Sie sind auf Seite 1von 6

Articles

Ronald D. Ziemian* DOI: 10.1002/stco.201810011


Jean C. Batista Abreu

Design by advanced analysis – 3D benchmark problems


Members subjected to major- and minor-axis flexure

Steel design codes continue to be expanded to permit the use of may be curved and/or the compressive axial force varies
more advanced methods of non-linear analysis. Designers look- along a member’s unbraced length.
ing to employ such methods need to validate their analysis soft- Key requirements for applying a design method that
ware and, just as importantly, verify their ability to utilize it prop- relies heavily on a rigorous second-order analysis are con-
erly. The literature contains many benchmark problems and firmation of the software’s capabilities and, just as impor-
­results to help achieve this, but nearly all are limited to two-­ tantly, the designer’s proficiency in using such software.
dimensional behaviour. This paper is intended to contribute a new Given that the literature now contains many plane frame
set of benchmark problems in order to help satisfy the need for a examples that can be used for such verifications, the pur-
database of examples in which accurate modelling of three-­ pose of this paper is to provide results for three benchmark
dimensional or spatial behaviour is essential. problems in which significant twist develops primarily as a
result of second-order effects. These three-dimensional ver-
1 Introduction ification studies, which focus on beams and beam-columns,
are intended to represent only a few of the array of bench-
More and more specifications worldwide are providing de- mark problems that an engineer should complete before
signers with opportunities to take advantage of the ever-in- using sophisticated levels of analysis to design steel struc-
creasing capabilities of non-linear frame analysis software. tures.
One such example is the American Institute of Steel Con-
struction’s Specification for Structural Steel Buildings. In 2  Why a little twist hurts
2010 this specification [1] included completely revised pro-
visions for design by inelastic analysis. With an emphasis Consider the beam-column shown in Fig. 1. The member
on taking advantage of the potential redistribution of is subjected to a compressive force Pu and uniformly dis-
member and connection forces and moments which can tributed load ωuy, which is assumed to track the direction
occur as a result of localized yielding, these provisions rep- of gravity and therefore always remain vertical. For now, it
resent a logical extension of the AISC’s direct analysis is further assumed that no out-of-plane distributed load
method, which first appeared in the AISC specification in and end moments are applied (ωux = 0, MX = 0, MY = 0).
2005 [2]. Most recently, the AISC specification [3] has The analytical model also includes an initial out-of-plane
again been expanded to provide opportunities for design- imperfection with amplitude δox at mid-span. In applying
ers who employ a second-order elastic analysis that is more the given loads, the member will not only bend in- and out-
rigorous than the traditional P-Δ and P-δ methods expected of-plane, but will also twist. Based on the resulting mid-
for the application of the original direct analysis method. span deflections δx, δy and θ, an equilibrium analysis on the
By using an analysis that also accounts for twist when for- deformed shape will result in the following major- and mi-
mulating equilibrium equations on the deformed shape nor-axis bending moments at mid-span:
and includes the explicit modelling of system sway and
member out-of-straightness imperfections as well as an in- Mux′ = Mst cos θ − M wk sin θ
(1)
elastic stiffness reduction factor, the compression strengths Muy′ = Mst sin θ + M wk cos θ
of members may now be taken as their cross-sectional
strength. Although this new approach is based on elastic where Mst = ωuy L2/8 + Pu δy and Mwk = Pu (δox + δx).
analysis and therefore does not permit an inelastic redistri-
bution of internal forces and moments, it can be particu- It is important to note that the moments Mux′ and Muy′,
larly useful for design situations in which the definitions of which refer to the local cross-section axes at mid-span, are
the unbraced lengths or effective lengths are not readily the moments that are intended to be used in any interac-
apparent. Such cases include, but are not limited to, arches tion equation design checks.
and unbraced Vierendeel girders, in which the member To gain some insight into the magnitude of the twist-
ing rotation q, let us consider the case in which the com-
* Corresponding author: pressive axial force is inconsequential, Pu ≈ 0. Substituting
ziemian@bucknell.edu

24 © Ernst & Sohn Verlag für Architektur und technische Wissenschaften GmbH & Co. KG, Berlin · Steel Construction 11 (2018), No. 1
R. D. Ziemian/J. C. Batista Abreu · Design by advanced analysis – 3D benchmark problems

Fig. 1.  General description of member investigated

the moments from Eq. (1) into AISC’s governing interac- Fig. 2.  Upper bounds on permitted twist angle θ
tion equation, given by

