This research investigates the in-plane seismic performance of glulam frames with buckling restrained braces (BRBs) through experimental testing, numerical modelling and design approach development.
With the advancement of engineered wood products (EWPs) and digital fabrication technology, there is an increasing interest and implementation of EWPs for mid-rise and high- rise buildings (also called mass timber buildings) around the world. However, the elastic modulus of timber is only around one-third of reinforced concrete and one-twentieth of structural steel. Additionally, limited ductility is assumed during mass timber building design due to the possibility of timber’s brittle failure in tension. Seismic considerations usually govern the design of lateral force resisting systems (LFRS) in earthquake-prone countries like New Zealand. The relatively lower elastic modulus and limited ductility of timber may cause uneconomical member sizes and increase the number of LFRS (e.g. shear walls and braces). These limitations motivated this research with the main objective to improve the seismic performance of mass timber building using a timber-steel hybrid system.
Experimental tests were conducted for BRB-braced glulam frames (BRBGFs). Following the capacity design approach, BRBs were designed as ductile elements while timber members and BRB-timber interface connections were designed as non-ductile elements. Two 8 m wide and 3.6 m high full-scale BRBGFs were built and tested under cyclic loading. Dowelled connections with inserted steel plates were used in one specimen to connect the glulam members and BRBs, while screwed connections with steel side plates were used in the other specimen. The test results showed that replacing the traditional timber braces with BRBs significantly increased the energy dissipation capacity and minimized the damage in the connections as well as glulam members. The BRBGF with the dowelled connections (S-D) had more initial slips than the BRBGF with the screwed connections (S-S), but both specimens had comparable performance after the serviceability limit state (SLS) load level.
Component-based numerical models were developed in OpenSees to investigate BRBGFs with general configurations. The test data of S-D and S-S were first used to calibrate the numerical models. Then, parametric studies were conducted to investigate the influence of connection stiffness and initial slips on the cyclic performance of BRBGFs. It was shown that the component-based numerical models represented the force-drift responses, accumulated energy dissipation and BRB deformations of S-D and S-S well. When the connection relative overstrength factor os,con was over the BRB overstrength factor os,BRB, the connections were sufficiently stiff to engage BRBs. The strength and stiffness of BRBs and initial slips caused by manufacturer tolerances had a negligible effect on the ultimate strength and energy dissipation under cyclic loading.
A direct displacement-based design (DDBD) approach was developed for the BRBGF system to avoid the complicated process of numerical modelling and facilitate the application of the hybrid system. The critical parameters for extending the DDBD approach to the BRBGF system were first discussed including the displacement profile, yield drift, connection stiffness, hysteresis damping ratio and displacement reduction factor in. Then, the component-based numerical modelling method was used to build one-bay one-storey BRBGFs and verify the critical parameters by pushover analyses and nonlinear time-history analyses (NLTHA). Moreover, the DDBD approach was used to design a set of BRBGF buildings with three, six, and nine storeys. The multi-storey BRBGF models were built in the OpenSees and analysed under a set of ground motions to verify the DDBD approach. The pushover analyses showed that the stiffness of BRB-timber connections needed to be considered when estimating the yield drift of BRBGFs. The NLTHA of one-bay one-storey BRBGFs showed that the relationship between in and ductility factor µ for the Takeda fat model was also suitable for BRBGFs on the conservative side. The NLTHA results of the multi-storey BRBGF models confirmed that the DDBD approach effectively controlled the inter-storey drift ratios of the BRBGF system under seismic loads.