Beruflich Dokumente
Kultur Dokumente
e-mail: andreadiassilva@yahoo.com.br
e-mail: icaroprado@gmail.com
e-mail: ricardo@em.ufop.br
ABSTRACT
A new computational tool for the advanced static and dynamic analyses of steel framed structures
based on the Finite Element Method has been developed and is presented herein. The factors that
influence the structural system behavior are included in the numerical models and formulations
implemented in the CA-ASA computational tool. Among these factors are those resulting from
larger displacements (second-order effects), and the effects introduced when considering semi-rigid
connections and material inelasticity. The influence of geometrical imperfections and residual
stresses are also considered. An evaluation of the static and dynamic stability of various structural
systems, with rigid and semi-rigid connections, whose analytical or numerical solutions are found in
literature, is herein presented. This latter evaluation is used for validating the numerical
methodologies and formulations developed by the authors and implemented in the computational
system proposed.
Keywords: CS-ASA, Advanced analysis, Nonlinear static and dynamic analysis, Geometric
With the development of the civil, naval, oceanic and aeronautical industries, and the appearance of
more resistant materials and new constructive techniques, in diverse problems of the structural
mechanics, the nonlinear behavior of the systems begins to be relevant and should be considered in
the analyses. As a consequence, engineering interest in designing lighter and more slender
structures grows and the final structural system product becomes more competitive. Moreover,
these lighter and more slender structures usually cannot be analyzed and designed without their
dynamic effects also being considered. The unknown levels and characteristics of the dynamic
response can lead to system failure during the cyclic loading application due to the accumulation of
structural damages.
Therefore, with the changes of structural conceptions and requirements of more necessary
verification of the behavior of the structures, nowadays there is an increasing search for
computational systems with resources for nonlinear static and dynamic analyses. However, even
though the computational tools for nonlinear models evolve quickly, available commercial
programs with all these resources are still rarities [24]. Among these, should be mentioned the
A new computational system for advanced structural analysis (CS-ASA) is presented in this
article. The main characteristic of the program is the accomplishment of the nonlinear static and
dynamic analyses of steel plane framed structures. In these analyses, the geometric nonlinearity, or
connections, and steel inelasticity. The influence of initial geometric imperfections (out of
plumbness and out of straightness) and residual stresses can be also considered. The introduction of
all these nonlinear effects in the numerical models and formulations defines a Structural Advanced
Analysis. This methodology for the analysis and design of steel structures makes it possible to
establish the stability and strength limit of the structural system in a direct manner, without the
necessity of separate verification of each members capacity [7]. Related studies have been
developed with the objective of examining steel structure nonlinear behavior, and the following
works can be cited: Ziemian [26], Liew [9], Barsan and Chioren [3], Chan and Chui [5], Maleck
[10], Sekulovic and Nefovska [18, 19] and Thai and Kim [23]. Structural systems with known
numerical or analytical solutions are herein analyzed to validate the nonlinear formulations adopted
Developed in Fortran 95/2003 language [12] and based on Finite Element Method (FEM), the
CS-ASA performs, as already mentioned, the nonlinear static and dynamic analysis of steel framed
structures. The three steps of the numerical simulation process (pre-processing, analysis and post-
processing) are illustrated in Figure 1. Pre-processing consists in the numerical model definition
(data files input). The data related with the model is then processed to obtain the structural results
(nodal forces, nodal displacements, among others) that will be printed in text files (post-
processing). These files are exemplified in Figure 2. The nonlinear effects simulated are also shown
in Figure 1.
Recently, Prado [14] developed an interactive graphics preprocessing tool, named CS-ASA
Preprocessor, whose primary purpose is to assist the analyst in using the CS-ASA program.
Besides the structural modeling (finite element mesh, boundary conditions, applied loading, semi-
rigid joints, among others), information regarding the type of analysis as well as the solution
strategy can be generated by preprocessor developed. As result of structural modeling and analysis
definition, data files are produced, here called neutral files, which can be used by CS-ASA users as
input information. All the necessary information to perform a structural modeling will be recorded
in the FILEIN1.d file shown in Figure 2, through the use of CS-ASA Preprocessor (Figure 3a). The
data to perform a nonlinear static analysis, or even a vibration analysis of a pre-loaded structure, are
recorded in the FILEIN2.d file, and Figure 3b shows that the dialogues present in CS-ASA
Preprocessor can help the analyst in the generation of this second data file. The FILEIN3.d input
file is necessary when a transient dynamic analysis is performed. Besides the nonlinear FE
formulation, other related data on this file are: time step number, time increment and iterative
process tolerance. The nodal displacement, velocity and acceleration initial conditions, the viscous
damping and the dynamic excitation function shall also be indicated. As the two previous files, this
third input file can also be generated using the CS-ASA Preprocessor (Figure 3c).
