Evaluation And Modification Of T-stub Theoretical Model Based On Statistical Analysis

Since the Beiling Earthquake in the United States and the Osaka-Kobe Earthquake in Japan,the rotational deformation capacity of joints in steel frame structures has been gradually taken into account,and semi-rigid joints are gradually favored by designers because of their ability of giving consideration to both bearing capacity and deformation capacity.Among the research methods of semi-rigid joints,Eurocode 3 has adopted the component method for its clear physical meaning,strong applicability and high computational efficiency.Among the components,the equivalent T-stub in the tension zone contributes the most significant rotational deformation capacity to the whole joint,thus it is considered to be the key component of semi-rigid beam-column joints.However,research on the tension performance of T-stubs is still insufficient.Firstly,the influence of bolt preload on the initial stiffness of T-stubs is generally neglected.Secondly,there is no criteria to evaluate the accuracy of the existing models in terms of predicting the ultimate bearing capacity and ultimate deformation capacity of T-stubs.Thirdly,the experimental data of researchers are used only to verify and evaluate their own models respectively,thus has not been fully utilized.To this end,the database of tension T-stub tests is established in this study,which is then used to evaluate and modify the existing theoretical models.Firstly,53 T-stub specimens are selected from the existing literature with unspecified parameters being reasonably assumed.The initial stiffness,ultimate bearing capacity and ultimate deformation capacity of T-stubs are taken as the evaluation indexes of mechanical properties.The tensile test database of T-stubs is then established,which provides a basis for the evaluation of existing theoretical models and model modification.Secondly,starting from the classical beam theory adopted by Eurocode 3,the analytical models proposed by Jaspart,Piluso,Beg,Spyrou,Yu,Yang and Lin are introduced in detail,and their accuracy in calculating the initial stiffness,ultimate bearing capacity and ultimate deformation capacity of T-stubs are then evaluated based on the established test database.The results show that the preload of bolts has significant effect on the initial stiffness and should be taken into account in stiffness analysis.In addition,the accuracy of the existing theoretical models is not completely consistent under different failure modes when calculating the ultimate bearing capacity,and none of the models can accurately predict the ultimate deformation capacity of T-stubs.Finally,the model modification is conducted based on the analytical models proposed by Jaspart,Piluso,Spyrou and Yang.The calculation method of initial stiffness is revised by taking the influence of bolt pretension into account,and the incremental method for calculating bolt bearing capacity is modified by considering the increase of bolt force from yield to fracture under the failure mode “bolt failure with yielding of the flange”.The revised model is then assessed by the test database.The results show that the modified model has good accuracy in calculating the initial stiffness with or without bolt preload,and can also accurately predict the ultimate bearing capacity under various failure modes.Consequently,the modified model has advantages in evaluating the strength and initial stiffness of T-stubs,which may be applied to semi-rigid structural design.However,the predicted values are lower than corresponding test results for ultimate deformation capacity,which needs to be further studied in the future.