Mux′ Muy′ analysis capabilities. Given that most of today’s structural


(2) + ≤ 1.0
ϕMnx ϕMny engineering designs are completed using finite element
analyses that employ elastic line or framework elements,
and, assuming q is small (cos q = 1 and sin q = q), yields the focus of this paper is to establish such benchmark re-
sults. With regard to torsional resistance, analysis pro-
Mst M θ grams often differ in their ability to model uniform and/or
(3) + st ≤ 1.0
φMnx φMny non-uniform torsion. Programs employed in this study
which only account for the continuous flow of shear
Solving Eq. (3) for an inherent limitation on q results in stresses on the section, commonly called uniform or St.
Venant torsional resistance, include SAP2000 [4] and
 −1  Strand7 [5]. Analysis software used which also models
 Mst  Mnx
θ≤ 
(4) − 1 warping restraint torques include the line elements within
 φMnx   Mny
  ADINA [6], ANSYS [7] and ABAQUS [8]. In all cases, ana­
lysis results are provided from models in which the
For the case of no axial force and a range of major- to mi- cross-sectional properties were provided; in other words,
nor-axis available strength ratios Mnx/Mny, Fig. 2 provides the general or integrated properties, such as the second
a plot of the required-to-available strength ratio for major moment of area, were defined as input. Although not re-
axis bending Mst/ϕMnx versus the maximum values of twist ported here, nearly identical results are obtained when the
θ permitted per Eq. (4). The curves indicate that attempts dimensions of the sections are provided as input, with the
to utilize a member’s major-axis bending capacity to the software computing the integrated section properties. To
full will require that the amount of twist resulting from an provide comparisons with results obtained using sec-
initial out-of-plane member imperfection (or some other ond-order inelastic or geometric material non-linear anal-
perturbation, such as a small amount of out-of-plane eccen- yses (GMNIA), complete shell and brick element models
tricity of the applied load) will need to be limited. It should were analysed using Strand7 [5]. With the exception of the
also be noted that using cross-sections with large major- to shell and brick element results presented, so-called shear
minor-axis available strength ratios further exasperates this deformations were not included in any of the analyses that
limitation on θ – a situation that, for efficiency purposes, employed line elements.
might often be the case for members being designed to re-
sist primarily major-axis bending. Again, the reason for this 4 Common properties and support conditions of members
behaviour is that assessing equilibrium on the deformed ­investigated
shape (reality!) requires that large major-axis demands be
resolved into minor-axis demands as the section twists. If All of the beams and beam-columns studied in this paper
the member has a comparatively smaller amount of mi- have the same geometric properties, material properties
nor-axis available strength, such minor-axis demands can and support conditions. The member comprised a
have a significant impact on the design interaction equa- W460×971 (W18×65) cross-section with a span and un-
tion checks. braced length of 6.1 m (240 in). The material stiffness and
strength are E = 200 GPa (29 000 ksi), G = 77 GPa
3  Analysis software employed
1 d = 467 mm (18.2 in), bf = 193 mm (7.59 in), tw = 11.4 mm
Several commercially available analysis programs were (0.45 in), tf = 19.1 mm (0.75 in), A = 12 300  mm4 (19.1 in2),
used to produce the results presented in this paper. In gen- Ix = 445×106 mm4 (1070 in4), Iy = 22.8×106 mm4 (54.8 in4),
eral, all of these programs are known to possess rigorous J = 1140×103 mm4 (2.73 in4), Cw = 1140×109 mm6
three-dimensional second-order or geometric non-linear (4240 in6)

