International Journal of Steel Structures 17(2): 775-787 (2017)

DOI 10.1007/s13296-017-6031-7
ISSN 1598-2351 (Print)
ISSN 2093-6311 (Online)

www.springer.com/journal/13296

Buckling Length Analysis for Compression Chord in

Cold-Formed Steel Cantilever Truss
Nirut Konkong, Trakool Aramraks, and Kitjapat Phuvoravan*
Department of Civil Engineering, Faculty of Engineering, Kasetsart University, Bangkok 10900, Thailand

Abstract

The lateral buckling of a cold-formed steel truss structure was investigated. Both an experimental test and finite element
analysis presented a good correlation in the failure mechanism of specimens. The critical buckling load of the compression truss
chord was analyzed using a Southwell plot and the analytical method. The transfer matrix method and the governing equation
of an Euler column were used to analyze the critical buckling load of a multi-step, uniform chord memberwith lateral spring
supports. Expressions for the buckling length factor presented according to AISI (2012) and the EN 1993-1-3 (2006)
specification were compared by experimental testing and the analytical method. The buckling length factorof the experimental
test and the analytical model confirmed that the buckling length of a compressed truss chord was less than the distance between
the lateral bracing.

Keywords: cold-formed steel cantilever truss, finite element analysis, lateral buckling, buckling length factor, multi-step
uniform column

1. Introduction mainly focused on the lateral bracing of the truss chord

and rigid side support was assumed constant along the
Cold-formed steel has been used by many segments of truss length. These results showed that a truss with
the building industry. In building construction, the channel intermediate elastic supports will buckle depending on
sections are used in steel framing systems such as the stiffness of the elastic supports. A compression truss
columns, beams and truss structures. Cold-formed steel is chord is laterally supported at an intermediate point by a
formed using a cold reduction process in which the web member of the truss. The chord member can be
procedure increases the strength and hardness and decreases idealized as a uniform bar with several spring supports.
the weight. Advances and progress in material sciences The stability problems of a barwith several elastic supports
have invented high-performance construction materials using the governing differential equation for buckling
such as the very thin, cold-formed steel section with high were studied by Li (2001 and 2003). The results of the
yield strength. Therefore, improvement of reliable and present method are in agreement with those calculated by
cost effective design procedures is essential (Bedair, 2014). the finite element and the Ritz method.
For example, by assuming the lateral buckling of the truss In the design procedure for cold-formed steel truss
chord element as an isolated element, the restraining of structures, the compression capacity of the truss element
the diagonal members are assigned zero stiffness. In can be estimated as a column formula which depends on
reality, the diagonal members also contribute to the the buckling length factor of the element. The buckling
protection of the lateral stability of structures. length factor isused to correct the actual unrestrained
The lateral bracing requirement is necessary for lateral length of a compression element to account for the dominant
stability protection of the compressed structural member. boundary conditions. The design standards AISI (2012)
The problem of lateral stability of the truss chord is well and EN 1993-1-3 (2006) -Design of steel structures- Part
known, and research work on the topic has already been 1-3 assign a buckling length that is equal to the distance
published by Iwicki (2007 and 2010). These studies between the lateral bracing in the chords. However,
recent studies have shown that the buckling length was
Received August 4, 2016; accepted December 6, 2016; less than the distance between the bracing in the chords
published online June 30, 2017 (Iwicki, 2003; Biegus, 2011; Jankowska-Sandberg and
© KSSC and Springer 2017
Kołodziej, 2013). The chords are generally made using
*Corresponding author continuous profiles to which web elements are joined
E-mail: fengkpp@ku.ac.th directly or by gusset plates. Thus, the lateral buckling
776 Nirut Konkong et al. / International Journal of Steel Structures, 17(2), 775-787, 2017

behavior of the chord element is impacted by the rigidity 2. Experimental Test

of the web elements, joint stiffness, metal roofing material
and the non-uniform axial loads (Biegus, 2015). The main purpose of the investigations was to
In the current research, the lateral buckling of a cold- determine the load and deformation relationship between
formed steel truss was analyzed using the experimental varied angles of diagonal truss members. The specimen
test, finite element analysis (FEA) and the analytical types were able to be classified bythe angles of the
method. The analytical model was developed to estimate diagonal truss members which rotate changing from the
the critical buckling load and the out-of-plane buckling reference line. Four series of specimens were prepared
length factor (K-factor) of the compression truss chord. It with the angle of the diagonal at values of 19, 42, 66 and
included the effect of web element stiffness. The K- 90 degrees, respectively.
factors as recommended by AISI (2012) and, EN 1993-1- A typical specimen of the cantilever truss structure is
3 (2006) were compared with the experimental test and illustrated in Fig. 1. The truss length (L) is 3.0 m, the
the analytical model. depth in the middle (H) is 0.5 m. The samples were made

Figure 1. Test setup.

