification nor to truncation errors, and therefore appears particularly convenient for modal testing on real structures.