Steel Construction 11 (2018), No. 1 25


R. D. Ziemian/J. C. Batista Abreu · Design by advanced analysis – 3D benchmark problems

(11 154 ksi) and Fy = 345 MPa (50 ksi). When results are Given that there is no axial force, most engineers
provided in accordance with AISC’s Design by Advanced and design codes would assume that a first-order elastic
Elastic Analysis method, all section properties (A, Ix, Iy, J, analysis is adequate. For those referring to the AISC
Cw) are reduced by a factor of 0.8 when computing re- speci­fication, the reduction factor of 0.8 associated with
quired strengths. When employing AISC’s Design by Ad- the direct analysis method for frame design would not be
vanced Inelastic Analysis method, the original section applied. Using this, the beam would be shown to deflect
properties and material yield strength are multiplied by a vertically in-plane and horizontally out-of-plane, but no
reduction factor of 0.9. twist would be computed along its length. With q = 0, and
All members are simply supported (pinned support per Eq. (1), the resulting mid-span moments would be
at one end, roller support at the other) with longitudinal equal to the applied end moments. It should be further
rotation or twist at the member ends restrained. Member noted that the specific values defined for the applied end
ends are assumed to be warping-free, with cross-flange moments were purposely selected such that the beam
bending not resisted at the supports. The axial, flexural would be assessed as adequate in meeting the require-
and uniformly distributed loads provided below are as- ments of AISC’s interaction equation, Eq. (2). For refer-
sumed to be factored and are applied proportionally. All ence, the major- and minor-axis demands equal 85 and
applied loads track the original global coordinate system. 15 % of the member’s available major- and minor-axis
For example, a uniformly distributed gravity load is ap- ­moment capacities respectively; M ux’/fMnx = 0.85 and
plied in the vertical direction throughout the loading his- Muy’/fMny = 0.15. Given that the interaction equation
tory; in other words, the applied loads do not track the results in a value of unity for an applied load ratio
­
element axes. ALR = 1.0, the member proportion could even be defined
For this compact section, unbraced length and sup- as optimal. A summary of these results is provided in the
port conditions, the design capacities of the member are second column of Table 1.
computed to have an axial strength fP n = 952 kN A rigorous second-order analysis that accounts for
(214 kip), major-axis flexural strengths fMnx = 380.9 kNm twist and thus assesses equilibrium on the deformed shape
(3371 kip-in) with Cb = 1 for a uniform moment distribu- provides significantly different results. The third column of
tion and fMnx = 434.2 kNm (3843 kip-in) with Cb =1.14 Table 1 provides results for an analysis that models uni-
for a moment gradient produced by a uniformly distrib- form or St. Venant torsion only. Finally, the fourth and fifth
uted load, and a minor-axis flexural strength fMny = columns of Table 1 provide results for analyses that model
114.4 kNm (1013 kip-in). With these values, the major- to both St. Venant and warping torsion, with the fifth column
minor-axis available strength ratios are 3.33 with Cb = 1.0 also including the 0.8 reduction on stiffness required by
and 3.79 with Cb = 1.14. It should also be noted that the the AISC’s Design by Advanced Elastic Analysis method.
cross-section axial strength is fPns = fPy = 3825 kN Although a summary of the observations that can be made
(860 kip). from these results will be provided at the conclusion of this
paper, it should be noted that second-order inelastic anal-
5  Benchmark problems and results yses – which employ either 4080 shell elements or 9120
brick elements with both explicitly modelling yielding ac-
All of the problems below were created and analysed using centuated by the presence of residual stresses – show that
imperial units (kips and inches). Results presented below the member would fail at an applied load ratio of approx.
were obtained by converting moments and deflections ALR = 0.78, or 78 % of the applied end moments. Note that
from imperial to SI units. these latter analyses included a reduction factor of 0.9 on
both the material stiffness and yield strength (0.9E, 0.9Fy),
5.1  Beam subjected to biaxial bending but no axial force per the provisions of the AISC’s Design by Advanced Ine-
lastic Analysis method.
Most structural engineers and design codes associate sec-
ond-order effects with additional moments produced by 5.2  Beam-column with in-plane distributed load
eccentricities due to laterally displacing gravity loads in
frames (P-D sway moments) and axial loads in members Using the same geometry, member properties and support
(P-d member moments). The primary purpose of this exam- conditions as defined above, the applied loadings shown
ple is to provide numerical results for a situation in which in Fig. 1 include a range of combinations of in-plane uni-
no axial force exists but there are still significant differ- formly distributed load wuy and axial force Pu. In all cases,
ences between results obtained ensuring equilibrium on the out-of-plane distributed load and applied end moments
the deformed shape versus the undeformed shape; in other are taken as wux = 0, MX = 0 and MY = 0. For each combi-
words, second-order effects can still be significant even in nation of loading, three sets of results are provided in Ta-
the absence of axial force. ble 2. It should be noted that the AISC has adopted signif-
Using the member properties and support conditions icant portions of Table 2 for inclusion in their commentary
defined above and with reference to Fig. 1, the applied on the 2016 Specification for Structural Steel Buildings [3].
loading includes only the equal and opposite end moments All results are based on assuming that the AISC’s direct
MX = 323.7 kN-m (2865 kip-in) and MY = 17.17 kN-m analysis method is being applied, and thus include a stiff-
(152 kip-in). The axial force and uniformly distributed ness reduction factor of 0.8. Included in these results are
loads shown in Fig. 1 will be taken as Pu = 0 and wux = wuy values obtained for the AISC’s governing interaction equa-
= 0, and without including any initial imperfections, e.g. tion for assessing members subjected to flexure and axial
member out-of-straightness. force, which is

