[Link].

500
STEEL STRUCTURES
Autumn Semester 2024/Lecture 3
Kristo Mela
Associate Professor
Tampere University
[Link]@[Link]
+358 40 849 0563

©Kristo Mela (2024) 1

 STRUCTURAL ANALYSIS

©Kristo Mela (2024) 2

 ANALYSIS OF STEEL STRUCTURES
• Structural analysis is the process of determining the internal Structural analysis of a bem
(linear elasticity)
force, deformations and other outputs of a structure which
are needed in design calculations. q
• In simple cases, structural analysis can be carried out analytically,
for example, by beam theory.
• In practice, numerical methods, such as the finite element method
(FEM), are employed. L

• Typically, structural analysis is based on theory of elasticity. Ultimate limit state:

• material model follows Hooke’s law (σ = Eε), although the 1 1
displacements can be large. A: 𝑀𝐸𝑑 = 𝑞 𝑈𝐿 2 B: 𝑉𝐸𝑑 = 𝑞 𝑈𝐿
8 2
• all methods of analysis are allowed, as long as they are applicable
for the design situation at hand.
Serviceability limit state:
5 𝑞 𝑆𝐿 4
A: 𝑣𝐸𝑑 =
Question: why can the theory of elasticity be applied and why it is applied? 384 𝐸𝐼

©Kristo Mela (2024) 3

 STRUCTURAL ANALYSIS
MATERIAL MODELS FOR STEEL
• EN 1993-1-1 allows structural analysis by theory ACCOUNTING FOR YIELDING

of plasticity under certain conditions.

• Idea is to utilize redistribution of force that occurs
when yielding occurs.
• Analysis according to plasticity can mean the limit
load analysis (based on plastic mechanisms) or
nonlinear analysis which allows gradual yielding of
structures.
• This course focuses on structural analysis by theory of
elasticity.
• Additional information about applying plasticity in
structural analysis, see EN 1993-1-1, clause 5.4
EN 1993-1-5, Figure C.2

©Kristo Mela (2024) 4

 STRUCTURAL ANALYSIS BY ELASTICITY
• First-order analysis
1) Forces and displacements (inputs and outputs)
depend on each other linearly.
2) Assumption: displacements, rotations, and strains are
small.
3) Equilibrium equations are written for the initial
geometry (undeformed structure).
- Familiar from the basic courses on structural mechanics
and strength of materials.
• Second-order analysis
FIRST-ORDER SECOND-ORDER
1) Forces and displacements depend on each other
nonlinearly. Bending moment
2) Strains and rotations are small, displacements can be
large. (geometric nonlinearity)
 = v( L)
3) Equilibrium equations are written for the deformed
structure. Differential equation of deflection
• In second-order analysis, the effect of displacements on (”shallow” displacement fields = small rotations)
internal forces is taken into account. EIv( x) = − M ( x)

©Kristo Mela (2024) 5

 SECOND-ORDER EFFECTS
• Second-order effects mean the effect of deformations of
the structure on internal forces and other outputs.
• In particular, second-order effects on frames concern bending
moments and horizontal displacements due to axial forces.
• First-order analysis does not ”detect” these effects, because it is
based on the initial geometry of the structure.

• Second-order effects can be divided into effects on the

complete structure, so-called P-Δ effects, and local effects
along the axis of a member, so-called P-δ-effects.
• P-δ effects are included in the design equations of buckling,
Figure: Simões da Silva et al. (2012)
but they can also be ”activated” in structural analysis by
appropriate modelling.

