5.

METHODS OF ANALYSIS
5.1 First-Order Elastic Analysis

In first-order elastic analysis a linear

relationship between the applied loading F
and the deformations () is assumed. The
internal force distribution in the frame is
assumed to be unaffected by the
displacements in the frame.

Frame analysis can therefore be conducted

according to linear elastic principles The
frame responds according to line 1 in
Figure 4.

5.2 First-Order Rigid-Plastic Analysis

Rigid-plastic analysis (or the application of

simple plastic theory) neglects the effects of
elastic deflections and assumes that all
structural deformation takes place in discrete
regions, called plastic hinges, where plasticity
has developed. When using first-order, rigid-
plastic theory only the collapse condition is
addressed. This condition occurs when
sufficient plastic hinges are assumed to have
formed to convert the structure into a
mechanism. Thus the path by which this stage
is reached, i.e. the sequence of formation of
the hinges and any intermediate distributions
of internal forces, are not considered. Figure 5
illustrates the concept for 3 cases of vertical,
horizontal and combined loading, whilst curve
2 in Figure 4 gives the frame response
according to this approach. Due to the form of
the analysis, no information is provided on the
magnitude of the deflections. The analysis
gives only that all stiffness is lost at the collapse load and deflections therefore (in
theory) become uncontrolled.

5.3 Elastic Critical Load

Using the methods described in Lectures 6, it is possible to calculate the buckling

loads for frames under suitably idealised loading. Depending upon the content and
complexity of the frame, several different buckling modes, each with its associated
elastic vertical load, may be possible. For the simple portal frame shown in Figure 6,
both a symmetrical and an antisymmetrical mode are possible; the latter will usually
be avoided with a much lower critical load. Once again the analysis provides no
information on the magnitude of the deflections; it simply identifies a particular load
level. The curve 3 in Figure 4 gives the representation of the critical load obtained by
an elastic buckling analysis.

5.4 Second-Order Elastic Analysis

In second-order elastic analysis the effect of

elastic deformations on the internal force
distribution is taken into account. The result is
a transition from the linear analysis line 1 at
low loads to the elastic critical line 3 at large
deflections. For frames the second-order
effects may be separated into 2 parts:

 reduction in the effective bending

stiffnesses of individual members due to
compressive loading.
 a destabilising effect due to the
overturning moment produced by the vertical
loads acting through the horizontal deflections
caused by the lateral loads.

5.5 Second-Order Rigid-Plastic Analysis

If the deformations that may develop as a result of the formation of the plastic
collapse mechanism are allowed for when formulating the equilibrium of the frame,
then the result is the developing mechanism curve of line 5 in Figure 4. This curve
shows that equilibrium can only be maintained with a reduction in the level of the
applied loads.
5.6 First-Order, Elastic-Plastic Theory

If a linear elastic analysis is modified to allow for reductions in frame stiffness with
the progressive formation of plastic hinges at increasing levels of the applied load,
then the response curve of line 6 is obtained. This line exhibits progressive loss of
stiffness as each plastic hinge is formed and eventually merges with the rigid-plastic
line 2.

5.7 Second-Order, Elastic-Plastic Analysis

When the analysis that traces the formation of plastic hinges also allows for the effects
of deformations in setting up the governing equations, then line 6 is modified
somewhat into line 7. Line 7 initially follows the first-order elastic line 1 but diverges
from this line to follow the second-order elastic line 4 as destabilising effects become
more significant. Formation of the first plastic hinge - which occurs at a slightly lower
applied load than is the case with the first-order, elasto-plastic analysis due to the
larger deformation associated with second-order analysis - further reduces the
stiffness, causing line 7 to diverge from line 4. This divergence becomes more
pronounced as more plastic hinges form. The peak of this curve corresponds to the
failure load predicted by this type of analysis. At large deformations line 7 tends to
merge with the curve for the mechanism, line 5.