26 Steel Construction 11 (2018), No. 1


R. D. Ziemian/J. C. Batista Abreu · Design by advanced analysis – 3D benchmark problems

Table 1.  Results for beam subject to biaxial bending and no axial force

Effects at First-order Rigorous second-order elastic analysis


mid-span elastic analysis J ≠ 0, Cw = 0, E J ≠ 0, Cw ≠ 0, E J ≠ 0, Cw ≠ 0, 0.8E
ALR 1.0 1.0 1.0 1.0
Mux′ kNm (kip-in) 323.7 (2865) 293.7 (2599) 321.3 (2843) 304.2 (2692)
Muy′ kNm (kip-in) 17.2 (152) 137.3 (1215) 43.8 (388) 112 (992)
dy mm (in) 16.9 (0.665) 53.2 (2.09) 19.5 (0.768) 50.1 (1.97)
dx mm (in) 17.5 (0.689) 107 (4.20) 38.4 (1.51) 111 (4.39)
q rad (deg) 0.00 (0.00) 0.384 (22.0) 0.0828 (4.74) 0.300 (17.2)
AISC Int. Eq. (2) 1.00 1.97 1.23 1.78
ALR, Int. Eq ≈ 1 1.00 0.80 0.88 0.81
Mux′ kNm (kip-in) 323.7 (2865) 256.4 (2269) 283. 7 (2511) 260.0 (2301)
Muy′ kNm (kip-in) 17.2 (152) 39.0 (345) 29.5 (261) 36.7 (325)
dy mm (in) 16.9 (0.665) 16.3 (0.641) 16.0 (0.628) 20.0 (0.787)
dx mm (in) 17.5 (0.689) 33.8 (1.33) 26.7 (1.05) 40.1 (1.58)
q rad (deg) 0.00 0.0978 (5.60) 0.0505 (2.89) 0.0874 (5.01)
AISC Int. Eq. (2) 1.00 1.01 1.00 1.00

P  Muy′  produced by cross-flange bending along the length of the


u + 8 M ux′ + Pu
  ≤ 1.0 when ≥ 0.2 member. In applying this design method, Eq. (5) would be
φPn 9  φMnx φMny  φPn
(5) assessed using an available compressive strength that is
Pu M Muy′  Pu based on the member’s cross-sectional strength, fPn =
+  ux′ +  ≤ 1.0 when < 0.2
2φPn  φMnx φMny  φPn 3825 kN (860 kip).

For this specific member and loading, he AISC’s available 5.3 Beam-column with in-plane and out-of-plane distributed
major- and minor-axis moment strengths would be taken loads
as fMnx = 434.2 kNm (3843 kip-in), with Cb =1.14 for a
moment gradient produced by a uniformly distributed This benchmark problem is nearly the same as the previ-
load, and fMny = 114.4 kNm (1013 kip-in). The available ous problem, except that i) an out-of-plane uniformly dis-
axial strength would be a function of the design method tributed load wux is applied in addition to the in-plane dis-
being employed. tributed load wuy and axial force Pu, and ii) the initial mem-
In the rows labelled (a) in Table 2, the member is as- ber out-of-straightness imperfection is never included.
sumed as nominally straight and a second-order elastic Although the AISC’s Design by Advanced Elastic Analysis
analysis would result in in-plane behaviour only; no twist, would require modelling of a member imperfection, this
out-of-plane deflection or minor-axis bending moments provision has been purposely disregarded in order to pro-
would be computed. In applying the AISC’s direct analysis vide a benchmark problem with a geometry that is rela-
method provisions, Eq. (5) would be assessed using an tively easy to create. In Table 3, results are provided with
available compressive strength that is based on an effective rows (a), (b) and (c) corresponding to the analysis assump-
length equalling the unbraced length of the member, tions presented in the previous section.
fPn = 952 kN (214 kip).
In accordance with the AISC’s Design by Advanced 6 Conclusions
Elastic Analysis method, additional results are provided in
rows labelled (b) and (c). In this case a more rigorous elas- Using a laterally unbraced I-shaped member subjected to
tic analysis procedure is employed which ensures equilib- various combinations of in-plane and out-of-plane loading
rium and compatibility are satisfied on the deformed shape effects as well as an applied axial force, results are pro-
of the member, and thus includes second-order effects that vided which are intended for benchmarking and verifying
can be attributed to both P-δ and twist. This design method non-linear frame analysis software. In addition to provid-
also requires an initial out-of-plane member imperfection, ing numerical results from commercial programs that vary
which in this study is modelled by distorting the original by the degree to which torsional resistance is modelled,
geometry of the member in the shape of a sine curve with several general observations can be made, including…
an amplitude of L/1000 at mid-span. The analyses used to 1. For situations in which the major-axis flexure capacity
produce the results in rows (b) and (c) differ only with re- of an I-shaped member is closely approached, only a
gard to the method used for modelling torsional resistance. small amount of twist need occur before such major-axis
The analytical results given in row (b) only include the demands are resolved into minor-axis demands.
torsional resistance provided by the St. Venant stiffness GJ, 2. Second-order or geometrical non-linear analysis is re-
whereas the results in row (c) come from an analysis that quired in order to assess such demands, even for cases
also includes the warping resistance of the section ECw in which there is no axial force.

