Professional Documents
Culture Documents
net/publication/320624526
Influence of shear lag coefficienton circular hollow sections with bolted sleeve
connections Civil Engineering Engenharia Civil
CITATIONS READS
2 120
4 authors, including:
Some of the authors of this publication are also working on these related projects:
ANÁLISE NUMÉRICA-EXPERIMENTAL DE LIGAÇÕES TIPO LUVA COM PERFIS TUBULARES DE AÇO View project
All content following this page was uploaded by Lucas Roquete on 26 October 2017.
Civil Engineering
Engenharia Civil
1. Introduction
Steel tubular structural sections Kvocak (2008), Martinez-Saucedo and from the tubular sections with flange
allow a higher load capacity for axial Packer (2009), Freitas and Requena studied by McGuire (1968), shown in
force, and torsion as well as for com- (2009), and others. Figure 1(a), the continuity of the profile
bined effects. Recommendations for Square hollow sections (SHS), is by a sleeve connection, which pres-
the use of circular hollow sections are rectangular hollow sections (RHS), ents a much more elegant appearance
generally based on CIDECT (2008). and circular hollow sections (CHS) are and allows the use of standardized
Many researches have been performed the tubular sections found in the mar- elements, as shown in Figure 1 (b).
in order to develop new techniques and ket. This article deals with steel CHS This new type of connection has been
propose solutions for the proper use of submitted to tensile force with bolted studied by Vieira et al. (2011), Silva
these sections as seen in works done by connections. The connection is studied (2012) and Amparo (2014), Amparo
Munse and Chesson (1963), Beke and for use in a truss system. Differently et al. (2014).
REM, Int. Eng. J., Ouro Preto, 70(4), 393-398, oct. dec. | 2017 393
Influence of shear lag coefficienton circular hollow sections with bolted sleeve connections
Figure 1
Tubular connections.
The sleeve connection is consti- of connections transfers the loading tion on the region of the cross-section
tuted by two tubes of the same diame- from the tube by its contact with the area with holes. This work shows
ter (outer tubes) connected by another bolt. This characteristic shows the one study about the influence of the
tube with smaller diameter and bolts presence of the shear lag coefficient shear lag coefficient in CHS with
(inner tube), Figure 1 (c). This type indicated by the high stress concentra- sleeve connection.
To represent the connection, the tested under tension in a controlled gauges were used for the test instrumen-
prototypes are formed by an outer tube servo-hydraulic press with 200kN of tation. Figure 2 shows the general scheme
and an inner tube (sleeve) with aligned capacity. An LVDT (Linear Variational of the experiment.The displacement rate
and crossed bolts. The prototypes were Displacement Transducer) and strain used was set to 0.4 mm/min.
(a) (c)
(b) Figure 2
Scheme of the prototypes and the experi-
ment: (a) aligned bolts; (b) crossed bolts;
(c) prototype in the servo-hydraulic press.
The prototypes have geometric groups, as shown in Table 1. Each group bolts and groups B and C are prototypes
property variations: thickness and diam- had a variation in the number of bolts. with crossed bolts. X and Y is the steel resis-
eter. Therefore they were separated into 3 Group A has prototypes with staggered tance of outer and inner tube, respectively.
Thickness (mm)
Thickness (mm)
Diameter (mm)
Diameter (mm)
Prototypes
identification
fy fu fy fu
(MPa) (MPa) (MPa) (MPa)
After experimental analysis, it was are shown in Amparo et al. (2014) and culated according to the scheme shown
observed that in some prototypes, the Amparo (2014), including one expression in Figure 3 (a) and (b); the figure also
bolts showed a flexural mode. In some for the bolt bending failure. shows the possible fracture lines for the
cases the limit state of bolt bending is The gross area and net area were two configurations of the connection. The
dominant, and not the yielding of the calculated considering the tubular cross-sections analyzed are identifiedas
gross section nor the fracture through the section as a rectangular prismatic bar 1-1, 2-2 or 3-3. This adaptation can be
effective net area, group A. More details (rectangular plate). The net area is cal- found in Amparo (2014).
394 REM, Int. Eng. J., Ouro Preto, 70(4), 393-398, oct. dec. | 2017
Lucas Roquete et al.
(a) (b)
Figure 3
Configurations used to calculate
the gross area and the net area,
(a) aligned bolts and (b) crossed bolts.
Ae=Ct An (1)
The value considered for the shear to calculate the failure modes and the modes werethe yielding of the gross sec-
lag coefficient (Ct) was equal to 1.0 for the theoretical values adapted according to tion (Y.G.S. - Nt), the fracture through the
calculation of theoretical values. Brazilian code ABNT NBR:2008, using effective net area (F.N.A. - Nt), and the
Table 2 shows the expressions used Ae as shown in Equation 1. The failure shear failure of bolt (S.B. - Fv).
Ag is the gross area; fy is the yielding resistance; LA-3-X3Y2 360.8 378.2 313.5
An is the net area; fu is the ultimate resistance; CB-3-X2Y3 271.9 259.3 313.5
Table 3 shows the experimental maxi- the experimental maximum load capacity
mum load capacity and the average load of for all of the prototypes of each group.
