Bending resistance of flat elliptical tube skeleton sleeve screw splicing in solar greenhouse

Greenhouses are one of the most commonly-used facilities in modern agriculture. A solar greenhouse skeleton can also be constructed with the flattened-oval steel tubes. However, it is still lacking in the mechanical response of the spliced joints in the solar greenhouse skeletons, leading to limited application in engineering practice. This study aims to elucidate the dominant failure mechanisms of such joints. A reliable analytical expression was also established for their initial bending stiffness. Experimental observations were integrated with the numerical simulations. The constitutive relations of the tube material and connecting screws were first characterized before the simulations. A streamlined full-scale experimental program was then performed to verify the fidelity of the numerical model. Measured responses were also compared with the theoretical predictions. A systematic numerical investigation was conducted using a calibrated model. After that, 60 configurations were defined by five tube-section geometries, three screw quantities, and four splice lengths. The moment–rotation features werethen extracted to examine the pre-yield mechanical behavior. The results revealed that there were three failure modes: (1) Compressive crushing of the upper tapered segment was characterized by progressive local buckling and plastic indentation ofthe tube wall, corresponding to the superior load-bearing performance. The compressive region sustained the substantial deformation before the instability occurred. (2) Screw-hole wall bearing or shear failure often occurred when the local bearing stresses exceeded the material capacity, leading to the abrupt degradation of the load transfer between the connected segments. The rotational stiffness substantially reduced the ultimate moment. (3) A coupled interaction between tapered-end crushing and screw-hole degradation was represented to produce a compound and highly sensitive failure. The relative dominance of the two coupled processes was varied in the geometric tolerances and assembly conditions, resulting in greater variability in the joint stiffness and capacity. A parametric analysis demonstrated that the initial bending stiffness depended mainly on the tube-section height, screw quantity, splice length, and assembly gap. Large section heights enhanced the stiffness and bendingresistance, whereas there was a decrease in the susceptibility to local deformation, particularly when an insufficient number of screws was employed. The joints with only three screws—particularly when combined with the splice lengths shorter than 0.8H—shared early rotation localization near the screw group, thus promoting premature stiffness loss. Lengthening the spliceregion engaged a larger portion of the tube wall in the load-transfer mechanism. The stress concentrations were reduced for themore desirable deformation that was governed by gradual compression at the upper taper. A splice-length range of 0.8H-1.2H enhanced the stiffness among all section types, and then delayed the local damage. The moment–rotation curves demonstrated an initial linear regime followed by progressive stiffness degradation. Curve fitting was performed on all configurations. It was found that the initial bending stiffness was incorporated with the sectional properties and screw-group mechanics. Numerical simulation was verified by the high accuracy with a conservative boundary, suitable for the structure and performance evaluation. Additionally, the configurations with only three screws or splice lengths below 0.8 H were not recommended for the greenhouse skeletons, leading to low stiffness and premature failure, particularly when the tube-section height exceeded 80 mm. The stiffness and capacity formulas were directly incorporated into the global strength and stability of the greenhouse skeletons. Moreover, the simplified full-scale validation, calibrated simulation, and analytical modeling were combined to transfer into the structural joint configurations structural joint configurations. The rotational stiffness coefficient K_θ of the splicing node is positively correlated with the splicing length ratio L_s. Using the "K-L_s" expression obtained from the research, numerical analysis was conducted on the splicing nodes. Compared with the experimental results, the numerical analysis results were safer and the root mean square error was about 10%. The finding can also provide a transferable framework for the bending stiffness in the lightweight steel construction.