 Buckling Length Analysis for Compression Chord in Cold-Formed Steel Cantilever Truss 777

from cold-formed lipped channel sections (C-102×51× specimens were tested using a hydraulic jack to apply a
12.5×1.0) with a yield strength of 550 MPa (G550). M5 concentrated load on the joint of the top chord member.
Grade 8.8 high strength steel bolts were used on all tested Linear variable differential transformers (LVDTs) were
configuration connections. Acantilever truss was set up in used to measure the displacement values with continuous
a rectangular steel frame which formed the skeleton of recording by the data acquisition system. Strain gauges
the test apparatus. The lateral support of the top chord wereused to measure the strain distribution in the axial
was provided by a roller connected to the steel frame. The direction of the bottom chord. The arrangement of the

Figure 2. Finite element models for the truss structure.

778 Nirut Konkong et al. / International Journal of Steel Structures, 17(2), 775-787, 2017

strain gauge is shown in Fig. 1(b). During testing, the σtrue = σtest ( 1 + εtest ) (1)
concentratedload was applied incrementally by the jack
εtrue = ln ( 1 + εtest ) (2)
and operated until the observed truss buckled and the
system lost its stability. The labeling format of the specimen where σ test and ε test are stress and strain obtained from
was CST-θ-n. where CST denotes the cold-formed steel tensile testing and ε true and ε test are data for use in FEA
cantilever truss, θ is the angle of the diagonal chord and as shown in Fig. 2(e). The cold-formed steel was obtained
n specifies the number of the repeated test. through tensile testing of coated coupons according to
ASTM A370-07b (2007) as shown in Table 1(a). The true
3. Finite Element Analysis stress and strain data of cold-formed steel were emplaced
into the multi-linear isotropic model. The linear elastic
The FEA was then performed using the ANSYS software model was used for the M5 Grade 8.8 steel bolt. The steel
package to develop the finite element model of the cantilever bolt properties were obtained through ASTM A325M
truss structure. The cold-formed steel was conducted (1994) as shown in Table 1(b).
using shell elements (SHELL281) which are eight-node The element mesh size was controlled to be close to the
elements with six degrees of freedom at each node, i.e. 1.0 aspect ratios of both shell and solid elements. The
translations in x, y and z-directions, and rotations about x, boundary conditions were assigned according to the test
y and z-axes as shown in Fig. 2(a). The steel bolt was specimen supporting conditions by restraining the displace-
modeled using the solid elements (SOLID186) which are ments of the respective nodes as shown in Fig. 2(f). To
twenty-node elements with three degrees of freedom at provide a solution for the model, nonlinear buckling analysis
each node, i.e. translations in the nodal x, y, and z- was obtained to achieve the ultimate load capacity of the
directions as shown in Fig. 2(b). The elements between truss structure that provides greater accuracy than a linear
the steel bolt and cold-formed steel were modeled using elastic analysis (Rust, and Schweizerhof, 2003). The non-
the contact elements (CONTAC174 and TARGET170) as linear buckling analysis procedure was imposedthe imperfect
shown in Fig. 2(c) (ANSYS, 2010, 2014). The pure shape onto the existing geometry. The first buckling mode
penalty contact algorithm was used for solving constrained shape of linear buckling analysis was transformed into
optimization problems. It can be thought of as placing the geometry model (Imaoka, 2005; Zhou et al., 2009).
stiffness springs between the two parts that have come Geometric nonlinearity analysis was achieved in ANSYS’s
into contact with each other as shown in Fig. 2(d). The static module by a large defection option. Large deflection
actual stress-strain data obtained from cold-formed steel is categorized as either large deflection or large strain
tensile testing was converted to the true data for use in which depended on the element type. It is the concept
FEA by using Eq. (1)-(2): used to describe geometric nonlinearities.

