Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Scribd
Upload
178 views

Practical Application of The "General Method" of EN 1993-1-1

In the second article of this series, Dr József Szalai of ConSteel Solutions demonstrates practical examples where the “General Method” of EN 1993-1-1 shows advantages compared to the conventional approaches.

Uploaded by

VivaCharles
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
178 views

Practical Application of The "General Method" of EN 1993-1-1

In the second article of this series, Dr József Szalai of ConSteel Solutions demonstrates practical examples where the “General Method” of EN 1993-1-1 shows advantages compared to the conventional approaches.

Uploaded by

VivaCharles
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

Technical

Practical application of
the General Method of
EN 1993-1-1
In the second article of this series, Dr Jzsef Szalai of ConSteel Solutions demonstrates
practical examples where the General Method of EN 1993-1-1 shows advantages compared
to the conventional approaches.

The key component: general elastic stability analysis


When verifying the stability of a structural component the critical issue
is always the correct assessment of the possible forms of buckling. In
conventional design procedures the buckling forms are taken into account
by calculating the elastic critical resistances (Ncr and Mcr) for the associated
buckling shapes. Thus the determination of the appropriate elastic critical
force is the most significant step of the design method (Step 3 in the previous
article) but at the same time has the potential to be the most complicated
step. It is very important to recognise that the calculation of an elastic critical
force always implies an assessment of the buckling shape the resistance and
the buckling shape are linked.
Designers are used to the concept of using an effective length (to
determine a buckling length) though this can be a significant simplification
of the problem, particularly if the system of applied forces and the restraint
system is complex, because the buckling mode in these circumstances is not
easily identified or assessed. In the general method the elastic critical load
amplifiers are calculated considering the complex system of forces based on
more realistic compound buckling shapes. Thus the general method provides
solutions for a number of design irregularities irregular members such as
tapered or haunched members, irregular buckling shapes due to special (for
example eccentric) supports or restraints on the member by the calculation
of the compound buckling shape of a realistic structural model.
The general method requires that designers consider the buckling
behaviour, which in turn has the advantage of placing the designer more
directly in control of the frame behaviour, which may well lead to more
appropriate solutions.
Although elastic critical forces calculated using the buckling length are
suited for simple hand calculation, the determination of the elastic critical
load amplifiers required for the general method are usually much more
complicated and the use of some appropriate numerical method or software is
necessary. The efficient use of the general method requires the use of software
in which the all the necessary compound buckling modes lateral, torsional,
lateral-torsional buckling, buckling about an eccentric restraint axis etc. can
be calculated properly. Currently, such software products are quite rare since
the general analysis of buckling which includes torsional modes requires
particular analysis tools. One solution is to use shell finite elements when
creating the global structural model. This can make modelling the building
very complicated and the analysis results can be difficult to handle, meaning
this option is practically never used by engineers.
Another solution is the use of special beam finite elements which are
able to accommodate torsional buckling modes. There are a number of
published finite elements fulfilling these requirements, including a 7 degreeof-freedom finite element which includes the effect of torsion and warping
of the cross section. This element is implemented into the structural design
software ConSteel which has been used in this article to calculate the design
examples.

30

NSC
May 11

Design examples
Two simple examples are presented to examine the consequences of the two
main simplifications of the conventional stability design method: the isolation
of the structural member from the surrounding structure and the separation
of the flexural and lateral-torsional buckling modes.
Example 1: Influence of buckling mode separation
The first example, shown in Figure 1, is a simple column fixed at the bottom
and pinned at the top subjected to compression and bending. The column has
two intermediate (eccentric) supports to one of the flanges.

Figure 1. Geometry, loading and internal forces of Example 1


The primary problem with this simple column is that the eccentric
intermediate supports generate an irregular buckling situation where
neither of the well-known pure buckling forms (pure lateral or pure torsional
buckling for compression or pure lateral-torsional buckling for bending)
can be separated. EN 1993-1-1 provides some simplified design formulas
for beams supported on the compression flange (Section 6.3.2.4) but only
for those subjected to pure bending. For pure compression there are no
applicable rules for columns with intermediate restraints to only one flange.
Moreover these formulas cannot be applied in any procedure that includes
interaction of the pure buckling forms. If the conventional method is to
be applied considering the pure buckling forms, the problem should be
somehow modified in order to be able to determine the pure elastic critical
forces. Two usual simplifications are considered: (1) to assume concentric
intermediate lateral supports and (2) to assume concentric intermediate
lateral and torsional supports. In Table 1 (over page) the actual configuration is
illustrated by the elastic critical forces and associated buckling shapes for