CS-ASA
Computational System for Advanced Structural Analysis
Data Input
ANALYSIS
Result Exit
[15]. The IUP interface system and the CD two dimensional graphics library resources are also used
http://www.tecgraf.puc-rio.br/cd).
Looking to provide different analysis options of the structural system modeling, several
beam-column element formulations were adopted. The formulations implemented in CS-ASA for
the simulation of the second-order, connections and inelasticity effects, referenced in Figure 1, are
shown in Table 1. The researchers related with these formulations are also indicated in Table 1. The
These effects appear due to geometrical modification of the structure during the body deformation
process. Large displacements and rotations, but small deformations are supposed. The formulations
SOF-1 and SOF-2 are based on the Bernoulli-Eulers theory of beams and follow the updated
Lagrangian approach. The SOF-3 formulation is derived from Timoshenkos theory, enabling
second-order analysis of thick structures, and the total Lagrangian reference is used.
Unlike the criteria currently adopted in structural design, the joints between members do not
behave as in idealized conditions: perfectly rigid or ideally pinned. The connections used in current
practice present semi-rigid behavior that can substantially influence the structural system stability
and force distribution. Here, the semi-rigid connection is modeled as a spring element linking the
structural members. The connection elements are physically attached to the ends of the beam-
column element, maintaining the equilibrium and compatibility conditions, and the joint degrees of
freedom are embedded in the beam- column stiffness tangent relationship. The springs presence
introduces relative rotations in nodes of that elements ends by modifying the equations that
describe the nonlinear behavior of the structural system. Three different procedures are used to
modify such equations that consider the effects of the connections flexibility upon programming
The steel inelasticity process changes the internal force distribution in the profile cross section
and in all structure. Here, the study of the frame inelastic behavior is performed through the PHF
formulations (plastic-hinge formulations), which follow the basis of the refined plastic-hinge
method. This method monitors the process of cross section early yielding until its total
plastification, and residual tensions can also be considered. The plasticity effects associated with the
parameter changes with the level of loading by tracking the plastic region progress on the element
ends. In PHF-1 formulation, this parameter is defined from the plastic strength surface provided by
AISC [1]. In PHF-2 formulation, which follows the BS5950 [4] standard requirements, this
parameter is evaluated from the internal forces and geometrical characteristics of the section.
Finally, the AAF formulations (advanced analysis formulations), which were obtained by
modifying the PHF formulations, consider the influence of the connection flexibility in the
structures second-order inelastic analysis. Details on all these formulations and about the nonlinear
The CS-ASA still does not have an integrated graphical post-processor for a quick visualization
of structural response. The generated program results are then stored in files of different extensions
for subsequent treatment. These files are exemplified in Figure 2. The equilibrium paths, resultant
internal forces diagrams, degree of member section plastification, modes and frequencies of
vibration, connection and section hysteretic behavior, displacements, velocities and accelerations
time histories, and element internal forces are some of the results generated by the program.
3. SOME MODELING USING THE CS-ASA
This section shows some applications made with the CS-ASA program. Initially, Figure 4a
illustrates the static analysis carried out with a classic structural system known as Lee Frame [17].
By its sensitivity to second-order effects, this slender frame is commonly used to validate the
nonlinear solution strategies and the geometric nonlinear finite element formulations. The
equilibrium paths, with load and displacement limit points, were obtained by controlling the
horizontal and vertical displacements (u and v) of the point of load application. The complete
curves were achieved using an efficient load increment and iteration strategies coupled to the
Newton-Raphson method. Several of these strategies are implemented in CS-ASA and are also
described in [21].
