A Parametric Survey of The Influence of The Semi-Rigid Connections On The Shakedown of Elasto-Plastic Frames
A Parametric Survey of The Influence of The Semi-Rigid Connections On The Shakedown of Elasto-Plastic Frames
A Parametric Survey of The Influence of The Semi-Rigid Connections On The Shakedown of Elasto-Plastic Frames
139–147 (2006)
Abstract
In the past it was generally assumed that the connection of the beams and columns of the steel-framed
multistorey structures are either rigid or pinned. In the reality, however, they are semi-rigid. This
circumstance influences significantly the behaviour of the entire structure, therefore, it has to be taken
into consideration in the analysis and design. In this paper a parametric study is presented to analyse
the influence of the semi-rigid connections on the shakedown of elasto-plastic steel framed structures
under multi-parameter static loading. To control the plastic behaviour of the structure bound on
the complementary strain energy of the residual forces is also applied. The semi-rigid behaviour is
modelled by appropriate internal springs at the beam column-connections. The formulation of the
problem yields to nonlinear mathematical programming which is solved by the use of an iterative
procedure. The parametric study is illustrated by the solution of an example.
Keywords: semi-rigid connection, shakedown analysis, mathematical programming, nonlinear pro-
gramming.
1. Introduction
The plastic analysis and design methods are widely used in structural engineering
practice since they provide information about the post yield behaviour and load
carrying capacity of structures and by utilizing the plastic reserve they generally
lead to significant saving in material. The "cost" of these advantages, however, is
that the plastic deformations accumulated during the loading history might exceed
the plastic deformation capacities of the structural elements, and the structure might
become unserviceable.
Hence, it is evident that the determination or at least the assessment of plastic
deformations and residual displacements and their consideration in plastic analysis
and design of structures, especially in case of variable and repeated loads, is an
important requirement.
The main structural elements of the steel-framed multi-storey structures are
the columns, the beams and their connections. The assumption that the stiffnesses
of the connection is either rigid or pinned has been widely applied in the past.
140 J. LÓGÓ et al.
However, the real behaviour of the connections is never as ideal as it was assumed,
but they behave somewhere in between these limits, such they are semi-rigid. This
circumstance which can influence significantly the behaviour of the structure has to
be taken into account in the analysis and design. To model these types of connections
there are different methods which are presented in national standards, as well. (e.g.
EUROCODE 3) [23].
A typical moment-rotation relationship of semi-rigid connections is illustrated
in Fig. 1.a. The diagram can be approximated by bilinear or linear relationships
shown in Fig. 1.b and Fig. 1.c, respectively. In the latter case r0 denotes the stiffness
and M0 is the design moment of the connection. According to the recommendations
and surveys based on linear approximation it is suggested to assume M0 to 2/3 of
the ultimate moment Mu of the connection [1, 5, 23]. In this paper the linear
approximation will be used. A more detailed study is under preparation in which
bilinear relationships will be applied.
M M M
Mu Mu
M0
M
r0
M M M
a; b; c;
The aim of this paper is to present a method for the shakedown analysis of
elasto-plastic frames with semi-rigid connections subjected to quasi-static multiple
loading and to conduct a parametric study of the influence of the semi-rigid con-
nections on the shakedown multiplier of the structure. In the present method the
plastic behaviour of the frame is controlled by
The effect of the semi-rigid connections are taken into consideration by an appro-
priately chosen spring coefficient in the elementary stiffness matrix of the beam
elements. The parametric study is illustrated by a numerical example where the
effect of the change of the spring coefficient is investigated.
A PARAMETRIC SURVEY... 141
2. Notations
In the paper the finite element method with i=1,2,…,n elements will be applied and
the following notations will be used:
The most important tool for controlling the plastic behaviour of structures is the ap-
plication of static and kinematic theorems of shakedown proposed by M ELAN [17]
and KOITER [12]. Using these theorems the accumulation of unrestricted plastic
deformations can be prevented, no information can be obtained, however, about the
magnitudes of the plastic deformations and residual displacements developed at the
shakedown of the structure. In this paper the static theorem of shakedown analysis
will be used [6, 13].
In the past decades the shakedown analysis has been supplemented by other methods
which provide further information in terms of bounds on the plastic deformations
and residual displacements. Bounding theorems were elaborated for the analysis
of elastoplastic bodies and structures under static loading (see e.g. KOITER [12],
M AIER [14, 15], P OLIZOTTO [18], P ONTER [20], T IN -L OI [21], K ALISZKY [7],
K ALISZKY and L ÓGÓ [8] C APURSO et. al. [3], D OROSZ [4]). Later, similarly to
the shakedown theorems, these theorems have been extended to topics concerning
dynamic loading, large deformations, strain hardening and nonassociated consti-
tutive laws and have been applied to the solution of several problems (see e.g.
C APURSO [2, 3], M AIER and V ITELLO [16], P OLIZOTTO [19], L ANGE -H ANSEN
[14], T IN -L OI [21, 22]).
