BIM-based precast building optimization for component sizing and rebar detailing considering constructability

Abstract

The utilization of precast concrete is rapidly gaining popularity in the construction industry due to its capacity to enhance building efficiency and quality in comparison to cast-in-situ concrete. Precast construction differs from cast-in-situ construction as it involves the manufacturing process of components performed within a factory, followed by the transportation of these components to the construction site for assembly. The use of repeating molds and consistent rebar layouts can significantly improve worker efficiency. Therefore, constructability factors, particularly those related to standardization, should be carefully considered. However, current research on precast construction primarily focuses on logistics and sequencing, neglecting the importance of standardization. Consequently, this study developed a BIM-based framework to examine the relationship between standardization and construction cost, incorporating the concept of standardization into a constructability score.

This research proposes a framework that incorporates Building Information Modeling techniques to extract semantic information from architectural plans. Subsequently, a gradient-based Optimality Criteria method is utilized to optimize the sizing variables of precast components. Additionally, a hybrid approach called NSGA-II-GD, which combines Non-dominated Sorting Genetic Algorithm II and Great Deluge Algorithm, is employed to optimize the rebar layout design of each precast component. While NSGA-II-GD improves computational speed and provides superior convergence and search space compared to other GA-based algorithms, it still requires significant operating time. To address this, a graph neural network (GNN) approach is adopted to predict the rebar layout of components based on the dataset generated from the NSGA-II-GD approach. The GNN prediction exhibits substantial time improvement while maintaining comparable results. Results from an illustrative example demonstrate the existence of an optimal point between construction cost and standardization, particularly for components subjected to similar stresses.

Date of Award	2023
Original language	English
Awarding Institution	The Hong Kong University of Science and Technology
Supervisor	Jack Chin Pang CHENG (Supervisor)