Submission for buildingSMART International «openBIM Award 2023»

• Category: Project Delivery Excellence / Construction for Infrastructure

Summary 

The conventional monitoring method for large-span special-shaped space grids cannot meet the requirements of rapid and comprehensive structural Deformation monitoring monitoring on site. In order to achieve high-precision, high-efficiency and whole process Deformation monitoring of large-span special-shaped space grid, three-dimensional laser scanning is used to obtain the point cloud data of the whole process of structure construction. After processing the point cloud data, it is aligned with the BIM model to complete the single component deviation analysis, construction process and morphological structure deviation analysis, and achieve Deformation monitoring of the whole process of construction. The comparison between the monitoring results and the data measured by total station shows that the data values and standard deviations measured by 3D scanning Deformation monitoring and total station are basically consistent, the data fluctuation is small, and the monitoring results are stable, which is suitable for Deformation monitoring throughout the construction process of large space steel structure projects.

In the factory processing stage, industrial 3D scanner is used to analyze and correct the deviation of large complex special-shaped steel members. The 3D Test model of key parts is generated through the reverse image of the industrial 3D scanner, and the complex components are located by combining with the high-precision photo positioning system, which greatly improves the measurement efficiency and accuracy. After the scanning is completed, a three-dimensional color difference deviation map is formed through fitting analysis with the Reference model in the computer software. For over offset parts, in addition to labeling the deviation value, any surface can also be cut, and the three-dimensional color difference deviation can be displayed on the two-dimensional view, which can intuitively see the size and direction of the deviation and quickly guide the factory to correct over offset components.

According to the on-site installation sequence, the digital models obtained by scanning the actual components are sequentially imported into the computer. The physical pre assembly process of the components is simulated in a computer virtual environment, and problems such as misalignment and bracket deviation that may occur during the on-site installation process of the actual processed components are analyzed. It is determined whether the relevant requirements of the "Steel Structure Construction Quality Acceptance Standard" are met. For those that do not meet the standard requirements, Provide practical and feasible deviation component adjustment plans during the factory processing stage. Compared with traditional physical pre assembly, it significantly improves construction efficiency, reduces resource investment, and does not require a large number of assembly sites.

The application of 3D laser scanning technology in quality management of steel structure construction stage mainly involves scanning the key monitoring parts and key construction stages of the structure through scanning instruments. It can also simulate the construction of key construction stages, providing good data for quality monitoring and acceptance work, and achieving real-time sharing of detection information. By scanning the structure with a scanning instrument, the obtained point cloud data is processed through data stitching, denoising, removing outliers, feature extraction, simplification, etc., and compared with the BIM model obtained from design drawing modeling to obtain monitoring data during the construction phase.

Before the scanning work during the on-site construction phase begins, a site survey is required to determine the number of scanned sites, target ball positions, and black and white target layout based on the project site layout, structural characteristics, etc. The placement of target balls should ensure that there are at least three shared target balls between adjacent sites for point cloud registration at different sites. Target paper should be placed at specific positions in the structure to control subsequent reference planes.

When conducting on-site data collection on the tested structure using a 3D laser scanner, there should be no obstruction between the scanning site and the tested object, and the distance between the two should be within the scanning radius to ensure that the structure can be scanned to the maximum extent possible. After the point cloud scanning is completed, each station conducts target scanning, which is divided into two parts: the first part is the target ball used for splicing data between stations in the later stage of scanning, ensuring that there are two common target balls between stations in principle; The second part is to scan the black and white targets arranged on the site, and obtain the coordinates of the black and white targets in the geodetic coordinate system through the total station for the positioning of point cloud data in the later stage. At the same time, the accuracy of 3D scanning monitoring data is verified by comparing the distance between target points and total station data.

After the completion of scanning during the assembly phase, all point cloud data is merged and the quality and accuracy of each scan are checked. Other excess scanning data is removed. If point cloud data needs to be located, coordinates need to be assigned to the model and the accuracy between coordinates needs to be checked to generate the final required point cloud model.

The roof shape of this project adopts parameterized design to optimize the surface shape. During the construction process, deviations in the steel structure will have an impact on the partition of the waiting floor aluminum plate ceiling. 3D laser scanning was used to collect point cloud data of the steel structure roof, and the point cloud data was imported into Rhino in. stp format for registration with the deepening design model. By using Grasshopper to extract the nurbs modeling parameters and point cloud data of the deepening design model, the nurbs modeling parameters and point cloud data are imported into Matlab using XML. The parameter is used to optimize the reverse reconstruction model and improve the accuracy of point cloud data feature extraction, achieving rapid extraction of deviations between entities and models. After obtaining the deviation of the steel structure, a generative design was used to deepen the design. The component boundary values obtained from Matlab analysis were read into Rhino in XML using Grasshopper. Without affecting the structural stress, the division of the aluminum unit plate of the nurbs surface layer and the size of the adapter were rationalized and optimized to achieve rapid extraction of machining coordinate data. Model reconstruction using the obtained data coordinates through Dynamo to obtain a Revit model, which is then coordinated and exported through lFC for manufacturing. OpenBlM achieves the entire process by providing a software neutral information exchange platform, with data integrity sufficient for manufacturing.

The use of 3D laser scanning technology for quality control of the entire construction process of large-span irregular space grids has significant advantages over traditional technologies in terms of accuracy, speed, and overall monitoring. The Deformation monitoring of the whole process of steel structure construction was carried out at Hangzhou West Railway Station, covering point cloud data and monitoring results of processing, assembly, lifting and structural formation. Compared with the measurement results of total station, it is proved that the accuracy and stability of the data obtained by 3D laser scanning meet the requirements of on-site monitoring.