The influence of static preloading on dynamic stability of a sinoidal arch is shown in Figure
4b. The geometric and connection nonlinearities are considered in this study. Nonlinear relationship
between the distributed load P and the archs lowest natural vibration frequency was verified for
different support conditions represented by the Sc parameter, which is defined as a function of arch
bending stiffness EI. The increase in support flexibility, decreasing the structural systems stiffness,
Figures 4c and 4d show the nonlinear static and dynamic analyses results of a two-story frame
with semi-rigid beam-column connections. The nonlinear behavior of these joints is simulated
through the moment-rotation curves that can be described by mathematical models, i.e.:
exponential, modified exponential, power with three or four parameters and multi-linear. Four types
of semi-rigid connections (named A, B, C and D) were considered. The initial stiffness of these
connections (Scini) is given in Figure 4c. Rigid connections (R) were also adopted. Geometric
imperfections were induced through the small intensity horizontal forces applied on the top of
floors. The static analysis, considering the material with elastic behavior, investigated the loading
capacity and stability of the structure. The load-top displacement curves show that the limit loads
are smaller for more flexible joints. In dynamic analysis, the frame with beam-column connections
of the type C is subjected to cyclic loading, and the history of the structure top horizontal
displacement is presented in Figure 4d. See that initially the displacement oscillates with variable
amplitudes characterizing the transient regime and after approximately 6s, enters into permanent
regime, which is perceived by the constant movement ranging around 5 cm. During the period of
constant response, the energy introduced into the structure by external load excitation and the
dissipated energy by the semi-rigid joints is approximately equal. This can be verified by looking at
Figures 4e and 4f are respectively related to static analysis where the effect of steel
inelasticity is considered, and to dynamic analysis performed with a seven-story frame with rigid
and semi-rigid joints of the types E and F. Again, see that the limit load factor increases with
connection flexibility reduction. For the frame with rigid connections, the influence of the beams
and columns plastic moment capacity in the structural behavior is evident. For the frame with semi-
rigid connection type F with nonlinear moment-rotation behavior, it is noticed that the connection
flexibility has considerable influence on the overall structural system behavior, since the connection
failure moment (101.7 kNm) is significantly less than the beams plastic moment (307 kNm).
Figure 4f shows the influence of hysteretic damping promoted by semi-rigid connections on the
Recently, a routine able to simulate the nonlinear cyclic behaviour of the material was
implemented in the CS-ASA system. Basically, this model is a natural extension of those already
implemented in CS-ASA for the static case. The methods of Newmark (integration process) and
Newton-Raphson (iteration process) are used for solving the nonlinear equations of motion. Some
examples considering the nonlinear inelastic time-history response of planar steel frames can be
found in [22]
6 120
P A1 : P h = 0.345; A 2 : P h = 0.079
0.2 L Present work Sc = 0
B1 : P h = 0.392; B2 : P h = 0.194
Present work 100 Sc = EI/L
C1 : P h = 0.521; C2 : P h = 0.474
L = 120
2 20
Sc
v A2 B2 A1 C2 C1
0
1 B1
P (N/mm)
-20
20 mm Sc h Sc
0
-40 Young's modulus: Cross section h = 20 mm
205 MPa
3 L = 1000 mm
Density: 78 kg/m
-1 -60
0 10 20 30 40 50 60 70 80 90 100 0.0 0.2 0.4 0.6 0.8 1.0 1.2
Displacements u, v P/h (MPa)
(a) Lee frame: static analysis (b) Shallow arch: free vibration analysis
15 24
A: Single web angle connections (Scini = 5424 kNm/rad) 80 Hysteresis M-c loops
Moment M (kNm)
14 (connection L)
B: Top and seat angle connections (Scini = 10759.8 kNm/rad)
13 20 40
C: Flush end plate connections (Scini = 12430 kNm/rad)
12 D: Extended end plate connections (Scini = 34804 kNm/rad) F(kN)
16 0
11 R: Rigid connections 7.5
P P
Displacement u (cm)
10 t(s) -40
F u 12 -7.5
Load P (MN)
9
R Sc W360x72 Sc 1s 1s
W310x143
8 -80
D 8 -8 -6 -4 -2 0 2 4 6 8
3.66 m
7 P P Rotation c (x10-3 rad)
6 Sc Sc
2F
4
L W360x72
5 C
W310x143
Young's modulus:
4 200 GPa 3.66 m 0
B
3 F = 0.001P
A
2 -4
Present work 6.10 m
1
Chan and Chui [5]
0 -8
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 0 2 4 6 8 10 12 14 16
Displacement u (cm) Time (s)
(c) Two-story frame: static analysis (d) Two-story frame: dynamic analysis
1.4 100
Present work: linear connections Rigid connections
1.3
Present work: nonlinear connections Linear connections type F
1.2 Sekulovic and Nefovska [19] 80
50.0 kN/m Nonlinear connections type F
1.1 F1 u
F2(kN)
1.0 2 67.5 kN/m 60
F2 24
Displacement u (m)
0.9
Load factor ()
R
0.6 1 67.5 kN/m 20
E F2
0.5 Young's modulus: 205 GPa
1 67.5 kN/m
0.4 Residual stress: 235 MPa F
2 0
F2 = 2 F1 = 24 kN
0.3 1 67.5 kN/m
R: rigid connections F2
0.2 Semi-rigid connections type E: Sc 1: IPB600 S c
Scini = 200 MNm rad; Mu = 450 kNm -20
1
2: IPB450
0.1 Semi-rigid connections type F: Beams: IPE400
Scini = 12.43 MNm rad; Mu = 101.7 kNm 8m
0.0 -40
0 10 20 30 40 50 60 70 80 90 100 110 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
Displacement u (cm) Time (s)
(e) Seven-story frame: inelastic static analysis (f) Seven-story frame: elastic dynamic analysis
Figure 4: CS-ASA structural applications.