Table 1(a). Material properties of cold-formed steel, G550

Test number
SD COV
CT-1 CT-2 CT-3 CT-4 CT-5 Mean
t (mm) 1.045 1.045 1.044 1.045 1.045 1.045 0.0004 0.04
w (mm) 12.64 12.65 12.59 12.58 12.52 12.59 0.0522 0.41
Fy (MPa) 608.5 605.8 605.9 617.5 606.5 608.8 4.97 0.81
Fu (MPa) 627.0 622.5 625.5 630.0 620.0 625.0 3.89 0.62
E (GPa) 213.51 213.49 213.49 213.58 213.58 213.53 44.18 0.02
Fu/Fy 1.03 1.02 1.03 1.02 1.02 1.02 0.0054 0.53
Remark: Fy is the yield stress, Fuis the ultimate stress, E is the modulus of elasticity, t is the plate thickness, w is the plate width, COV
is the coefficient of variation and SD is the standard deviation.

Table 1(b). Material properties of bolts

Test No.
SD. COV
BT-1 BT-2 BT-3 BT-4 BT-5 Mean
d (mm) 4.93 4.97 4.97 4.97 4.94 4.95 0.019 0.393
Fy (MPa) 930.0 949.0 927.0 976.0 872.0 930.8 38.23 4.107
Fu (MPa) 963.0 930.0 963.0 997.0 909.0 952.4 33.91 3.560
E (GPa) 204.08 204.08 204.09 204.10 204.09 204.08 7.39 0.003
Remark: d is the bolt diameter.
 Buckling Length Analysis for Compression Chord in Cold-Formed Steel Cantilever Truss 779

Figure 3. The failure mode of test and finite element models.

4. Experimental Results have shown that the Southwell plot can be theoretically
and practically used for estimating the lateral torsional
The experimental tests studiedthe lateral buckling behavior buckling load of the beam (Mandal and Calladine, 2002;
of the chord element with different angles of diagonal Kalkan, 2010). In the presentresearch, the axial load is
configuration. CST-19-n failed due to the combined lateral equal to the average elastic strain multiplied by the
buckling at the bottom chord with local buckling in the modulus of elasticity (E) and the area cross-section of the
diagonal element as shown in Fig. 3(a). CST-42-n and cold-formed steel section. The lateral deformation was
CST-66-n failed due to the combined distortion and especially considered at the maximum deformation point
lateral buckling in the compression chord as shown in from the LVDT. The Southwell plots for the compression
Fig. 3(b) and 3(c). CST-90-n failed due to lateral buckling chord truss are presented in Fig. 5.
at the bottom chord and due to connection failure as
shown in Fig. 3(d). The load-lateral deformation curve 5. Analytical Studies on the Truss Chord
and buckling shape of the chord member are shown in Member
Fig. 4. The results of the tests are summarized in Table 2.
The test results showed that it was common that the In general, the chord element is usually designed with
limitation of the load was taken as the maximum load in a constant stiffness along its length. Furthermore, it is a
the experimental load-deformation plots of the specimens. simplified design procedure in that the stiffness of other
Nonetheless, the loads were not represented as the members such as the diagonal and tension chord members
criticalbuckling load of the chord member. Thus, there is are neglected. However, in an effort to achieve effective
a need for an efficient and accurate method to determine cost designs, several researchers have developed an
the critical buckling load of the compression chord member analytical technique which considers internal factors such
from the experimental data. This efficient method was as the stiffness of the tension chord, and the vertical and
developed by Southwell (1932). The Southwell plot is a diagonal elements. The present research further developed
method used to estimate the buckling load (Pcr) of a the practical analytical method to solve the stability
column from its experimental data. problems of the compression chords with the stepped
The concept of the Southwell plot is the calculation of internal forces together with aconsideration of the lateral
the slope of lateral deformation versus the lateral deformation stiffness from the diagonal and vertical members.
over the axial load curve, as it can be analytically proved The transfer matrix method and the governing equation
that the slope is the critical load of the column. Studies for the buckling of an Euler column were used to simulate
780 Nirut Konkong et al. / International Journal of Steel Structures, 17(2), 775-787, 2017

Figure 4. Results for test specimens.

