A systematic intelligent arch-forming framework for large-span arch bridges is detailed in a new paper published in Engineering, addressing persistent construction barriers for arch infrastructure built within rugged mountain canyon environments. Authored by researchers from Chongqing Jiaotong University and Shenzhen University, the work outlines three interconnected technical modules covering arch-forming calculation, segment manufacturing control, and automated arch rib installation attitude adjustment, with full-scale validation on the 504 m main span Deyu Expressway Wujiang Bridge, a concrete-filled steel tube arch bridge located in Guizhou Province, China.

Mountainous construction sites create layered challenges for traditional cable-stayed fastening–hanging cantilever assembly, including extreme temperature swings, uneven canyon wind fields, limited staging space, high-altitude work hazards, and cumulative geometric offset of prefabricated arch rib segments. Conventional cable force calculation methods such as zero-moment, zero-displacement, and fixed-length cable approaches treat pre-tensioning and post-cable-removal structural states as isolated systems, lacking capacity to quantify how temporary cable tension shapes final bare arch alignment. Meanwhile, conventional “N+1” physical preassembly consumes extensive labor, land and heavy machinery, while single-point total station measurement delivers incomplete three-dimensional pose data and requires repeated manual on-site correction of arch rib segments.

The research first establishes a full-process multi-objective optimization calculation model linked by mechanical state connection equations that correlate cantilever construction force conditions with the geometry of the bare arch after temporary cable removal. The model integrates constraint equations for post-forming arch rib alignment tolerance, tower lateral deviation limits, and upper/lower bounds of allowable cable tension, paired with an objective function minimizing the coefficient of variation of cable forces across all construction phases. A nondominated genetic algorithm solves the constrained optimization to generate a unified one-time tensioning scheme compliant with structural safety and geometric targets.

Second, the team introduces a digital preassembly manufacturing control scheme built on terrestrial laser scanning point cloud data and building information modeling. The workflow captures high-fidelity point cloud models of individual arch rib segments via multi-station scanning, then applies a dual-weight point cloud-to-BIM registration algorithm. Semantic weights are assigned to structural components while distribution weights partition each chord’s longitudinal space, forming a weighted least-squares alignment function solved through singular value decomposition to derive optimal joint splicing offsets, eliminating reliance on repeated physical trial assembly during fabrication.

Third, an original-shape-restoration automatic attitude adjustment strategy governs on-site segment placement. The workflow converts manufacturing-coordinate point clouds into unloaded target installation geometry, superimposes finite-element-calculated deformations from self-weight and temperature loads to generate loaded target pose data, then registers scanned as-installed point clouds to compute a full 4 × 4 transformation matrix for three-dimensional correction. Separate adjustment mathematical models support small cantilever ground-jack support and large cantilever cable-stayed construction phases, calculating precise adjustment quantities for jacks or stay cables to enable single-step positioning of each arch rib segment.

The integrated intelligent arch-forming method was deployed across core construction phases of the Wujiang Bridge project, verifying consistent geometric control and operational efficiency gains. The paper notes the framework delivers actionable technical guidance for low-labor, resource-conserving, automated construction of large-span arch bridges built via prefabricated segment cantilever hoisting, and identifies further research needs to adapt the digital control workflow for cast-in-place concrete arch structures.

The paper “Intelligent Forming of Large-Span Arch Bridges: Methodology and Engineering Applications,” is authored by Jianting Zhou, Yanliang Du, Yin Zhou, Jinyu Zhu. Full text of the open access paper: https://doi.org/10.1016/j.eng.2025.10.022. For more information about Engineering, visit the website at https://www.sciencedirect.com/journal/engineering.