Академический Документы
Профессиональный Документы
Культура Документы
Downloaded from ascelibrary.org by Imperial College London on 08/09/15. Copyright ASCE. For personal use only; all rights reserved.
Abstract
There have been a number of research efforts worldwide to quantify the nature of
abnormal loading and determine means by which its effects may be mitigated
(Grierson 2003). A state-of-the-practice paper (Dusenberry 2002) summarizes
some significant technical design codes and research papers on the subject of
progressive collapse under abnormal loading. A recent paper (Marjanishvili 2004)
Downloaded from ascelibrary.org by Imperial College London on 08/09/15. Copyright ASCE. For personal use only; all rights reserved.
The general steps of the linear-static method for progressive collapse analysis are
as follows:
3. Having the analysis results, identify the occurrence of any plasticity and reduce
member stiffnesses accordingly. Code specified moment-axial force and moment-
shear force interaction criteria govern plastic behavior, e.g.(CISC 1993). Member
stiffness reduction may be achieved using commercial design-office software
through direct modification of input data files, in a variety of different ways, as
follows:
(a) Modify material properties (E, G, etc.) and/or member properties (A, I, etc.) so
as to reduce member stiffness in a virtual manner. A drawback to this approach is
that the member needs to be further discretized beforehand so that properties can
be modified only in the regions where plastic behavior occurs, which may result
in a large increase in the number of elements for the structure.
(b) Release each member degree-of-freedom (DOF) associated with plastic
behavior by modifying the corresponding member end-fixity factor to be ‘free’.
This is illustrated in Figure 1 for both semi-rigid and rigid behavior, where the
DOF is not released in the former case until both the strength capacity Fp and
deformation capacity fail have been reached. While this approach does not
increase the dimensions of the analysis, it only crudely approximates plastic
behavior and may lead to erroneous results (Grierson & Safi 2005).
(c) Introduce a spring element for each DOF at locations of the member where
plastic behavior may occur. Initially assign each spring stiffness to be either the
same as that of the member (semi-rigid connection) or of infinite value (rigid
connection). After the member reaches its first-yield capacity Fy, progressively
modify the spring stiffness until both the strength capacity Fp and deformation
capacity fail have been reached, at which point the DOF is released. This
procedure is illustrated in Figure 2 for the initial spring stiffness being either
semi-rigid or rigid. This approach approximates plastic behavior quite well and
gives results that reasonably agree with those found by more rigorous analysis
(Grierson & Safi). The number of DOF for the structure is not increased by this
stiffness reduction method when using design-office software that has embedded
4. Having the reduced member stiffnesses, determine if any existing members are
sufficiently damaged as to be eliminated from the structure. A member is
eliminated from the structure if its buckles, or its plastic deformation capacity has
been reached a both end-sections. Upon eliminating a member, its released
reactions are applied as point loads on end-section nodes while its selfweight and
loading are applied as debris loads on members below. If no members are
Downloaded from ascelibrary.org by Imperial College London on 08/09/15. Copyright ASCE. For personal use only; all rights reserved.
eliminated, reanalyze the structure accounting for the reduced member stiffnesses
and go back to step 3. Otherwise, go to step 5.
5. Reanalyze the structure accounting for both the reduced member stiffnesses
and any eliminated members to check overall stability. If it is found possible to
complete the analysis because the structure is yet stable, go back to step 3. If it is
not found possible to complete the analysis because the structure is unstable as
indicated by its stiffness matrix being singular, the analysis process is terminated
(i.e., essentially, progressive collapse has been triggered).
F F
Fp Fp
F F
Analysis1 Analysis1
F1 F1
Analysis n Analysis n
Fn Fn
Analysis N Analysis N
Fp Fp
Ke-n Ke1 Ke-n
Ke1
Ke-N Ke-N
Fy DOF Release Fy DOF Release
Ki Ki
y p 1 n fail y p 1 n fail
Plastic deformation may occur at the end sections of each of the six column
members, and at the end and mid span sections of each of the four beam members.
Linear-static progressive collapse analysis of the framework is conducted using
the DOF-Release stiffness reduction method (Figure 1), and the results are
compared with those found using a rigorous progressive collapse analysis method,
(Grierson et al 2005).
As shown in Figure 4(a), it is initially assumed that column 6-9 fails due to the
action of some type of abnormal load (e.g., explosion, design error, etc.) and is
eliminated from the original structure. After column 6-9 fails, the remaining
structure experiences inelastic flexural behavior. The linear-static PCA determines
that five plastic hinges sequentially form up to a loading level that is only 22.68%
of the gravity service loads in Figure 3, at which beam 8-9 becomes unstable, as
shown in Figure 4(a). The rigorous PCA (Grierson et al 2005) determines the very
similar results shown in Figure 5(a), the only real difference being in the sequence
of plastic hinges formed in column 4-7.
After the fifth plastic hinge forms, beam section 8 reaches its flexural deformation
capacity and beam 8-9 is eliminated from the structure, and its gravity dead
weight w89 is applied as uniformly distributed debris loading with impact factor
= 1.0 on the lower floor beam 5-6, as indicated in Figure 4(b). Thereafter, the
linear PCA determines that four further plastic hinges sequentially form up to a
loading level that is 78.81% of the gravity service loads, at which point the
structure stiffness matrix became singular as the right bay of the frame fails as a
plastic collapse mechanism, as shown in Fig. 4(b). The rigorous PCA (Grierson et
al 2005) determines the nearly similar results shown in Figure 5(b), the
differences being in the sequence of plastic hinges formed and the fact that
progressive collapse is triggered at the higher load level of 89.63% of the gravity
service loads.
Conclusions
7 8 51.1 N/mm 9
W200×19
W360×16
W 530×66 W
360×1
4572
5 109.5 N/mm
Downloaded from ascelibrary.org by Imperial College London on 08/09/15. Copyright ASCE. For personal use only; all rights reserved.
4 W360×19 6
W
W200×2
W360×1
W
6096
6
x
1 2 3
6096 mm 14630 mm
8
7 1 5 9
3 4 7 8
2
4 4 7
5 6 8
5 6
9
Sequence of plastic 6
i = 0.7881
= 0.2268 hinge formation
1 3 1 2 3
2
Ground level
8
7 1 5 9 7
2 8
4
3
4 4 6
5 7
6 5 6
9
This study was supported by the Natural Science & Engineering Research
Council of Canada, and bythe Power & Water Institute of Technology, I.R. Iran.
References
American Society of Civil Engineers (ASCE), (2002), Minimum Design Loads for
Downloaded from ascelibrary.org by Imperial College London on 08/09/15. Copyright ASCE. For personal use only; all rights reserved.