Table 2. Results summary of cantilever truss structures

Lateral deformation (mm)
Load applied (N)
Series n Experimental FEA
Experimental FEA L1 L2 L3 L1 L2 L3
CST-19-n 1 5,970 0.60 -1.28 -8.12
2 5,730 5,619 0.56 -1.54 -7.38 0.84 -0.42 -6.75
3 5,638 0.56 -1.85 -7.38
Mean 5,779 0.57 -1.56 -7.63
COV 2.97 4.59 -18.36 -5.59
CST-42-n 1 6,321 -2.35 -4.36 -6.72
2 6,904 6,532 -2.35 -4.36 -6.50 -0.97 -3.61 -6.25
3 5,685 -2.18 -4.52 -6.74
Mean 6,303 -2.29 -4.42 -6.65
COV 9.67 -4.21 -2.06 -1.96
CST-66-n 1 5,933 -2.46 -6.42 -10.71
2 6,624 6,547 -2.46 -6.40 -10.28 -2.11 -5.37 -8.64
3 7,158 -2.35 -6.67 -10.72
Mean 6,571 -2.42 -6.50 -10.57
COV 9.35 -2.81 -2.32 -2.36
CST-90-n 1 6,860 -5.44 -9.96 -10.16
2 6,700 6,813 -5.51 -10.03 -10.45 -4.71 -8.01 -10.39
3 6,779 -5.72 -10.20 -8.78
Mean 6780 -5.56 -10.06 -9.80
COV 1.18 -2.58 -1.20 -9.10
 Buckling Length Analysis for Compression Chord in Cold-Formed Steel Cantilever Truss 781

Figure 5. Southwell plots of compression chords.

the lateral buckling behavior of cold-formed steel compression torsional spring concept which is also related to the
chord members as presented in Fig. 6. torsional stiffness of the cross-section of the tension
The compression chord can be idealized for buckling chord as shown in Eqs. (5) and (6):
analysis as a continuous beam that is braced by elastic
kcon ϕcon = Fld sinθ (5)
translation springs, which correspond to the stiffness of
the diagonal member at each panel point as shown in Fig. in which
6(b). The analytical model was analyzed using the concept
Δrot
of Hooke’s law to determine the lateral spring stiffness. ϕcon = -------------
- (6)
ld sinθ
Hooke’s law is a principle of physics that states the
amount of force applied to a spring is proportional to the where kcon is the torsion stiffness in the tension chord
amount of deformation. The force (F) applied to a spring member as kcon=(GJ)t,cL/(IacIbc), (GJ)t,c is the torsional
is determined by Eq. (3): rigidity for the tension chord element, L, lac and lbc are the
lengths of tension chord elements, ld is the length of the
F = ks – i ΔT (3)
web element (diagonal or vertical member), and ϕcon is
where ks−i is the spring stiffness and ΔT is the total the rotation angle of the web element referred to in Fig.
deflection. The lateral spring stiffness was obtained from 7. Thus, the rotational deflection can be defined by Eq.
a plane truss analysis considering the cantilever beam (7):
with a load at the end as shown in Fig. 6(c). The lateral
1 2
spring stiffness was newly proposed and included for the Δrot = -------- F ( ld sinθ ) (7)
kcon
torsion stiffness of the tension chord and the flexural
stiffness of the web element in the truss. The web element Consider a cantilever beam with a concentrated load
was modeled as a vertically oriented, elastically cantilevered acting upward at the free end. The flexural deflection at
beam. The total deflection at the end of the cantilever the free end is given by Eq. (8):
beam, can be derived from the sum of the rotational
3
deflection of the beam (Δrot), and flexural deflection Fl d
Δflex = --------------
- (8)
(Δflex) using Eq. (4): 3 ( EI )d
ΔT = Δrot + Δflex (4)
where ( EI )d is the flexural rigidity of the web element.
The rotational deflection was calculated using the The total deflection at the tip of the web element includes
782 Nirut Konkong et al. / International Journal of Steel Structures, 17(2), 775-787, 2017

Figure 6. Analytical model of the compression chord in cantilever truss.