©Kristo Mela (2024) 6

 FIRST-ORDER ANALYSIS
LOADS BENDING MOMENT HORIZONTAL DISPLACEMENT
qz = 20 kN/m
Fx = 10 kN

39.3 mm 39.1 mm

Load factor 22.5 kNm

𝛼 = 1.0

-15 -15

Fx = 30 kN qz = 60 kN/m
117.9 mm 117.3 mm

67.5 kNm
𝛼 = 3.0

-45.02 -44.98

©Kristo Mela (2024) 7

 FIRST-ORDER ANALYSIS
v
DISPLACEMENT

• In linear (= first-order) analysis, the

displacements and internal forces change Loading
proportionally to the loads. 𝛼 = 1.0 factor 𝛼 = 3.0
• Loads can be increased abritrarily large, and the
structural responses increase in same proportion.
• Real structures do not behave like this: the range
of applicability of the linear structural model is
limited.
• Calculating displacements and internal forces in

Load factor 
load combinations is simple:
• Responses (displacements, bending moments, axial
forces, etc.) of each load case is multiplied by the
corresponding load factor and added together.

Column displacement v
©Kristo Mela (2024) 8
 FIRST-ORDER ANALYSIS
BENDING MOMENT
Example:
Loads acting on a frame:
i) Permanent load G, and ii) variable load Q.

Bending moments at column base due to these loads are

MG and MQ.

In ultimate limit state (ULS), the factored loads are:

1.15G and 1.5Q.

Bendind moment [kNm]

Bending moment in column base in the ULS load
combination is
MEd =1.15MG + 1.5MQ

Column displacement v
©Kristo Mela (2024) 9
 SECOND-ORDER ANALYSIS
LOADS BENDING MOMENT DISPLACEMENT
qz = 20 kN/m 43.4 mm 43.2 mm
Fx = 10 kN

22.6 kNm

𝛼 = 1.0

-16.3 -16.3

Fx = 30 kN qz = 60 kN/m 165.0 mm 164.4 mm

𝛼 = 3.0 68.4 kNm

-60.0 -60.0
©Kristo Mela (2024) 10
 SECOND-ORDER ANALYSIS
• For α = 1, second-order analysis yields COLUMN BASE MOMENT AND
about 10% greater displacement and HORIZONTAL DISPLACEMENT
about 9% greater bending moment than Quantity α 1st 2nd 2nd Order/1st
first-order analysis (in this example!). Order Order Order
Displacement [mm] 39.3 43.4 1.10
• For α = 3, second-order analysis yields 1.0
Bending moment 15.0 16.3 1.09
about 40% greater displacement and [kNm]
about 33% greater bending moment. Displacement [mm] 117.9 165.0 1.40
3.0
• When loads are tripled (α = 3), bending Bending moment 45.0 60.0 1.33
[kNm]
moment increases 3.67-fold and
α … load factor
displacement increases by the factor 3.8.
• Bending moment of the beam increases
with the same ratio as for first-order
analysis.

©Kristo Mela (2024) 11

 SECOND-ORDER ANALYSIS

• Second-order analysis provides larger 2nd order analysis

Load factor 
internal forces and deformations than
first-order analysis in general. v

• Results of first-order analysis can be on

the unsafe side (too small!).
• In some cases, results of first-order
Column displacement v

analysis are sufficiently accurate, i.e.

the difference to second-order results

Load factor 
2nd order analysis
is ”small”.
• Question: when does the first-order
analysis suffice?

©Kristo Mela (2024) 12

Column displacement v
 BRACED FRAME Fx = 10 kN
qz = 20 kN/m

• Horizontal displacements of braced frames

are small in general.
• Second-order analysis provides nearly
identical displacements to first-order analysis. braces

• Same applies for bending moments.

1st order analysis: 2nd order analysis:
displacements displacements
• Horizontal stiffness of the frame clearly affects
0.3mm 0.4mm 0.3mm 0.4mm
the magnitude of the second-order effects.
• Question: how to quantify lateral stiffness and
what kind of criterion should be imposed on the
acceptability of first-order analysis?