5.8 Second-Order, Plastic Zone Analysis

If the spread of plasticity both through the cross-section and along the member length
is taken into account, instead of assuming that it is concentrated into the desirable
regions of the plastic hinges, then the resulting type of analysis is usually termed
plastic zone theory. It provides an even closer representation of actual behaviour and
leads to a curve similar to line 7.

6. COMMENTS
In principle, any of the above approaches to frame analysis may be adopted. In
practice, some of the effects may be found to be of little real significance for certain
classes of structure, e.g. for many low-rise frames second-order effects are very small
and may reasonably be neglected. Certain cases may also arise where particular forms
of response should be avoided, e.g. for buildings containing heavy cranes which will
cause repeated loading, elastic design is normally employed.

The more complex approaches will almost certainly require the use of suitable
computer software to implement the volume of calculation. It is therefore important to
select an approach which is compatible with both the accuracy required and the level
of importance of the project under consideration.

When calculating deflections at working load levels for the purpose of checking
serviceability, it is usual to employ only linear elastic analysis.

7. FRAME CLASSIFICATION
In order to provide guidance on the most appropriate type of analysis to use in
particular cases, Eurocode 3 has introduced the idea of frame classification [1]. A
double condition is used: (i) braced or unbraced (ii) non-sway or sway.

7.1 Braced Frames

 A frame may be classified as braced if its sway resistance is supplied by a

bracing system with a response to in-plane horizontal loads which is
sufficiently stiff for it to be acceptably accurate to assume that all horizontal
loads are resisted by the bracing system.

This may be further quantified as:

 A steel frame may be classified as braced if the bracing system reduces its
horizontal displacements by at least 80%.

For such frames first-order elastic or plastic theory should be used.

When the above conditions are not satisfied the frame must be considered as
unbraced.

When designing the bracing system:

 The effects of the initial sway imperfections, given in Clause 5.2.4.3 of

Eurocode 3 [1] and discussed later in these notes, in the braced frame shall be
taken into account in the design of the bracing system.
 The initial sway imperfections plus any horizontal loads applied to a braced
frame, may be treated as affecting only the bracing system.
 The bracing system should be designed to resist:

- any horizontal loads applied to the frames which it braces.

- any horizontal or vertical loads applied directly to the bracing system.

- the effects of the initial sway imperfections (or the equivalent horizontal forces)
from the bracing system itself and from all the frames which it braces.

 Where the bracing system is a frame or sub-frame, it may itself be either sway
or non-sway.

7.2 Unbraced Frames

An unbraced frame may be classified as a non-sway frame according to Clause 5.2.5.2

of Eurocode 3 [1] providing:

 Its response to in-plane horizontal forces is sufficiently stiff for it to be

acceptably accurate to neglect any additional internal forces or moments arising
from horizontal displacements of its nodes.

This classification may be further quantified as:

 A frame may be classified as non-sway for a given load case if the elastic
critical load ratio VSd /Vcr for that load case satisfies the criterion:

VSd /Vcr  0,1

where VSd is the design value of the total vertical load.

Vcr is its elastic critical value for failure in a sway mode.

 Beam-and-column type plane frames in building structures with beams

connecting each column at each storey level may be classified as non-sway for
a given load case if the following criterion is satisfied. When first-order theory
is used, the horizontal displacements in each storey due to the design loads
(both horizontal and vertical), plus the initial sway imperfection applied in the
form of equivalent horizontal forces, should satisfy the criterion:

where  is the horizontal displacement at the top of the storey, relative to the bottom
of the storey.

h is the storey height.

H is the total horizontal reaction at the bottom of the storey.

V is the total vertical reaction at the bottom of the storey.

Both requirements follow from the idea that, if satisfied, the load-carrying resistance
determined by neglecting sway effects will be only a ten per cent less than that
calculated by including such effects. This approach is, in turn, based upon the
Merchant-Rankine concept for estimating the true ultimate load of a frame that fails
by some from of inelastic instability from a knowledge of its elastic critical load and
its first-order, rigid-plastic collapse load. Both loads are relatively straightforward to
calculate.

The original Merchant-Rankine formulae for the failure load Vsd is:

where:

Vcr is the elastic critical load.