the sum of the flexural deflection and rotational deflection shown in Fig. 8. Using linear static analysis, the unit
as shown in Eq. (9): force was applied to the truss model that related to the
specimen types in the experimental test.
3 2
Fl d Fld sinθ
ΔT = ⎛ -------------- -⎞
The governing equation for the buckling of an Euler
- + ------------------------------------- (9)
⎝ 3 ( EI )d ( GJ )t, c L ⁄ ( lac lbc )⎠ column for a multi-step chord can be written as shown in
Eq. (11):
Thus, the equation for the equivalent lateral spring
4 2
stiffness prediction of web element is as shown in Eq. d yi 2 d yi
--------- + ki --------- = 0 (11)
(10): dx dx
F F
ks – i = ----- = ---------------------------------------------------------------- 2
ΔT Fld
3 2
Fld sinθ where ki = Pi ⁄ ( EI )i , Pi is the axial compressive force
⎛ --------------- + ------------------------------------- -⎞
⎝ 3 ( EI )d ( GJ )t, c L ⁄ ( lac lbc )⎠ multi-step column, (EI)i is the flexural rigidity of the
multi-step column and i is the number of segments. The
3 ( EI )d ( GJ )t, c L general solution for the buckling of an Euler column is
= ----------------------------------------------------------------------------
2 2
- (10)
ld ( 3 ( EI )d lac lbc sin θ + l d ( GJ )t, c L ) given by Eq. (12):
yi ( x ) = Ai + Bi x + Ci sin ( ki x ) + Di cos ( ki x ) (12)
The α i-parameter has a distribution ratio of the external
axial force actions resulting from the load applied in the where the constants Ai, Bi, Ci, and Di can be obtained by
experimental test. The external axial force action of the using boundary conditions. The relationships among the
member can be calculated using linear static analysis as displacement, rotation, bending moment and shearing
 Buckling Length Analysis for Compression Chord in Cold-Formed Steel Cantilever Truss 783

Figure 7. Simplified torsion stiffness in tension chord and flexural stifness in web element.

Figure 8. Distribution ratio of the external axial forces (αi).

force of an Euler column can be written using Eq. (13): Euler buckling and Eq. (11) can be expressed as Eq. (14):
2
dy dy ⎧ yi ( x ) ⎫ ⎧ Ai1 ⎫
θ i ( x ) = -------i , Mi ( x ) = –( EI )i --------2-i and ⎪ ⎪ ⎪ ⎪
dx dx ⎪ θi( x ) ⎪ ⎪ Bi2 ⎪
⎨ ⎬ = [ Hi ( x ) ] ⎨ ⎬ (14)
dM dy ⎪ Mi( x ) ⎪ ⎪ Ci3 ⎪
Vi ( x ) = ---------i – P -------i (13) ⎪ ⎪ ⎪ ⎪
dx dx ⎩ Vi ( x ) ⎭ ⎩ Di4 ⎭

Thus, the relationship between the general solution of where:

784 Nirut Konkong et al. / International Journal of Steel Structures, 17(2), 775-787, 2017

Figure 9. Relationship between the parameters at the two ends of segment i and i+1.

1 x sin ( ki x ) cos( ki x )
0 1 kcos( ki x ) –ksin ( ki x )
[ Hi ( x ) ] = (15)
0 0 Psin ( ki x ) Pcos ( ki x )
0 –P 0 0

The relationship of column parameters at the two ends

of an i-step column is shown in Fig. 9. It can be written
as Eq. (16):

⎧ yi ( li ) ⎫ ⎧ yi ( 0 ) ⎫
⎪ ⎪ ⎪ ⎪
⎪ θi ( li ) ⎪ ⎪ θi( 0 ) ⎪
⎨ ⎬ = [ Ti ] ⎨ ⎬ (16)
⎪ Mi ( li ) ⎪ ⎪ Mi( 0 ) ⎪ Figure 10. Discontinuity of joint note in analytical model.
⎪ ⎪ ⎪ ⎪
⎩ Vi ( li ) ⎭ ⎩ Vi ( 0 ) ⎭