©Kristo Mela (2024) 13

 GLOB AL STABILITY OF A FRAME
Generalized eigenvalue problem:
• Magnitude of second-order effects can be
(linear stability, finite element formulation)
estimated by examining the buckling modes
(eigenmodes) of the frame. Ku = αKGu
• Requires solution of the generalized eigenvalue
problem of linear stability theory. K … stiffness matrix
• In basic courses, similar problem for single KG … geometric stiffness matrix
members in compression has been solved α … load factor (eigenvalue)
(buckling according to Euler’s theory). u … buckling mode (eigenvector)

• The smallest load factor (eigenvalue), αcr, for Buckling mode 1 Buckling mode 2
which the frame buckles acts as an indicator
for the magnitude of second-order effects.
• Design criterion for the applicability of first-order
theory can be expressed in terms of this value
(αcr). α1 = 10.46 α2 = 126.22

©Kristo Mela (2024) 14

 APPLIC ABILITY OF FIRST-ORDER ANALYSIS –
Eurocode
EN 1993-1-1, 5.2.1(3):

First order elastic analysis may be used

αcr load factor, for which elastic instability of the structure in a global mode is
attained
Fcr Elastic critical buckling load for global instability mode based on initial elastic
stiffness
FEd design loading on the structures

©Kristo Mela (2024) 15

 GLOB AL FRAME STABILITY
• Load factor αcr is obtained by linear stability analysis.
• similar to Euler buckling load for individual members in
compression.
• ”Global mode” means that the buckling mode is a sway
mode.
• Smallest load factor does not necessarily correspond to αcr non-sway mode
but the shape of the corresponding buckling mode is relevant.
• Sway mode: relative (horizontal) displacement of ends
of a buckled member.
• Non-sway mode: ends of buckled members do not
move horizontally relative to each other. sway mode

Figure: ESDEP 16
©Kristo Mela (2024)
 SECOND-ORDER EFFECTS
EN 1993-1-1, 5.2.1(4)B:

• Portal frames with shallow roof slopes (slope not steeper

than 1:2) and beam-and-column type plane frames in
buildings may be checked for sway mode failure with first
order analysis if αcr > 10 for each storey.
• If the axial compression in the beams or rafters is not
significant, the following approximation can be used**:

HEd … is the design value of the horizontal reaction at the bottom of the storey to * Axial compression is not significant, if
the horizontal loads and fictitious horizontal loads; NEd < 0,09Ncr,
VEd … is the total design vertical load on the structure on the bottom of the storey; where Ncr is the buckling load of the
δΗ,Εd … is the horizontal displacement at the top of the storey, relative to the bottom member when it is considered as
of the storey, when the frame is loaded with horizontal loads (e.g. wind) and hinged at its ends.
fictitious horizontal loads which are applied at each floor level ; ** This is quite a rough estimate of αcr.
h … is the storey height. In practice, αcr should be obtained by
appropriate software.
©Kristo Mela (2024) 17
 SECOND-ORDER EFFECTS
EN 1993-1-1, 5.2.2(5)B:

• If 3 ≤ αcr < 10, for single storey frames, second order

sway effects due to vertical loads may be calculated by
increasing the horizontal loads and other possible sway
effects according to first order theory by the factor

• For multi-storey frames, second order sway effects may

be calculated by means of the above method, if the
structure has sufficient regularity (see EN 1993-1-1,
5.2.2(6)B)
• If αcr < 3, a more accurate second order analysis is
applied.
• Such structures are very flexible horizontally, and the deflection
limits of the serviceability limit state is often exceeded.
©Kristo Mela (2024) 18
 STRUCTURAL ANALYSIS – SUMMARY
1. Once the structural model is complete, first-order analysis is performed.
• Order of magnitude for responses is obtained and behaviour of structural model is checked.
2. Carry out linear stability (buckling) analysis.
• Determine αcr for global (sway) modes.
3. If αcr > 10, second-order effects need not be considered.
4. If αcr < 10, second-order analysis must be performed.
• This is rather straightforward with modern software, but it is important to understand the basics
of nonlinear structural analysis.
• In simple cases, second-order effects can be approximated by ”hand formulae”.

