Contents lists available at ScienceDirect Soil Dynamics and Earthquake Engineering journal homepage: www.elsevier.com/locate/soildyn Impact of displacement demand reliability for seismic vulnerability assessment at an urban scale L. Diana a, ⁎ , A. Manno b , P. Lestuzzi a , S. Podestà c , C. Luchini c a EPFL, ENAC, IIC, IMAC, Bâtiment GC, Station 18, CH-1015 Lausanne, Switzerland b DEIB, Politecnico di Milano, Via Ponzio, 34/5, 20133 Milano, Italy c DICCA, University of Genova, Via Montallegro, 1, 16145 Genova, Italy ARTICLE INFO Keywords: N2 method Seismic vulnerability assessment Lin & Miranda proposal N2 method optimization Non-linear time-history analysis Damage distribution Displacement demand ABSTRACT This paper addresses seismic vulnerability assessment at an urban scale by focusing on the displacement demand determination for building damage prediction. The study is based on the comparison of urban seismic damage distributions obtained by the displacement demand computed using non-linear time-history analysis (NLTHA) with three simplified methods. These methods include the N2 method, the Lin & Miranda proposal and an optimized version of the N2 method. Comparing the different damage distributions from the three simplified methods with the one obtained by time- history analysis helps understanding the reliability of displacement demand determination. The study is carried out on Sion and Martigny, two typical Swiss cities. For the case of Sion, results clearly show that using N2 method may lead to significant overestimation of damage grade distribution. The use of Lin & Miranda method and optimized version of N2 improves the damage prediction in both cases. For the other studied case of Martigny, N2 method and Lin & Miranda proposal are not accurate. The optimized version of N2 method provides stable and reliable results. 1. Introduction Several methods for large-scale seismic risk assessment have emerged within the past few decades, especially after destructive earthquakes occurred in Europe. In regions affected by earthquakes, direct evidence of seismic effects on structures is useful for building typology classification and damage scale definition. Methods have been developed in Italy [10,15,4,41], Turkey [11], US [14,20], Japan [39], Portugal [37], Switzerland [24], Canada [38,44], Spain [40] and France [18]. Main goals of these works are to provide methods for reliable urban risk analysis and to generate possible earthquake scenarios useful in seismic risk and damage management as well as in urban planning. The final aim of such investigations is to plan programs for seismic risk mitigation and management of emergency in the case of the occurrence of an earthquake. Seismic vulnerability assessment is also central for urban planning even for cities subjected to moderate earthquake ha- zard. Therefore, a well-formed damage scenario helps to determine focus areas for urban development and areas to regenerate. In the last few years, research focused on procedures to accelerate urban seismic vulnerability assessment. Introduction of data-mining to reduce the costly process of drawing up an inventory of building characteristics on the field (e.g. [19]) and the application for rapid risk evaluation of pre-populated databases of seismic, building inventory and vulnerability parameters for nearly real-time analysis (e.g. [36]) have been proposed. Although for the analysis of a single building damage is deliberately overestimated, such an overestimation is not desirable at an urban scale. An unreliable urban seismic vulnerability assessment leads to incorrect building damage predictions with several problems in risk management. Most approaches available for seismic vulnerability assessment of buildings are based on empirical methods and mechanical methods. Empirical methods combine observed statistical post-earthquake da- mage with a predefined value of macroseismic intensity. By contrast, mechanical methods apply parameters that define the structural re- sponse to results of the refined hazard analysis. Mechanical methods determine the damage a structure suffers for a given earthquake. Within the framework of mechanical based vulnerability approaches, some methods, such as displacement-based methods, describe the response of structures with capacity curves of structural behaviour in non-linear domain. Each point on the capacity curve is associated with a given https://doi.org/10.1016/j.soildyn.2018.05.002 Received 10 October 2017; Received in revised form 1 May 2018; Accepted 1 May 2018 ⁎ Corresponding author. E-mail address: lorenzo.diana@epfl.ch (L. Diana). Soil Dynamics and Earthquake Engineering 112 (2018) 35–52 0267-7261/ © 2018 Elsevier Ltd. All rights reserved. T