Steel Construction 11 (2018), No. 1 27


R. D. Ziemian/J. C. Batista Abreu · Design by advanced analysis – 3D benchmark problems

Table 2.  Results for beam-column subjected to in-plane distributed load


Load combination No. 1 2 3 4
Pu kN (kip) 0 333.6 (75.0) 556.0 (125.0) 778.4 (175.0)
wuy kN/m (kip/ft) 58.4 (4.0) 43.8 (3.0) 29.2 (2.0) 14.6 (1.0)
(a) 271.2 (2400) 207.1 (1833) 139.8 (1237) 70.8 (626)
Mux′ kNm (kip-in) (b) 269.6 (2386) 206.3 (1826) 139.5 (1235) 70.5 (624)
(c) 271.1 (2399) 207.0 (1832) 139.8 (1237) 70.8 (626)
(a) 0 0 0 0
Muy′ kNm (kip-in) (b) 29.1 (258) 26.4 (234) 21.7 (192) 34.9 (309)
(c) 6.3 (56) 11.7 (104) 15.8 (140) 32.1 (284)
(a) 14.7 (0.580) 11.2 (0.443) 7.58 (0.299) 3.84 (0.151)
dy mm (in) (b) 17.6 (0.694) 13.3 (0.524) 8.69 (0.342) 5.11 (0.201)
(c) 15.0 (0.589) 11.7 (0.460) 8.09 (0.318) 4.71 (0.186)
(a) 0 0 0 0
dx mm (in) (b) 24.6 (0.967) 24.2 (0.951) 21.2 (0.833) 35.5 (1.40)
(c) 5.44 (0.214) 11.0 (0.435) 15.6 (0.616) 32.8 (1.29)
(a) 0 0 0 0
q rad (deg) (b) 0.108 (6.19) 0.0790 (4.53) 0.0471 (2.70) 0.0358 (2.05)
(c) 0.0233 (1.33) 0.0290 (1.66) 0.0266 (1.52) 0.0260 (1.49)
(a) 0.62 0.77 0.87 0.96
AISC Int. Eq. (5) (b) 0.87 0.75 0.58 0.62
(c) 0.68 0.62 0.53 0.60
(a) 1.61 1.29 1.15 1.04
ALR, Int. Eq ≈ 1 (b) 1.03 1.07 1.13 1.09
(c) 1.28 1.26 1.25 1.11
ALRult GMNIA 1.23 1.15 1.16 1.10
Note: (a) traditional GNA, dox = 0; (b) rigorous GNIA, J ≠ 0, Cw = 0; (c) rigorous GNIA, J ≠ 0, Cw ≠ 0