NOTE: cr depends on the magnitude of loads. Therefore, it needs to be determined separately for each load
combination.

©Kristo Mela (2024) 19

 STRUCTURAL ANALYSIS – SUMMARY
• Usually, structural analysis is based on elasticity (see. beam theory, finite element method).
• Design value of resistance can still be based on plasticity.
• Slenderness makes steel structures susceptible to second-order effects which are reflected on
internal forces and deformations especially for sway frames.
• Second-order (elastic) analysis is geometrically nonlinear analysis, with small rotations, but large
displacements.
• First-order (linear) analysis should always be performed as well as linear buckling analysis.
• Simple hand calculations of structural mechanics can be employed to check the internal
forces and deformations obtained from numerical analysis.
• Sway of a frame can be affected by
• Making joints rigid
• Increasing column profiles
• Adding bracing to the frame
©Kristo Mela (2024) 20
 IMPERFECTIONS

©Kristo Mela (2024) 21

 IMPERFECTIONS: GENERAL
• Real structures deviate from their idealized counterparts in SWAY IMPERFECTION
terms of geometry and material properties.
• Geometric imperfections:
• Members are always a bit tilted after installation.
• Members are not perfectly straight.
• Material imperfections:
• Residual stresses due to manufacturing remain in members.
RESIDUAL STRESSES
• Imperfections affect both the structural responses and
design of members.
• In calculations, equivalent geometrical imperfections are used,
that reflect the effects of all types of imperfections.
(EN 1993-1-1, 5.3)

©Kristo Mela (2024) 22

 IMPERFECTIONS vs. TOLERANCES
ℎ
∆≤ ±
a) b) 300
• Standard EN 1090-2:2018 presents ∆≤ ±
𝐿
1000
requirements for tolerances related to the
execution (fabrication and installing) of steel
structures, including:
• Member bow, (out-of-plumb) (a)
• Initial member sway (out-of-straightness), (b)
• Eccentricities (c)
c)
• These effects are different than the equivalent ∆ ≤ ±5mm e ≤ 5 mm
geometric imperfections.
• Unintentional deviations from tolerances
should be considered in design.
• Especially the effect of eccentricities exceeding the
tolerances should be considered.
EN 1090-2

©Kristo Mela (2024) 23

 IMPERFECTIONS: GENERAL
Global imperfection
a) Global imperfections:
• Initial sway of vertical structures.
• Considered in particular for sway frames.
• In general, it may be good to consider global
imperfections in all cases.
• Global imperfections may be neglected if horizontal
forces are sufficiently large.
Local imperfection
b) Local imperfections:
• Member bow.
• Includes the effect of out-of-plumb and residual stresses.
• Can often be neglected in structural analysis; local
imperfections are included in buckling design rules.

©Kristo Mela (2024) 24

 IMPERFECTIONS: GENERAL
• Imperfections can be included in the
Global imperfection Local imperfection
structural model by
a) modifying the geometry of the structure
b) adding equivalent forces to the model
• Idea of equivalent force: it causes the same
maximum bending moment to the member
as the axial force NEd when the axial force is
applied on the imperfect member.
• Use of equivalent forces requires knowing
the axial force NEd of a member
• NEd are obtained from first-order analysis.
• Magnitude of equivalent force depends on the
load combination. MEd = ϕLNEd MEd = NEde0,d

©Kristo Mela (2024) 25

 GLOBAL IMPERFECTIONS
(EN 1993-1-1: 5.3.2)

• Global (sway) imperfections are

considered in all relevant directions,
but only in one direction at a time.
• Equivalent forces are placed so that
their effect on the frame is the most
onerous.
• For example, the sway of main frame
columns corresponds to the wind
direction.
Wind load
• In multi-storey frames, equivalent
horizontal forces are added to each
storey.
©Kristo Mela (2024) 26
 GLOBAL IMPERFECTIONS
(EN 1993-1-1: 5.3.2)

Initial sway:

m … is the number of columns in a row including

Because αm ≤ 1 and αh ≤ 1, a conservative
only those columns which carry a vertical
value of initial sway is
1 load NEd not less than 50% of the average
𝜙 = 200 = 0,5% (= 0.29)
value of the columns in the vertical plane
considered.