Vpl is the first-order, rigid-plastic collapse load.

From this it is clear that when Vcr >> Vpl, then Vsd ~ Vpl.

Non-sway frames should be designed using first-order elastic or plastic theory to resist
safely the arrangements of loads that lead to the most severe combinations of internal
forces and moments in the individual members and connections. The effects of
restraint to columns in improving their stability should be taken into account by using
the concept of effective buckling length as explained in Lecture 7.7.

Frames that do not meet the above requirements must be designed as sway frames.

7.3 Sway Frames

Sway frames shall be analysed under those arrangements of the variable loads which
are critical for failure in a sway mode. In addition, sway frames shall also be analysed
for the non-sway mode.

The initial sway imperfections, and member imperfections where necessary, shall be
included in the global analysis of all frames.

The allowance for imperfections in the analysis of sway frames is intended to cover
effects such as lack of verticality, lack of straightness, residual stresses, etc. It is
expressed in Eurocode 3 by means of a set of equivalent geometrical imperfections
[1]. These imperfections are not actual construction tolerances but, because they are
intended to represent the effect of a number of factors, are likely to be larger than such
tolerances. The form specified in Eurocode 3 is:

 The effects of imperfections shall be allowed for in frame analysis by means of

an equivalent geometric imperfection in the form of an initial sway
imperfection  determined from:

 = kc ks o

with o = 1/200

kc = [0,5 + 1/nc]0,5 but kc  1,0

and ks = [0,2 + 1/ns]0,5 but ks  1,0

where nc is the number of columns per plane.

ns is the number of storeys.

 Columns which carry a vertical load NSd of less than 50% of the mean value of
the vertical load per column in the plane considered, shall not be included in n c.
 Columns which do not extend through all the storeys included in n s shall not be
included in nc. Those floor levels and roof levels which are not connected to all
the columns included in ncshall not be included when determining ns.
 These initial sway imperfections apply in all horizontal directions, but need
only be considered in one direction at a time.
 The possible torsional effects on the structure of anti-symmetric sways, on two
opposite faces, shall also be considered.
 If more convenient, the initial sway imperfection may be replaced by a closed
system of equivalent horizontal forces, see Figure 7.
 In beam-and-column building frames, these equivalent horizontal forces should
be applied at each floor and roof level and should be proportionate to the
vertical loads applied to the structure at that level, see Figure 8.
 The horizontal reactions at each support should be determined using the initial
sway imperfection and not the equivalent horizontal forces. In the absence of
actual horizontal loads, the net horizontal reaction is zero.
First-order or second-order analysis may be used. If the analysis is first-order, second-
order effects may be allowed for in an appropriate way when designing the columns
by using the results of a first-order analysis and either:

- using amplified sway moments, or

- using the sway-mode buckling lengths.

 When second-order elastic global analysis is used, in-plane buckling lengths for
the non-sway mode may be used for member design.
 In the amplified sway moments method, the sway moments found by a first-
order elastic analysis should be increased by multiplying them by the ratio:

where VSd is the design value of the total vertical load.

Vcr is its elastic critical value for failure in a sway mode.

 The amplified sway moments method should not be used when the elastic
critical load ratio VSd/Vcr is more than 0,25.
 Sway moments are those associated with the horizontal translation of the top of
a storey relative to the bottom of that storey. They arise from horizontal loading
and may also arise from vertical loading if either the structure or the loading is
asymmetrical.
 As an alternative to determining VSd/Vcr directly, the following approximation
may be used in beam-and-column type frames:
where , h, H and V are as defined previously.

 When the amplified sway moments method is used, in-plane buckling lengths
for the non-sway mode may be used for member design.
 When first-order elastic analysis with sway-mode in-plane buckling lengths is
used for column design, the sway moments in the beams and the beam-to-
column connections should be amplified by at least 1,2 unless a smaller value is
shown by analysis to be adequate.

Rules for the application of plastic analysis procedures to sway frames are given in
Clause 5.2.6.3 of Eurocode 3 [1].