A high‐accuracy numerical model of long‐span spatial structures with as‐true geometry based on a self‐developed MASM method

Abstract

In the context of Industry 4.0, digital twin technology has emerged as a key enabler of the digital transformation in civil engineering by supporting intelligent operation and maintenance through high‐precision modeling. For long‐span spatial structures, construction deviations and complex geometries necessitate modeling based on as‐built data. However, existing methods still struggle to achieve automated, efficient, and high‐precision modeling for such complex structures. This study employs a self‐developed processing method for measurement point clouds, namely, Multi‐member Automatic Structural Modelling (MASM) method, to reconstruct the accurate geometry of scanned long‐span spatial structures. The approach employs modular segmentation of point clouds, extraction of skeleton points, and fitting of members and nodes to efficiently rebuild the accurate structural geometry. The MASM‐generated model can be imported into ANSYS for high‐accuracy numerical simulation of long‐span spatial structures. To evaluate the accuracy and applicability of the model, the model based on a self‐developed MASM method was compared with the ideal design model and a defect model based on code assumptions. The results demonstrate that the MASM‐generated model exhibits higher accuracy and better reflects the actual structure, effectively capturing real deviations. Future work will focus on fully automating the workflow from point cloud acquisition to high‐accuracy numerical simulation. This will further improve the applicability of the method for digital twin and performance analysis of long‐span spatial structures.