LA-3-X3Y2 425.7
440.4
LA-4-X3Y2 448.7 446.9
The Figure 4 shows the curves of sible to observe that the load capacity of through the effective net area (Table 2).
tension load (P) versus displacement for these prototypes was above the theoretical The theoretical values of group A use the
3 prototypes of the group A. It was pos- yielding gross section and the fracture shear lag coefficient Ct equal 1.0.
REM, Int. Eng. J., Ouro Preto, 70(4), 393-398, oct. dec. | 2017 395
Influence of shear lag coefficienton circular hollow sections with bolted sleeve connections
Figure 4
Curves of tension load (P) versus
displacement, prototypes of group A.
The average load correspond- acterize the different failure modes. In crossed bolts, the experimental load failure
ing to the fracture of the prototypes this group, the bending in bolts was the mode in the net area had a lower value than
of the group A was 440.4kN (Table first failure mode of the connection, the theoretical load, as shown in Figure
3), 16.45% higher than the expected observed experimentally. 5(a) and (b). Figure 5 shows the curves of
theoretical value (Table 2). Changes In the prototypes of the groups B and tension load (P) versus displacement for the
in the experimental curve slopes char- C, which have a sleeve connection with prototypes of groups B and C, respectively.
(a)
(b)
Figure 5
Curves of tension load (P)
versus displacement: (a) prototypes
of group B;(b)prototypes of group C.
396 REM, Int. Eng. J., Ouro Preto, 70(4), 393-398, oct. dec. | 2017
Lucas Roquete et al.
The prototypes of the group B had a different behaviour compared with type of connection given by the eq. (2).
had the average load value correspond- the prototypes of group A; all experi- This equation was proposed by AISC
ing to the fracture equal to 244.5kN mental curves of tension load (P) versus (2005) and ABNT NBR 8800:2008 for
(Table 3). This value is 10.08% lower than displacement were below the theoretical a slotted tube connected to a gusset plate,
the expected theoretical value (Table 2). values. In these prototypes, the value 1.0 (Figure 6), adapted to sleeve connections.
Meanwhile, for the prototypes of group was used as the shear lag coefficient. So, it Table 4 shows the experimental shear lag
C, the average load value corresponding is necessary to adopt a value that reduces coefficient determined by dividing the
to the fracture was equal to 413.0kN the net area of the effective net area, im- average value for the experimental load
(Table 3), 17.07% lower than the expected proving the results. corresponding to the fracture by the net
theoretical value. The international standards do not area and the ultimate stress of the inner
The prototypes of groups B and C provide a shear lag coefficient for this tube (tube where the fracture occurred).
ec
Ct = 1- (2)
lc
where: Ct is the shear lag coefficient, ec is the connection eccentricity, lc is the length of connection.
Figure 6
Representation of
value ec in tubular section.
3 0.909 0.760
CB
4 0.887 0.840
The prototypes with crossed bolts do Consequently, the stress distribution is dif- NBR. However, the use of the equation
not have a geometric configuration equal ferent in these prototypes. This fact could proposed by AISC (2005) and ABNT
to the model presented by ABNT NBR explain the differences shown in Table 4 NBR 8800 (2008) provides conservative
8800:2008 (Figure 3 (b) and Figure 6). between the experimental Ct and the Ct by results and therefore is reliable.
4. Conclusions
The goal of this work was perform consideration of the shear lag coefficient net area to an effective net area, since
theoretical and experimental evaluations (Ct) was based on the equation proposed the fractures occurred at levels below
of the influence of the shear lag coefficient by AISC (2005) and NBR. the load stipulated as standard.
on the new connection known as sleeve The methodology used in the ex- The international and national
connection. This connection is constituted periments was adequate to evaluate the standards do not yet include the sleeve
by steel hollow sections, and the proto- new connection. For group A, the ex- connections. The results and the analy-
types were formed by two tubes, the outer perimental results were good and showed ses proved the need and suitability of a
tube and the inner tube, connected with high capacity load resistance, while the formulation for the calculation of the
aligned bolts or crossed bolts in 90º. The flexural failure mode was dominant, shear lag coefficient for crossed bolts.
inner tube represented the sleeve. and the Ct=1.0. It was observed that the Thus, an equation based in AISC/NBR
The theoretical analysis of the fracture through the net area, in the to calculate the Ct, was proposed. The
modes of failure was based according prototypes with crossed bolts, needs to results showed good correlation with the
to ABNT NBR 8800:2008, and the consider a coefficient that reduces this experimental results.
Acknowledgments
The authors are grateful for the support from FAPEMIG, CNPQ, CAPES and Vallourec do Brazil.
REM, Int. Eng. J., Ouro Preto, 70(4), 393-398, oct. dec. | 2017 397
Influence of shear lag coefficienton circular hollow sections with bolted sleeve connections
References
398 REM, Int. Eng. J., Ouro Preto, 70(4), 393-398, oct. dec. | 2017