Experimental investigation and simulation analysis of cast-steel joints under vertical pressure
Zhihao Li,Yizhong Zhang,Wenfeng Du,Liming Zhu
DOI: https://doi.org/10.1038/s41598-024-62138-4
IF: 4.6
2024-05-18
Scientific Reports
Abstract:The joint made of cast steel is frequently utilized within a treelike column structure to ensure a smooth transition. It is of great significance in ensuring the overall structural safety, but currently, the mechanical property and bearing capacity of this type of joint cannot be fully understood. This study investigates the load characteristics of three-forked cast steel joints through concrete experiments, finite element analysis, and regression method formula derivation, filling the gap in mechanical properties and calculation formulas of forked cast steel joints. Initially, a comprehensive model of the cast-steel joint, sourced from a practical engineering, underwent vertical load testing. Detailed scrutiny of stress distribution and vertical displacement of the tested joint was conducted based on the experimental outcomes. Subsequently, a finite element model of the tested joint was constructed using SolidWorks and subjected to analysis via ANSYS. The numerical findings were juxtaposed with experimental data and extrapolated to encompass other parametric scenarios. Ultimately, a regression analysis method was employed to derive a calculation formula for the load-carrying capacity of branch-bearing cast-steel joints. The regression analysis method can accurately obtain the load-bearing capacity calculation formula for tree-shaped joint models and can be extended to determine corresponding branch and main pipe dimensions, as well as the deviation angle between branches and the main pipe, under known load conditions. This improves design efficiency and accuracy. Comparative analysis reveals a substantial concurrence among experimental, finite element analysis, and formula-based predictive outcomes. The maximum error between experimental results and those obtained from finite element analysis is 9.02%. The maximum error between the results calculated using the load-bearing capacity formula derived from regression methods and those from finite element analysis is only 1.9%.
multidisciplinary sciences