On the basis of the upper bound theorems developed by C APURSO [2], C A -
PURSO , C ORRADI and M AIER [3], it was suggested by K ALISZKY and L ÓGÓ
[9, 10, 11] that the complementary strain energy of the residual forces (Wp ) could
be considered an overall measure of the plastic performance of structures. Using
this idea the plastic deformations can be controlled by the following constraint:
n
1X r
Q Fi Qri − Wp0 ≤ 0. (1)
2 i=1 i
Here Wp0 is the allowable residual complementary strain energy. In the lack of
other data it may be assumed, for example, that Wp0 = αWe , where We is the
complementary elastic strain energy calculated from the largest load acting on the
structure and α is an appropriately chosen multiplier (say 1 ≤ α ≤ 3).
The above equation becomes much simpler if the residual moments (MiR ) are
used to compute the residual complementary strain energy, the residual normal and
shear forces are taken into consideration at the calculation of the residual moments
only and the torsion moments of the beams are neglected:
n Zℓi
1 X1
Wp = M iR (s) 2 ds. (2)
2E i=1 Ii
0
A PARAMETRIC SURVEY... 143
If the beam and column elements are loaded only at the nodes the residual bending
moment function MiR (s) is linear along the elements and omitting the details Eq.
(2) can be written as follows:
n
1 X ℓi h R 2 R 2
i
R R
(3)
Wp = Mi1 + Mi1 Mi2 + Mi2 .
6E i=1 Ii
Here Mi1
R
and Mi2
R
are the residual bending moments at the ends of the i-th element.
The problem to be solved is to determine the maximum load multiplier mmax under
the conditions that the frame is strong enough to carry the loads mmax Ph , shakes
down, and satisfies the constraint on plastic deformations given by Eq. (3). Hence,
maximize m (4)
subject to
G∗ Mr = 0; (5)
−2S0i σyí ≤ mMieh + Mir ≤ 2S0i σyí , (h = 1, 2, ..., s), (i = 1, 2, ..., n); (7)
n
1 X ℓi h R 2 R 2
i
R R
− Wp0 ≤ 0; (9)
Mi1 + Mi1 Mi2 + Mi2
6E i=1 Ii
Here m and Mir are unknown. The flexibility F (r) and stiffness K (r) matrices of
the frame depends on the spring coefficient values of the beam-column connections.
5. Example
In the following the influence on the shakedown parameter (m) of the allow-
able residual strain energy (Wp0 ) and the stiffness of the semi-rigid connections
characterized by the spring coefficient (r) will be used.
All the necessary data of the problems are given in Fig. 2. The external
loads are F1 =1kN and F2 =4kN. The yield stress and the Young’s Modulus are σy
=21kN/cm2 and E = 2.07 · 106 kN/cm2 . The cross-sectional data of the beam are:
AB =28.5cm2 , IB = 1943cm4 , SB0 = 130.0cm3 , while for the columns
Ac = 39.0cm2 , Ic = 3891.6cm4 , Sc0 = 210.0cm3 . At nodes 2 and 4 the beam-
column connections can be fixed, semi-rigid with variable internal spring values
r and pinned. If the connection is semi-rigid, the behaviour is represented by a
spring which value is varied from 10 kNcm/rad (≈pinned connection) to 1.00E+21
kNcm/rad (≈ fixed connection). At that time the 2/3 reduction rule was not applied
at the semi-rigid connections.
Considering the intermediate case when W0p2 = 2.00 kNm the difference between
the shakedown parameters belonging to the pinned and rigid connections and related
to the latter one is ∼20%.
25,00
23,00
shakedown parameter
21,00
19,00
17,00
15,00
13,00
0,00 0,50 1,00 1,50 2,00 2,50 3,00 3,50 4,00
allowable residual strain energy
Fig. 3. Variation of the shakedown parameter in terms of the allowable residual energy
The figures show that depending on the magnitude of Wp0 the variation of
the stiffness of the semi-rigid connections influences significantly the shakedown
parameter. Between the two limit cases (i.e. the pinned connection and the rigid
one) the variation of the shakedown parameter might reach ∼35%. The above results
show the importance of the effect of the semi-rigid connections on the shakedown
of structures, but from this special example, general statements cannot be made.
6. Conclusion
In the paper a parametric study is presented to analyse the influence of the semi-
rigid connections on the shakedown of elasto-plastic steel-framed structures under
multi-parameter static loading. In addition to the shakedown analysis, to control
the plastic behaviour of the structure, bound on the complementary strain energy
of the residual forces is also applied. The semi-rigid behaviour is modelled by
appropriate internal springs at the beam-column connections. The formulation of
the problem yields to nonlinear mathematical programming which is solved by the
use of a sequential quadratic algorithm. The applied numerical method is relatively
simple and can be conducted easily.
146 J. LÓGÓ et al.
25,0000
23,0000
shakedown parameter
21,0000
19,0000
17,0000
15,0000
13,0000
1,00E+01 1,00E+03 1,00E+05 1,00E+07 1,00E+09 1,00E+11 1,00E+13 1,00E+15 1,00E+17 1,00E+19 1,00E+21
spring values
Acknowledgements
The present study was supported by The Hungarian National Scientific and Research Foun-
dation (OTKA) (grant T042993) and the Ministry of Education (TET grant F-44/05).
A PARAMETRIC SURVEY... 147
References