©Kristo Mela (2024) 27

 GLOBAL IMPERFECTIONS
(EN 1993-1-1: 5.3.2)

(4)B For building frames sway imperfections may be disregarded when:

HEd ≥ 0,15 VEd
where
HEd … total design horizontal load, including equivalent forces transferred by the
storey (storey shear)
VEd … is the total design vertical load on the structure on the bottom of the storey.

• Equivalent loads due to sway imperfections aim to activate additional internal force due to initial sway.
• Determining HEd requires calculating the equivalent loads: in practice the above condition can be checked without
equivalent loads. If it is satisfied without equivalent loads, then equivalent loads are not needed.

©Kristo Mela (2024) 28

 EXAMPLE: Equivalent load
q1
q2
• Total design vertical load
R = q1L+q2L = 300 kN
• Axial forces of columns:
NEd,1 = 93.43 kN h 1 2 3
NEd,2 = 164.02 kN
NEd,3 = 42.55 kN
• Average axial force of columns Nave = 100 kN. L L
Because NEd,3 < 0.5Nave, the factor m is
determined based on columns 1 and 2 (m = 2). q1 = 20kN/m, q2 = 10kN/m, L = 10 m, h = 8m

©Kristo Mela (2024) 29

 EXAMPLE: Equivalent load
q1
q2
HEd,1

HEd,2 HEd,3

h 1 2 3

L L
so
q1 = 20kN/m, q2 = 10kN/m, L = 10 m, h = 8m

HEd,1 = ϕNEd,1= 0.0031∙93.43 = 0.2861kN These forces are added at the top of each
HEd,2 = ϕNEd,2= 0.0031∙164.02 = 0.5022kN column, acting in the same direction.
HEd,3 = ϕNEd,3= 0.0031∙42.55 = 0.1303kN
©Kristo Mela (2024) 30
 LOCAL IMPERFECTIONS
(EN 1993-1-1: 5.3.2)
E Q U I VA L E N T F O R C E
• Local imperfection of a member in form of an initial bow
affects especially flexural buckling design.
• Can be utilized also in design for lateral-torsional buckling.
• Initial bow imperfection includes out-of-plumb and residual e0
stresses.
• Bow imperfection is greater than the straightness tolerance of EN
1090-2 (L/1000).
• Magnitude of bow imperfection depends on buckling curve e0
and analysis type. e0
refers to
6.2 theory used
for cross-section
verification
e0

EN 1993-1-1,Table 5.1 Uniform load

is added on the member
©Kristo Mela (2024) 31
 EXAMPLE: Local imperfection
q1 = 20kN/m
q2 = 10kN/m
Member 1:
qd = 0.4 kN/m
NEd = 93.43 kN, L = h = 8m
HEA profile, buckling curve ’b’ (EN 1993-1-1, Table 6.2)
so e0 = L/250
1 2 3 h = 8m

8 N Ed L 8 N Ed 8 93.43 kN
qd = 2
 = = 
L 250 250 L 250 8m
L =10m L
kN kN
= 0.374  0.4
m m
EN 1993-1-1, Table 5.1

©Kristo Mela (2024) 32

 LOCAL IMPERFECTIONS
(EN 1993-1-1: 5.3.2)

• Initial bow imperfection and residual stresses are

included in the buckling design rules.
• In many cases, local imperfections need not be considered in
structural analysis. e0

• For sway frames (αcr < 10), local imperfections are

included in structural analysis, when determining the end
forces and end moments for member stability checks,
for each member for which the following conditions e0
hold: e0

1. Member is in compression;
2. Member as at least one moment resistant joint and
3. The following condition is satisfied:
NEd ≥ 0,25Ncr e0
where Ncr is the elastic critical buckling force for the member
considered as hinged at its ends.
• For non-sway frames, local member imperfections may
be omitted from the structural model.

