Abstract: An experimental investigation of branch-rotated T joints was carried out in which the branch was rotated by 45[degrees] with respect to its longitudinal axis. The main experimental parameters were the ratio of the width of the branch to the width of the chord, [beta]', with 0.38 [less than or equal to] [beta]' [less than or equal to] 1.00, and the ratio of the width of the chord to the thickness of the chord, 2[gamma], with 16.7 [less than or equal to] 2[gamma] [less than or equal to] 33.3. Experimental results from 27 specimens showed that the ultimate strength increases and the failure mode changes under compression loading due to the rotated branch. The failure modes were out-of-plane bending yielding of the upper flange of the chord for [beta]' [less than or equal to] 0.85 and buckling of the chord web for 0.85 < [beta]' [less than or equal to] 1.00. Theoretical analysis was then carried out to compute the ultimate strength using a yield-line model for [beta]' [less than or equal to] 0.85 and web buckling model for [beta]' = 1.00. Linear interpolation was used for 0.85 < [beta]' < 1.00. Design formulas are presented for estimating the strength of T joints with a rotated branch.
Key words: hollow structural section, truss T joint connections, yield-line instability, chord web instability, design equations.
Resume : Une etude experimentale a ete realisee sur des raccords en T avec branche orientee, dans laquelle une branche etait orientee a 45[degrees] par rapport a l'axe longitudinal. Les principaux parametres experimentaux etaient le rapport de largeur du branchement par rapport a la membrure, [beta]', ou 0,38 [less than or equal to] [beta]' [less than or equal to] 1,0, avec un rapport largeur sur epaisseur de la membrure, 2[gamma], ou 16,7 [less than or equal to] 2[gamma] [less than or equal to] 33,3. Les resultats experimentaux de 27 echantillons ont montre que la resistance a la rupture augmente et que le mode de defaillance change sous un effort de compression en raison de la branche orientee. Les modes de defaillance etaient la rupture par flexion hors plan de la semelle superieure de la membrure pour [beta]' [less than or equal to] 0,85 et par flambage de l'ame de la membrure pour 0,85 < [beta]' [less than or equal to] 1,0. L'analyse theorique a ensuite ete effectuee pour calculer la resistance a la rupture en utilisant un modele de courbe de defaillance pour [beta]' [less than or equal to] 0,85 et un modele de flambage de l'ame de la membrure pour [beta]' = 1,0. Une interpolation lineaire a ete utilisee pour les valeurs 0,85 < [beta]' < 1,0. Les formules de calcul pour estimer la resistance des raccords en T munis d'une branche orientee sont presentees.
Mots cles: element de structure creux, connexions des fermes par raccords en T, instabilite de courbe de defaillance, instabilite de l'ame des membrures, equations de calcul.
[Traduit par la Redaction]
1. Introduction
The use of hollow structural sections (HSS) in welded trusses has become more widespread in recent decades due to the advantages of high torsional strength, light weight, and economy of design. There are many types of truss joints, and their behaviors are different from one another. The behavior of T joints composed of square hollow sections with the branch member under compression loading is generally governed by the ratio of the width of the branch to the width of the chord, [beta] [b.sub.1]/B (refer to Fig. 1 for definitions of [b.sub.1] and B). CIDECT (1986) specifies two basic failure modes depending on the value of [beta]. For [beta] [less than or equal to] 0.85, the general failure mode of T joints is out-of-plane bending yielding of the upper flange of the chord, and loads continuously increase with no distinct point of ultimate load. For 0.85 < [beta]' [less than or equal to] 1.00, chord web buckling generally governs the failure of T joints, and the ultimate strength increases slowly with an increase in [beta].
[FIGURE 1 OMITTED]
Since the strength of T joints increases as the ratio of the width
of the branch to the width of the chord increases, it is thought that if the branch is rotated by 45[degrees] so as to increase the effective value of [beta] to [beta]' = [(2).sup.1/2][beta], the strength could potentially be increased without any additional material or fabrication requirements. To examine the validity of this proposition, the behavior of T joints in which the branch is rotated by 45[degrees] with respect to its axis (hereinafter referred to as branch-rotated T joints) and welded to the upper flange of the chord was experimentally investigated, focusing on deformation and ultimate strength. To verify the branch-rotated effect, the behavior of normal T joints was also investigated. Simple compression tests were performed for both branch-rotated T joints and normal T joints by taking the ratio of the width of the branch to the width of the chord, [beta] or [beta]', and the ratio of the width of the chord to the thickness of the chord, 2[gamma]B/T, as the main experimental parameters (refer to Fig. 1 for the definition of T). Based on the experimental results, theoretical analysis is carried out and design strength formulas are derived.
When the behavior of a joint is governed by flange yielding, the strength of joints is generally computed by a yield-line method. Although Zhao and Hancock (1991) considered membrane action and strain hardening, most yield-line methods to determine the joint strength are simple analyses that do not consider membrane action or strain hardening. A typical example of the method is the CIDECT (1986) model, which is adopted by many standards (CISC 1997; European Committee for Standardization 2003). There are also …
No comments:
Post a Comment