Executive Summary : | In an earthquake, two horizontal components may often induce significant damage to structural systems (Roy et al., 2022). Owing to the arbitrary variations of the amplitude and direction of the horizontal components, a real motion is generally recorded by a pair of sensors preinstalled in orthogonal orientations. Further, orientations of structures relative to those of the pair of a motion may grossly differ. As such, the incidence angle of motions may influence structural response (Skoulidou and Romao, 2020). A proper seismic response assessment scheme, therefore, engages conducting nonlinear analysis of a structure under a pair of components for all incidence angles which is onerous for routine design (Roy et al., 2018). Recommendations for combining responses from separate analysis due to each component (unidirectional analysis) by approximate combination rules (ASCE/SEI 7-16, 2016) appear to endorse the views. Keeping the challenges of bidirectional analysis and influence of incidence angles in view, Roy et al. (2017, 2018) explicitly proposed to (i) rotate horizontal components of ground motions to the most preferred orientations, which can be defined in terms of ‘pure’ ground motion parameters; (ii) conduct unidirectional analysis and (iii) combine response quantities so obtained by suitable combination rules to estimate both demand and damage parameters. From a comprehensive scrutiny on biaxial interaction, incidence angles, and codal guidelines, such proposals are recommended in a very recent work (Roy et al., 2022). The identification of the preferred orientations, however, requires the description of a pair of components in advance, which is design contexts, is often unknown. In practice, a number of earthquake motions are often selected for a seismic scenario compatible with the design spectra. ‘In the design office, it is usually not practical or feasible....’ (Grant, 2011) to analyze structures for a large number of motions so selected. As such, ‘spectrum-matched records to estimate a mean/median seismic response with reduced dispersion’ (Hancock et al., 2006; Al-Atik and Abrahamson, 2010; Baker and Lee, 2018) are adopted. However, ‘spectrum matching simultaneously for two-components is still contentious’ (e.g., Grant, 2011; Reyes et al., 2014). Also, limited studies on spectral matching of two-components do not address the issue of response assessment to bidirectional excitations taking the issue of incidence angle into account. Against this backdrop, the major goal of the proposed work is set as follows: (i) Developing pair of spectra for spectral matching of two horizontal components which can be utilized to estimate actual mean/median response to bidirectional shaking through unidirectional shaking reduced dispersion can be achieved and (ii) presenting response statistics in an orderly manner shall aid to estimate response a-priori. Attainments of the goal so set shall surely simplify and improve upon seismic design practice. |