Table 3.  Results for beam-column subjected to in-plane and out-of-plane distributed loads
Load combination No. 1 2 3 4
Pu kN (kip) 0 222.4 (50.0) 444.8 (100.0) 667.2 (150.0)
wuy kN/m (kip/ft) 58.4 (4.0) 43.8 (3.0) 29.2 (2.0) 14.6 (1.0)
wux kN/m (kip/ft) 5.84 (0.4) 4.38 (0.3) 2.92 (0.2) 1.46 (0.1)
(a) 271.2 (2400) 205.9 (1822) 138.9 (1229) 70.3 (622)
Mux′ kNm (kip-in) (b) 248.3 (2198) 199.6 (1767) 137.4 (1216) 70.1 (620)
(c) 267.0 (2363) 203.8 (1804) 138.2 (1223) 70.2 (621)
(a) 27.1 (240) 26.6 (235) 25.4 (225) 22.1 (196)
Muy′ kNm (kip-in) (b) 111.6 (988) 66.4 (588) 41.8 (370) 26.8 (237)
(c) 54.2 (480) 44.1 (390) 34.9 (309) 25.5 (226)
(a) 14.7 (0.580) 11.2 (0.440) 7.54 (0.297) 3.81 (0.150)
dy mm (in) (b) 42.4 (1.67) 19.6 (0.773) 10.0 (0.394) 4.34 (0.171)
(c) 19.2 (0.755) 13.7 (0.539) 8.76 (0.345) 4.19 (0.165)
(a) 28.7 (1.13) 27.9 (1.10) 26.7 (1.05) 23.1 (0.909)
dx mm (in) (b) 97.0 (3.82) 62.7 (2.47) 41.7 (1.64) 27.4 (1.08)
(c) 52.1 (2.05) 43.7 (1.72) 35.6 (1.40) 26.4 (1.04)
(a) 0 0 0 0
q rad (deg) (b) 0.323 (18.5) 0.158 (9.05) 0.0700 (4.01) 0.0232 (1.33)
(c) 0.101 (5.79) 0.0685 (3.92) 0.0404 (2.31) 0.0165 (0.945)
(a) 0.86 0.86 0.95 1.02
AISC Int. Eq. (5) (b) 1.55 1.07 0.74 0.48
(c) 1.09 0.88 0.68 0.47
(a) 1.16 1.15 1.05 0.99
ALR, Int. Eq ≈ 1 (b) 0.84 0.97 1.12 1.12
(c) 0.92 1.05 1.18 1.22
ALRult GMNIA 0.93 1.00 1.09 1.18
Note: (a) traditional GNA, dox = 0; (b) rigorous GNIA, J ≠ 0, Cw = 0; (c) rigorous GNIA, J ≠ 0, Cw ≠ 0

28 Steel Construction 11 (2018), No. 1


R. D. Ziemian/J. C. Batista Abreu · Design by advanced analysis – 3D benchmark problems

3. The beneficial effects of employing non-linear analysis [6]  ADINA (v. 9.0.1), ADINA R&D, Inc., USA, www.adina.com
programs that account for both uniform and non-uni- [7] ANSYS Mechanical APDL (rel. 16.0), USA, www.anysis.
form (warping) torsional resistance result in reductions com
in deflections, twist and minor-axis bending moment [8]  ABAQUS/CAE (v. 6.14), Dassault Systèmes Americas Corp.,
USA, www.abaqus.com
demands.
Keywords:  three-dimensional benchmark problems; design; ad-
These benchmark problems are intended to be valuable vanced analysis
when verifying the proper use of advanced methods of
non-linear analysis, using a variety of software packages,
for situations in which the accurate modelling of spatial
behaviour is important. The problems also highlight impor- Authors
tant effects associated with twist which are relevant as the Ronald D. Ziemian, PhD
design profession moves towards employing three-dimen- Associate Dean of Engineering and Professor
sional non-linear analysis. Department of Civil and Environmental Engineering
Bucknell University
References 235 Dana Engineering Bldg.
701 Moore Ave.
[1] AISC: Specification for Structural Steel Buildings, ANSI/ Lewisburg, PA 17837
AISC 360-10, American Institute of Steel construction, Chi- USA
cago, Il, 2010. ziemian@bucknell.edu
[2] AISC: Specification for Structural Steel Buildings, ANSI/
AISC 360-05, American Institute of Steel construction, Chi- Jean C. Batista Abreu, PhD
cago, Il, 2005. Visiting Assistant Professor
[3] AISC: Specification for Structural Steel Buildings, ANSI/ Department of Mechanical Engineering
AISC 360-16, American Institute of Steel construction, Chi- Bucknell University
cago, Il, 2016. 318 Dana Engineering Bldg.
[4]  SAP2000 (v. 17.2), Computers and Structures, Inc., USA, 701 Moore Ave.
www.csiamerica.com Lewisburg, PA 17837
[5]  Strand7 (rel. 2.4.6), Strand7 Pty Ltd, Australia, www.strand7. USA
com jean.batistaabreu@bucknell.edu

Steel Construction 11 (2018), No. 1 29

Das könnte Ihnen auch gefallen