Inelastic deformation capacity of links is a factor that significantly influences design of steel eccentrically braced frames (EBFs). The link rotation angle is used to describe inelastic link deformation. The link rotation angle is generally calculated by making use of design story drifts that in turn are calculated by modifying the elastic displacements by a displacement amplification factor. This paper presents a numerical study undertaken to evaluate the displacement amplification factor given in ASCE7-10 for EBFs and the rigid-plastic mechanism used for calculating link rotation angles. A total of 72 EBFs were designed by considering the number of stories, the bay width, the link length to bay width ratio, and the seismic hazard level as the prime variables. All structures were analyzed using elastic and inelastic time history analyses. The results indicated that the displacement amplification factor given in ASCE7-10 provides unconservative estimates of the story drifts. On the other hand, the rigid-plastic mechanism provides conservative estimates of link rotations. Based on the results of the numerical study, a new set of displacement amplification factors that vary along the height of the structure and modifications to the rigid-plastic mechanism were developed. In light of the proposed modifications, the EBFs were redesigned and analyzed using inelastic time history analysis. The results indicated that the proposed modifications provide improvements for the displacement amplification factor and link rotation angle calculation procedures.respectively.The rainflow counting procedure that is explained by Richards and Uang [7] was applied to these response histories. Figure 4(b) and (d) presents ordered cycles obtained according to the rainflow counting procedure for LA22 and LA37, respectively. The figures on ordered cycles reveal that the LA22 ground motion produces cycles with lower demand but the number of damaging cycles is more. On the other hand, LA37 ground motion produces cycles with very high link rotation demands, but the number of these cycles is less. The sum of cycle ranges, which is a measure of the cumulative rotation demand, can be calculated from the ordered cycles by taking into account the cycles with ranges greater than 0.0075 radians as recommended by Richards and Uang [7]. The cumulative link rotation angle (i.e., the sum of cycle ranges) is 1.52 radians for both time histories given in Figure 4(a) and (c). The probability of link fracture can be calculated by comparing the demands with the capacities. For this purpose, recent tests conducted by Okazaki et al.[9] on links with A992 steel were considered. The cumulative link rotation angle obtained for 24 specimens was calculated. The experimental results reveal that the average of the cumulative link rotation is 1.55 radians and the standard deviation is 0.39 radians. The cumulative link rotation angle demands have an average and standard deviation of 1.33 radians and 0.78 radians under MCE ground Figure 3. Response of a nine-story EB...
