Arabian Journal for Science and Engineering https://doi.org/10.1007/s13369-018-3584-5 RESEARCH ARTICLE - CIVIL ENGINEERING Assessment of Structure-Specific Intensity Measures for the Probabilistic Seismic Demand Analysis of Steel Moment Frames Mohammad Javanmard 1 · Aliakbar Yahyaabadi 1 Received: 24 January 2018 / Accepted: 30 September 2018 © King Fahd University of Petroleum & Minerals 2018 Abstract Given the inherent uncertainty in seismic response, seismic performance assessment of structures should be conducted within a probabilistic framework. One of the most efficient probabilistic approaches is the IM-based probabilistic seismic demand analysis (PSDA). In this method, an intermediate parameter, which is known as the intensity measure (IM), is used to decouple the seismological and structural uncertainties. Two intensity measures of IM oc and ( S a ) rms were introduced for near-fault pulse-like records in previous research. These IMs are defined based on the optimal combination of spectral displacements and root-mean-square of spectral accelerations at effective periods, respectively. In this research, to consider the efficiency of these IMs under a set of 90 records that contains both near-fault and ordinary ground motion records, we conducted the PSDA for five moment-resisting frames with the number of stories ranges from 3 to 15. Results show that IM oc and the advanced intensity measure of IM 1I&2E exhibit the highest correlation with the expected damage for the most frames, especially moderate and relatively long-period ones. IM 1I&2E is defined based on the inelastic spectral displacement with the higher-mode modification. In addition, comparison of the drift hazard curve of different frames shows that by increasing the structural height, the amount of drift hazard will decrease. However, comparing to other cases, the reduction rate of drift hazard along with increasing the number of stories from three to six is significant. Keywords Probabilistic analysis · Seismic demand · Intensity measure · Moment-resisting frames · Efficiency · Drift hazard 1 Introduction Probabilistic seismic demand analysis measures the mean annual frequency of exceeding a specified seismic demand for a given structure [1]. In PSDA, the structural response and the ground motion intensity are quantified, respectively, via an engineering demand parameter (EDP) and the inten- sity measure (IM). Basically, incremental dynamic analysis (IDA) is used to forecast EDP value for future earthquakes [2]. In IDA, the intensity of each earthquake ground motion is increased after each nonlinear dynamic analysis, using IM as the seismic intensity scaling index. By combining IDA and hazard analysis results, we can estimate the probability that EDP exceeds a specific value of edp as follows: B Aliakbar Yahyaabadi A.yahyaabadi@ub.ac.ir 1 Faculty of Engineering, University of Bojnord, Bojnord 94531-55111, Iran λEDP (edp) =  G  EDP > edp|IM = im  ·|d λ IM (im)| (1) In Eq. (1), G(EDP > edp|IM = im) is the probability that EDP exceeds a given level of edp conditioned on a given intensity level of im, and the term λ IM (im) denotes the mean annual frequency of exceeding the value of im [1]. At a speci- fied intensity level, a structure will experience collapse under a fraction of earthquake records, therefore non-collapse and collapse data can be combined according to the total proba- bility theorem to obtain the probability of an EDP exceeding a specified value [1,3]: G  EDP > edp|IM = im  = G  EDP > edp|IM = im, NC  ( 1 - P C|im ) + P C|im (2) In Eq. (2), G(EDP > edp|IM = im, NC ) is the amount of G(EDP > edp|IM = im) for non-collapse data, and P C|im denotes the collapse probability at the intensity level of im. Equation (2) can then be substituted into Eq. (1) to esti- 123