where [ Ti ] is the transfer matrix, in which the parameters

⎧ yn ( l n ) ⎫ ⎧ yi ( 0 ) ⎫
are transferred from end x=0 to those at the lower end ⎪ ⎪ ⎪ ⎪
x=l of the i-step column. It can be written as Eq. (17): ⎪ θn ( ln ) ⎪ ⎪ θi( 0 ) ⎪
⎨ ⎬ = [ T]⎨ ⎬ (20)
–1 ⎪ Mn ( ln ) ⎪ ⎪ Mi( 0 ) ⎪
[ Ti ] = [ H ( li ) ] [ H ( 0 ) ] (17) ⎪ ⎪ ⎪ ⎪
⎩ Vn ( ln ) ⎭ ⎩ Vi( 0 ) ⎭
The interfaces of the two neighboring step columns are
in which [T]=[T n ][T n−1 ] ... [T 2 ][T 1 ]. The matrix [T ] has
required to be continuous for the displacement, slope,
the form shown in Eq. (21):
bending moment and shear force and can be written as
Eq. (18):
T11 T12 T13 T14
⎧ yi + 1 ( 0 ) ⎫ ⎧ yi ( l ) ⎫ T21 T22 T23 T24
⎪ ⎪ ⎪ ⎪ [T] = (21)
⎪ θi + 1 ( 0 ) ⎪ ⎪ θi( l ) ⎪ T31 T32 T33 T34
⎨ ⎬= ⎨ ⎬ (18)
⎪ Mi + 1 ( 0 ) ⎪ ⎪ Mi( l ) ⎪ T41 T42 T43 T44
⎪ ⎪ ⎪ ⎪
⎩ Vi + 1( 0 ) ⎭ ⎩ Vi ( l ) ⎭ Figure 10 showsthe continuity among the displacements,
bending moments and shear forces at the joint node of the
The continuity of the column parameters at the
i and i+1 step columns. However, there is a discontinuity
boundary of two neighboring i and i+1 step columns is
in the shear force at this point, caused by the axial
required to be continuous as shown in Eq. (19):
compression force (α i P) and deformation of the spring
representing the diagonal member stiffness. Thus, the
⎧ yi + 1 ( li + 1 ) ⎫ ⎧ yi ( 0 ) ⎫
⎪ ⎪ ⎪ ⎪ transfer matrix [Ti ] must be substituted by [T is ] which
⎪ θi + 1 ( li + 1 ) ⎪ ⎪ θi ( 0 ) ⎪ can be determined as shown in Eq. (22):
⎨ ⎬ = [ T(i + 1) ] [ Ti ] ⎨ ⎬ (19)
⎪ Mi + 1 ( li + 1 ) ⎪ ⎪ Mi ( 0 ) ⎪
⎪ ⎪ ⎪ ⎪ 1 0 00
⎩ Vi + 1( li + 1 ) ⎭ ⎩ Vi( 0 ) ⎭ 0 1 00
[ Tis ] = [ Ti ] (22)
The final equation of the multi-step column with lateral 0 0 10
spring support can be written as Eq. (20): ks, i Pi 0 1
 Buckling Length Analysis for Compression Chord in Cold-Formed Steel Cantilever Truss 785