©Kristo Mela (2024) 33

 STRUCTURAL ANALYSIS AND
IMPERFECTIONS Load factor 

• Imperfections are coupled with structural analysis

and member design.
• In particular the stability of frames and members is
connected with imperfections.

• Second order analysis can capture instabilities

• Moments and displacements become excessively large, if
the structure is unstable.
• Alternatively, the numerical calculations may diverge, i.e.
that equilibrium cannot be attained for the structure in Column displacement v
analysis.

• Buckling design rules of the standard EN 1993-1-1

include member bow imperfections and the
corresponding second-order effects.
• If these rules are applied, local imperfections need to be
considered in structural analysis.

©Kristo Mela (2024) 34

 STRUCTURAL ANALYSIS AND IMPERFECTIONS
EN 1993-1-1 provides three approaches for structural analysis and stability checks (5.2.2):
1. Perform 2nd order analysis and include global and local imperfections in the structural model.
• For design, it suffices to do cross-section checks and out-of-plane stability checks (lateral-torsional buckling, out-of-plane
flexural buckling).
• Idea: 2nd order analysis captures instabilities through large moments and displacements, making unstable structures
infeasible.

2. Perform 2nd order analysis and include global imperfections in the structural model.
• Member design: i) resistance of cross-sections; ii) stability check. In buckling verifications, the system length (length of the
member) can be used as buckling length.
• Idea: 2nd order analysis captures frame instabilities related to sway modes. Then, for the non-sway buckling modes, using
member length as buckling length is a conservative assumption.

3. Perform 1st order analysis without imperfections and employ appropriate buckling lengths for member
stability checks.
• Member design: i) resistance of cross-sections; ii) in-plane and out-of-plane stability. Buckling lengths must be determined
separately.
• Idea: appropriate buckling length with buckling design rules of EN 1993-1-1 account for second-order effects.
• Unclear: do second-order effects have to be considered in verification of cross-sections?
©Kristo Mela (2024) 35
 STRUCTURAL ANALYSIS AND IMPERFECTIONS
COMMENTS:

• Approach 2 is relatively straightforward to apply on modern software: the program performs second-order
analysis, and buckling design is done by the rules of EN 1993-1-1.
• Approach 1 takes more effort in structural modelling, but is tempting for the implied simplicity in design.
• Approach 3 is the simplest in terms of structural analysis, but determining buckling lengths is not always
straightforward.
• The rules of EN 1993-1-1 have clearly been written with beam-and-column frames in mind. For truss
structures, the three approaches require careful thought from the designer.
• Roof truss resembles a beam, but the compressed chord and braces can buckle.

• For beam-and-column frames, the second-order effects concern mostly columns.

• In the second-generation EN 1993-1-1 (published 2022, see SFS-Online) the various combinations of
analysis and imperfections has been clarified.
• In approach 3, second-order effects are not considered at all, but additional guidance can be given in the National Annex of
each country.

©Kristo Mela (2024) 36

 I M P E R F E C T I O N S – S U M M A RY
• Geometrical and material imperfections are taken into account in structural analysis: initial sway, initial bow, residual
stresses.
• These imperfections are not the same thing as execution tolerances of EN 1090-2..
• Certain imperfections (e.g. deviations of cross-sectional dimensions and material properties from their nominal values) are
included in partial safety factors γM, so they are not accounted for in structural analysis.
• Imperfections are typically treated as equivalent geometrical imperfections.
• These can be modelled as equivalent force systems.
• The equivalent forces must be determined separately for each load combination, unless a single conservative approximation is
used for all load combinations.
• Global imperfections (initial sway) concern sway frames in particular.
• They can be included also in the analysis of non-sway frames.
• Initial sway is consiered in all relevant directions, but in only one direction at a time. (EN 1993-1-1, 5.3.2(8)).
• Local imperfections (initial bow) can often be omitted from structural analysis.
• Initial bow is included in buckling rules, as are residual stresses.

