Experimental and analytical studies on mechanical performance of innovative energy-dissipating hold-down for CLT structures

Chen, Jiawei; Xiong, Haibei; Furuta, Tomoki; Lu, Yifeng; Abbas, Nadeem

doi:10.1016/j.conbuildmat.2021.125966

Cited by 12 publications

(3 citation statements)

References 38 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Each panel was fabricated with 35/35/35 mm layups and a total thickness of 105 mm. According to the formal research, the connection used in this study employed energyconsuming rubber to reduce the interlayer displacement and improve the anti-seismic property, as shown in figure 2 [7]. The loading procedure in Hassanieh and Zhou's study included CLT joint failure with steel panel yielding, self-tapping screw cutting, and CLT panel broking [8,9].…”

Section: Materials and Propertiesmentioning

confidence: 99%

See 1 more Smart Citation

RNN-Based Monotonic Loading Behavior Prediction of CLT Joints

Ding

Xiong

Yuan

et al. 2023

World Conference on Timber Engineering (WCTE 2023)

View full text Add to dashboard Cite

Cross-laminated timber (CLT) joints play an essential role in ensuring co-working performance between CLT structural parts and dissipating seismic energy when structures suffer earthquake. The variation of mechanical properties of CLT joints would further affect the whole structural safety. Thus, evaluating and predicting the loading behaviors of CLT joints is critical. In this research, the authors proposed a recurrent neural network (RNN)-based method to predict monotonic loading behaviors of CLT joints. Two RNNs models were adopted, including long short-term memory (LSTM) and gated recurrent unit (GRU). Further, the prediction performance of two RNNs models was compared using statistical and mechanical property index. Prediction results demonstrate that both LSTM and GRU models can predict the failure stage of the CLT joint with R 2 more than 0.90, S and T larger than 0.98. In the statistical index, the GRU model outperformed the LSTM model, although both models achieved the same accuracy in the mechanical property index. Overall, the RNN-based prediction method proposed in this study can provide an outstanding prediction performance for the monotonic mechanical behavior of CLT joints.

show abstract

Section: Materials and Propertiesmentioning

confidence: 99%

“…Based on the formal research [7], this study takes the coefficient of peak steel panel load S as the normalized value of the predicted peak steel panel load based on the real value: = 𝐹 , /𝐹 , (16) where FS,pre means the predicted peak steel panel load, the FS,real means the real peak steel panel load.…”

Section: Mechanical Property Indexmentioning

confidence: 99%

RNN-Based Monotonic Loading Behavior Prediction of CLT Joints

Ding

Xiong

Yuan

et al. 2023

World Conference on Timber Engineering (WCTE 2023)

View full text Add to dashboard Cite

show abstract

“…On the other hand, the timber substructures can be connected to the concrete slabs through two approaches. A fixed boundary condition can be realized by adopting rigid connections (e.g., angle brackets or bolts [ 19 ] ), while energy dissipation layers can be adopted by introducing energy dissipation devices (e.g., rubber bearings or dampers [ 20 ] ). Previous studies on the FaB system mainly focused on feasibility analysis, [ 18,21 ] seismic behaviors, [ 21,22 ] simplified models, [ 23 ] and floor spectrum.…”

Section: Introductionmentioning

confidence: 99%

Seismic reliability evaluation of a tall concrete‐timber hybrid structural system

Chen

Xiong

Ventura

2022

Structural Design Tall Build

Self Cite

View full text Add to dashboard Cite

Seismic reliability analysis is important for performance evaluation and structure design of novel structural systems since it can consider inherent uncertainties in seismic actions and structural properties. In this paper, seismic reliability analyses were performed on an innovative high-rise concrete-timber (FaB) hybrid structural system.The FaB system combines a concrete frame-tube structure with large story height as the main structure and multi-story timber structures on each concrete slab as substructures. To better assess the seismic reliability of the FaB system, response surface methods with either the Lasso regression approach or the machine-learning-based symbolic regression approach were conducted and compared. Uncertainties, including ground motions, elastic properties and ductility factor of the concrete main structure, mechanical properties of timber substructures and connections (i.e., rubber bearings), mass ratios, and fitting errors, were considered in formulating the limit-state performance functions. Tens of thousands of non-linear time-history analyses were run to establish the response database of the FaB system, and the reliability indexes of both the concrete main structure and timber substructures were calculated. The results show that the symbolic regression approach can formulate simpler response surface functions with similar results compared to the Lasso regression approach. The maximum difference of reliability indexes between the two approaches is less than 8.3%. For the concrete main structure, the seismic reliability indexes are in a range of 1.068 to 4.223 under different performance objectives and hazard levels, while those of timber substructures are in a range of 0.892 to 2.28. The corresponding non-exceedance probabilities of failure are between 81% and 99%, which reveals the safety and robustness of the FaB system. Meanwhile, the adoption of rubber bearings in the FaB system can enhance the seismic reliability indexes of timber substructures by 9% to 28% but has little influence on those of the concrete main structure. Outcomes of this study can serve as efficient tools to quantify the seismic performance and evaluate the performance-based seismic design methodology for the proposed tall concrete-timber hybrid structural system.

show abstract