The governing differential equation of column buckling Table 3. Stiffness of lateral spring support
was derived by considering the continuity of the column ks
parameters at the boundary of the two neighbors. The Sample
Eigenvalue equation for the buckling of a multi-step ks-1 (N/m) ks-2 (N/m) ks-3 (N/m)
column can be explained using Eq. (18). However, it is CST-19-n 31.51 31.49 31.51
independent of the boundary conditions for different cases CST-42-n 35.84 31.49 35.84
of restraints. The Eigenvalue equation for hinged-hinged CST-66-n 38.65 31.49 38.65
end supports and lateral spring supports is presented in CST-90-n 41.99 31.49 41.99
Eq. (23):
Table 4. αi -parameter
⎧ 0 ⎫ T11 T12 T13 T14 ⎧ 0 ⎫
⎪ ⎪ ⎪ ⎪ αi
⎪ θn ( ln ) ⎪ T21 T22 T23 T24 ⎪ θ n ( 0 ) ⎪ Name of series
α1 α2 α3 α4
⎨ ⎬= ⎨ ⎬ (23)
⎪ 0 ⎪ T31 T32 T33 T34 ⎪ 0 ⎪ CST-19-n 0.06 0.21 0.15 0.58
⎪V (l ) ⎪ ⎪ ⎪
⎩ n n ⎭ T41 T42 T43 T44 ⎩ Vn ( 0 ) ⎭ CST-42-n 0.08 0.20 0.21 0.52
CST-66-n 0.10 0.18 0.27 0.45
The non-trivial solution requires that the determinant of CST-90-n 0.11 0.17 0.30 0.42
the matrix [T ] is equal to zero. Thus, the Eigenvalue
equation is Eq. (24):
These results are also plotted in Fig. 11 and show a
T12 T34 – T14 T32 = 0 (24) comparison of the K- factors. These were obtained from
experimental tests, the analytical model, AISI (2012) and
Solving the Eigenvalue equation gives the k-parameter. Eurocode 3. The limit of the K-factor designed using
The critical buckling load of the multi-step column is AISI (2012) can be used, namely, K=1 for the compression
2
calculated as Pcr = k ⁄ EI . The critical buckling is equal element in trusses. For the EN 1993-1-3 (2006) specifications,
to the axial force in the chord member at overall stability. the value of the K-factor of the truss chord can be 0.9
The buckling length factors (K-factors) of the compression and 1 for in-plane buckling and out-of-plane buckling,
chords for the lateral buckling can be obtained by respectively.
experimental results and the analytical model for the The K-factors of the studied cases result in values of K
cantilever truss can be determined using Eq. (25): =0.95-0.98 which vary with the series groups of specimens.
These values showed that the buckling length of a
π EI
K = --- ------
- (25) compressed truss chord is less than the distance between
L Pcr
the lateral bracings. The overall loading capacity of the
truss was obtained from both the experimental test and
6. Buckling Length Factor of Compression FEA, and is related to the K-factor calculated by the
Truss Chord analytical model. The lower the K-factor, the higher the
overall loading capacity. Except for CHT-19, the effect of
The critical buckling of the truss chord member was local buckling of the diagonals was clearly noticed in the
calculated by setting the determinant of matrix [T ] to be experimental test and FEA, which limited the overall
zero. The lateral spring stiffness and α i-parameter are buckling. However, the lateral buckling mode was the
presented in Tables 3-4 which were used to create the second mode of failure in CHT-19, which was called the
matrix [T ]. post-buckling failure mode. Based on the concept of truss
The results of critical buckling loads and K-factors are buckling, the K-factor with the angle of the diagonal
tabulated in Table 5. In each of these tests, the critical changing was characterized as the critical buckling strength
buckling loads and K-factor were established lower than of the truss structure. This is an important consideration
design standards. for truss configuration design. Both the Southwell method

Table 5. Critical buckling load and buckling length factors of compression chord
% Load increase from
Critical buckling load (N.) K-factor (N.)
Name of series design standard (%)
Test Analytical Test Analytical Test Analytical
CHT-19-n 18,115 18,335 0.98 0.97 4.12 6.28
CHT-42-n 18,235 18,679 0.98 0.96 4.12 8.51
CHT-66-n 18,417 18,679 0.97 0.96 6.28 8.51
CHT-90-n 18,620 19,377 0.97 0.95 6.28 10.80
786 Nirut Konkong et al. / International Journal of Steel Structures, 17(2), 775-787, 2017

Figure 11. The comparison of the K-factor.

and the analytical model were presented in the previous procedure. The results of the K-factors of chords achieved
section. These were the effective methods for critical in this research have the exemplary characteristics of a
buckling prediction and can be applied to all the types of cantilever truss structure. The proposed procedure is an
truss structures. efficient method for buckling analysis of multi-step members
which can be used to the continuous truss chord element
7. Conclusions of all truss configurations.