4. FINAL REMARKS
A computational program developed for nonlinear static and dynamic analyses of steel framed
structures has just been presented. The numerical results obtained with this software demonstrate
behavior assessment of the steel structures were also verified. The graphical tool developed, the CS-
ASA Preprocessor, with its friendly interface, assisted the papers authors in the structural models
editing/visualization, and made the numerical study more efficient. It should be mentioned that a
graphical post-processor for the CS-ASA is now being developed and implemented.
Acknowledgments
The authors express their gratitude to CAPES, CNPq and FAPEMIG for the financial support
received in the development of this research. Special thanks go to Harriet Reis for her editorial
reviews.
REFERENCES
[1] AISC. Manual of Steel Construction, Load and Resistance Factor Design Specification for
Structural Steel Buildings. American Institute of Steel Construction, AISC, 1st edn., Chicago,
IL, 2003.
[2] R.V. Alves, Instabilidade no-linear elstica de estruturas reticuladas espaciais. Rio de
[3] G.M. Barsan and C.G. Chiorean, Computer program for large deflection elastoplastic analysis
[4] BS5950: Structural Use of Steelwork in Buildings. Part 1. British Standards Institution,
[5] S.L. Chan and P.P.T. Chui, Nonlinear Static and Cyclic Analysis of Steel Frames with Semi-
1991.
[7] W.F. Chen and S. Toma, Advanced Analysis of Steel Frames. CRC Press, Boca Raton,
Florida, 1994.
[8] J. Fish and T. Belytschko, A first course in finite elements, Wiley, 2007.
[9] J.Y.R. Liew, Advanced Analysis for Frame Design, Purdue University, West Lafayette, 1992.
(Ph.D. Dissertation).
[10] A.E. Maleck, Second-order inelastic and modified elastic analysis and design evaluation of
planar steel frames, Atlanta, GA: School of Civil and Environmental Engineering, Georgia
[11] E. Madenci and I. Guven, The Finite Element Method in Engineering Using Ansys, Springer,
2005.
[12] M. Metcalf, J. Reid, and M. Cohen, Fortran 95/2003 Explained. Oxford University Press,
[13] C. Pacoste and A. Eriksson, Beam elements in instability problems. Computer Methods in
[14] I.M. Prado, CS-ASA Preprocessor: programa grfico interativo de pr-processamento para
[15] J.L.M. Rangel Netto, R.F.G. Cerqueira, and W. Celes Filho, Introduo a Estrutura de
[16] SAP 2000 Advanced Integrated Software for Structural Analysis & Design, Version 10.0.1,
[17] K.H. Schweizerhof and P. Wriggers, Consistent linearization for path following methods in
nonlinear FE analysis. Computer Methods in Applied Mechanics and Engineering, 59: p. 269-
279, 1986.
[18] M. Sekulovic and M. Nefovska-Danilovic, Static inelastic analysis of steel frames with
flexible connections. Journal of Theoretical and Applied Mechanics, 31(2): p. 101-134, 2004.
[20] M. Sekulovic and R. Salatic, Nonlinear analysis of frames with flexible connections.
[21] A.R.D. Silva, Sistema computacional para a anlise avanada esttica e dinmica de
[22] A.R.D. Silva, W.L. Fernandes, R.A.M. Silveira, and P.B. Gonalves, Elastoplastic analysis of
plane frames under dynamics loading. In: Oate E. and Owen D.R.J., editors. XI International
[23] H.T. Thai and S.E. Kim, Pratical advanced analysis software for nonlinear inelastic analysis
[24] M.B. Wong, Plastic Analysis and Design of Steel Structures. Elsevier, 2009.
[25] Y.B. Yang and S.B. Kuo, Theory & Analysis of Nonlinear Framed Structures. Prentice Hall,
1994.
[26] R.D. Ziemian, Advanced methods of inelastic analysis in the limit states design of steel
structures, Ithaca, NY: School of Civil and Environmental Engineering, Cornell University,