©Kristo Mela (2024) 37

 EXAMPLE: PORTAL FRAME
gd = Permanent load
Add imperfections as equivalent loads. sd = Snow
Load values (kN/m) are factored design values wd = Wind

gd, sd

wd,1 wd,2 h

©Kristo Mela (2024) 38

 EXAMPLE: PORTAL FRAME
Global imperfections gd, sd

• Initial sway need not be considered if ϕNEd,1 ϕNEd,2

i. Frame is non-sway (αcr > 10), or
ii. HEd ≥ 0,15VEd.
wd,1
• Initial sway is considered by adding a h
wd,2
horizontal force ϕNEd at the top end
of each column.
i. Factor ϕ can be calculated accurately, or
a conservative approximation of ϕ = L
0.005 can be used.
Vertical loads: VEd = g d L + sd L
ii. For each column, its own axial force is
used. Horizontal loads: H Ed = wd ,1h + wd ,2 h

©Kristo Mela (2024) 39

 EXAMPLE: PORTAL FRAME
Local imperfections gd, sd

• Truss: local imperfections can be omitted

qd,1 qd,2
• Braces are pinned
• Compressed (top) chord is pin-jointed to the
columns
• Also: truss members are not buckling sway mode.
wd,1
h
• Columns: Column base joint is rigid, so wd,2
local imperfections must be included in
strucutural analysis, if
i. Frame is a sway frame and NEd > 0,25Ncr, where
Ncr is the elastic critical buckling load such that
the column is considered pinned at its ends.. L
- In this case an equivalent load qd is added as uniformly
distributed load acting in the same direction as the 1  2 EI
N Ed  ( g d + sd ) L N cr =
wind. 2 h2

©Kristo Mela (2024) 40

 EXAMPLE: PORTAL FRAME
gd, sd

• Equivalent loads due to global and local ϕNEd,1 ϕNEd,2

imperfections are incorporated in the actual
structural analysis.
• But only when the corresponding wd,1
imperfections need to be considered. qd,1 qd,2 h
wd,2

©Kristo Mela (2024) 41

 EXAMPLE: PORTAL FRAME
Numerical example of column bow imperfection rule:
Column section: HEA 260
I y = 10454 104 mm 4
I z = 3667 104 mm 4
Column length: h = 8000 mm
Buckling load with respect to major principal axis: Buckling load with respect to minor principal axis:
kN kN
 EI y
2  2  210 2
10454 104 mm 4
 EI z
2  2  210 2
 3667 104 mm 4
N cr , y = = mm = 3385 kN N cr , z = = mm = 1188 kN
2
h (8000mm) 2 h2 (8000mm) 2

0.25 N cr , y = 0.25  3385 kN = 846 kN 0.25 N cr , z = 0.25 1188 kN = 297 kN

Axial force of the column in typical frames Axial force of the column in typical frames
is often clearly smaller than this. can be of this magnitude.

©Kristo Mela (2024) 42

 FOR SELF-STUDY
• Details of structural analysis and imperfections:
• Simões da Silva et al. (2012): Chapter 2.3
• EN 1993-1-1, Clauses 5.1-5.2 and 5.3 (bonus: 5.4)
• Bonus: see EN 1993-1-1:2022, Chapter 7
• Coursework assignment:
• Consider the imperfections and equivalent horizontal loads and think how to incorporate them in the structural
model.
• Perform 1st order analysis for different load cases and combinations. Use the results to ensure that your structural
model is appropriate.
• Perform linear buckling analysis on the main frame. Explore how changing the column profile affects the results

Robot:

©Kristo Mela (2024) 43