Planar frame models considering system‐level interactions of a two‐story low‐damage concrete wall building

• Published in: Earthquake engineering & structural dynamics 2023-11, Vol.52 (14), p.4325-4351
• Main Authors: Gu, Anqi, Zhou, Ying, Lu, Yiqiu, Yang, Qun, Henry, Richard S., Rodgers, Geoffrey W.
• Format: Article
• Language: English
• Subjects: Building damage; Compatibility; Concrete; Damage; Deformation; Direction; Floors; Modelling; Strength; Uplift
• ISSN: 0098-8847; 1096-9845
• DOI: 10.1002/eqe.3954

A system‐level shake‐table test of a two‐story low‐damage concrete wall building was conducted in 2019. The test building consisted of unbonded post‐tensioned (UPT) walls with perimeter frames and different wall‐to‐floor connections to provide deformation compatibility. The test results highlighted that the measured lateral strength of the building notably exceeded the design value. This over‐strength was largely attributed to the compatibility deformations imposed on the floor systems. Previous simulation results from planar models also underestimated the building global response and lateral strength, so a modified model was developed to improve the accuracy by accounting for the system‐level interactions. The floor slab out‐of‐plane strength was assessed by the floor slab models using shell elements to quantify their response when subjected to deformations induced by the UPT wall uplift. In addition, the unintended effects of the beam‐to‐floor connection strength were also calculated. The strength contributions from the floor and connections were included in the planar frame models using lumped rotational springs located at the slotted‐beam joints. Lastly, the models were also updated to include flexibility in the connections of the energy dissipations at the wall bases based on test observations. Nonlinear time history analysis of the updated models for unidirectional loading cases demonstrated an improved calculation of the test results for both the global and local responses. When considering the floor interaction in the updated models, the absolute relative errors of peak overturning moments reduced from 35.4% to 15.6% averagely for the longitudinal direction, and from 23.1% to 14.1% averagely for the transverse direction.