The purpose of this research was to provide an Acknowledgments

experimental and analytical basis to predict the behavior
of a cold-formed steel cantilever truss. The experimental The authors wish to acknowledge the Kasetsart University
results were evaluated against the predicted values using Research and Development Institute (KURDI) for providing
the ANSYS software package. The local buckling, the financial support and to BlueScope Lysaght Limited
combined distortion with lateral buckling and the connection (Thailand) for material support.
failure mode were presented in the tests and FEA results.
The FEA including material nonlinearity, geometric References
nonlinearity and contact equations were investigated. The
FEA and experimental failure mode were compared and
AISI (2012). North American specification for the design of
presented. A good correspondence between the FEA and cold-formed steel structural members. American Iron and
experimental response was found for all the cases of truss Steel Institute, Washington D.C., USA.
specimens. The Southwell plot technique was applied for ANSYS (2010). Elements Reference Release 11.0. ANSYS
the analysis of the lateral buckling load of the compression Inc., Pennsylvania, USA.
chord member ANSYS (2014). ANSYS User’s Manual Revision 15.
The analytical solutions of the K-factors for the compression ANSYS Inc., Pennsylvania, USA.
chord element were calculated for a multi-step uniform ASTM (1994). A325M high strength bolts for structural
chord member with lateral spring supports subjected to steel joints [Metric]. American Society for Testing and
several concentrated axial loads that were determined by Materials, West Conshohocken. PA, USA.
an Eulerdifferential equation for buckling analysis with ASTM (2007). A370-07b standard test methods and definitions
the transfer matrix method. The transfer matrix method is for mechanical testing of steel products. American
adequate to implement the buckling problem with multi- Society for Testing and Materials, West Conshohocken.
step bars, lateral spring stiffness and concentrated axial PA, USA.
Bedair, O. (2014). “Design space representation of channel
forces. The arrangements of diagonal members affect the
members for industrial applications.” International
lateral stability of a cantilever truss structure. All the results Journal of Steel Structures, 14(3), pp. 529-538.
showed that the K-factors for a compression chord are Biegus, A. (2015) “Trapezoidal sheet as a bracing preventing
under the limits specified by AISI (2012) and Eurocode flat trusses from out-of-plane buckling.” Archives of Civil
3. K-factors predictions of AISI (2012) and Eurocode 3 and Mechanical Engineering, 15(3), pp. 735-741.
were too conservative for compression chord members Biegus, A. and Wojczyszyn, D. (2011). “Studies on buckling
subject to out-of-plane buckling. The intention of this lengths of chords for out-of-plane instability.” Archives of
analytical method is to act as a guideline in the design Civil and Mechanical Engineering, 11(3), pp. 507-517.
 Buckling Length Analysis for Compression Chord in Cold-Formed Steel Cantilever Truss 787

EN 1993-1-3 (2006). Eurocode 3: Design of steel structures: step non-uniform columns with arbitrary distribution of
Supplementary rules for cold-formed members and flexural stiffness or axial distributed loading.” International
sheeting-Part 1-3: General rules London. European Journal of Mechanical Sciences, 43(2), pp. 349-366.
Committee for Standardization, UK. Li, Q. S. (2001). “Buckling analysis of non-uniform bars
Imaoka, S. (2008). “Analyzing Buckling in ANSYS Workbench with rotational and translational springs.” Engineering
Simulation.” ANSYS Advantage. 2(1), ANSYS Inc., USA. Structures, 25(10), pp. 1289-1299.
Iwicki, P. (2007). “Stability of trusses with linear elastic Li, Q. S. (2001). “Buckling of multi-step non-uniform beams
side-supports.” Thin-Walled Structures, 45(10-11), pp. with elastically restrained boundary conditions.” Journal
849-854. of Constructional Steel Research, 57(7), pp. 753-777.
Iwicki, P. (2009). “Comparison of classical Winter's bracing Mandal, P. and Calladine, C. R. (2002). “Lateral-torsional
requirements of compressed truss chord with stability buckling of beams and the Southwell plot.” International
analysis of 3D truss-model.” Proc. 2009 Appl. Math. Journal of Mechanical Sciences, 44(12), pp. 2557-2571.
Mech., 9, pp. 247-248. Rust, W., and Schweizerhof, K. (2003). “Finite element limit
Iwicki, P. (2010). “Sensitivity analysis of critical forces of load analysis of thin-walled structures by ANSYS (implicit),
trusses with side bracing.” Journal of Constructional LS-DYNA (explicit) and in combination.” Thin-Walled
Steel Research, 66(7), pp. 923-930. Structures, 42(2-3), pp. 227-244.
Jankowska-Sandberg, J and Ko³odziej, J. (2013). “Experimental Southwell, R. V. (1932). “On the analysis of experimental
study of steel truss lateral-torsional buckling.” Engineering observations in problems of elastic stability.” Proceedings
Structures, 46, pp. 165-172. of the Royal Society of London, 135, UK, pp. 601-616.
Kalkan, I. (2010). “Application of Southwell method on the Zhou, Z., Meng, S. and Wu, J. (2009). “Stability analysis of
analysis of lateral torsional buckling tests on reinforced prestressed space truss structures based on the imperfect
concrete Beams.” International Journal of Engineering, truss element.” International Journal of Steel Structures,
Research and Development, 2(1), pp. 58-66. 9(3), pp. 253-260.
Li, Q. S. (2001). “Analytical solutions for buckling of multi-