Advisory desk
Technical

supports

actual eccentric support conditions

loading

N+M

elastic
critical
forces

cr= 2.05
Ncr= 1230 kN
Mcr= 307.5 kNm

pure N

pure M

Ncr= 2052 kN
Mcr = 712.5 kNm

buckling
shapes

Table 1

supports

concentric intermediate lateral restraints


pure N

loading
elastic
critical
forces

Ncr,x =
3192 kN

Ncr,z =
6306 kN

concentric intermediate lateral and torsional restraints

pure M
Mcr =
1647 kNm

pure N
Ncr,x =
9732 kN

pure M
Ncr,z =
6306 kN

Mcr=
2246 kNm

buckling
shapes

Table 2

Figure 2: Geometry and internal forces of Example 2

the combined loading used in the general method and for the pure cases
(N; compression, M; bending). It can be seen that even the pure cases do not
belong to pure lateral and lateral-torsional buckling shapes so the appropriate
pure mode separation and accordingly the determination of the buckling
lengths are impossible.
In Table 2 the two modified support situations are illustrated. If concentric
lateral supports are assumed, a pure lateral buckling resistance can be
determined, but this is clearly a different buckling behaviour than under the
actual restraints. The intermediate lateral supports do not prevent the twist
of the member, with different buckling lengths for flexural and torsional
buckling, meaning conventional calculation of the elastic critical bending
moment is impossible. Also, the relevant mode (torsional) is not taken into
account in the interaction formula.
Example 2: Influence of structural member isolation
In this example the frame shown in Figure 2 is examined as a part of a
complete structure. The frame is haunched and has pinned bases, braced
at the corners and in the middle of the rafters and subjected to 6.75 kN/m
distributed vertical load on the rafters. The stability design of the right column
is presented. In Figure 3 the buckling of the frame is illustrated due to the
compression and bending moments in the members and the results of the
general method are shown calculated at the top section of the column.
It can be seen that according to the general method the column is slightly
inadequate (105.6%). The column resistance may be recalculated by the
conventional interaction method defined in EN 1993-1-1 6.3.3 using the Method
1 (Annex A) for the calculation of the interaction factors. The buckling lengths
for both the lateral and lateral-torsional buckling are taken as the system
length, assuming that this is a correct estimation. The final utilization
37

32

NSC
May 11

Global Stability resistance (dominant)


Utilization

105.6%

Applied part of standard

6.3.4 (2)-(3), (4)b - (6.63, 6.64, 6.66) formula

ult,k

1.522

0.750

cr,op

1.450

LT

0.624

op

1.025

NEd

-79.5 kN

0.340

My,Ed

-249.3 kNm

1.165

Mz,Ed

0.0 kNm

0.582

NRk

2 322.3 kN

LT

0.490

My,Rk

400.3 kNm

LT

1.047

Mz,Rk

61.9 kNm

0.400

M1

1.0

Figure 3: Buckling shape and the results of the general method for the column

Technical/Advisory Desk

Practical application of the General Method of EN 1993-1-1


continued from p32
value was 82.1%, meaning that according to the conventional method the
column is adequate. The reason for the big difference between the two
utilization values is that the buckling shape of the frame involves both the
column and the rafter, so these two members should form a consistent unit
in stability design. The separation of the column and the determination of the
buckling length independently of the rafter produce an overestimation for the
column buckling resistance. This can be understood clearly if it is appreciated
that as one unit, the column and the beam together reach the elastic critical
state at a lower level of load than the column alone.

Conclusions
This article presented some examples of the application of the general
method. It was demonstrated that if a more realistic modelling and structural
analysis is possible i.e. a general stability analysis then a more realistic and
natural way for the stability design is to use the general method. The examples
also showed the importance of an accurate assessment of the buckling shapes
and the associated elastic critical values which can lead to safer and in other
cases more economic structural design.

AD 358
Design thickness of cold formed
members and sheeting
According to BS EN 10143:2006, for nominal thicknesses between 1.6 mm
and 2.0 mm, for a width of strip between 100 mm and 1500 mm, the normal
tolerance is +/ 0.15 mm and the special tolerance is +/- 0.09 mm.
The core thickness is thus:
tcor = tnom tmetallic coatings = 1.80 0.04 = 1.76 mm

The purpose of this Advisory Desk Note is to highlight the existence of, and
explain the implications of, an important but often over-looked definition
in BS EN 1993-1-3:2006 the design thickness of cold formed members and
sheeting. It is crucial that designers use the correct design thickness for cold
formed members and sheeting because small differences between design and
nominal thicknesses can lead to significant differences in section properties
and design resistances.
The definition concerned is in clause 3.2.4 (3) of BS EN 1993-1-3. The
wording in the clause is slightly confused but it does provide two expressions
for the design thickness, depending on the specified tolerance on thickness.
If the negative tolerance on the material thickness, tol, (expressed as a
percentage of the nominal thickness) is less than or equal to 5%, expression
(3.3a) gives the design thickness as:

t = tcor
(3.3a)
where the core thickness, tcor = tnom tmetallic coatings , in which tnom is the nominal
material thickness and tmetallic coatings should be taken as the total coating
thickness on both faces.
However, if the negative tolerance, tol, is greater than 5% then the design
thickness is given by expression (3.3b) as:

t = tcor (100 tol) / 95
(3.3b)
(Note that the above expression corrects a typographic error in the published
Standard.)
The tolerances on dimensions and shape for continuously hot-dip coated
steel sheet and strip are given in BS EN 10143: 2006. There are two types of
thickness tolerance given in BS EN 10143 - normal and special. The actual
tolerance value is dependent on several factors such as the steel grade,
nominal material thickness and the width of the strip.
Where normal tolerances apply, the negative normal tolerance values
are generally greater than 5% of the nominal material thickness, meaning
that expression 3.3b should be used. Where special tolerances apply,
although the negative tolerance values are also greater than 5%, according
to BS EN 1993-1-3, 3.2.4 (4) the design thickness may be taken as the core
thickness, tcor , irrespective of the magnitude of that tolerance.
The designer therefore needs to know, in addition to the nominal thickness
and the coating thickness, whether normal or special tolerances will be
specified for coated steel to BS EN 10143. To illustrate the difference between
the two alternatives, consider the following example.

Specification of different positive and negative tolerances


The approach described above to determine design thickness is based on
the assumption that the positive and negative tolerances for the material
thickness are the same, which is true when either normal or special
tolerances in accordance with in BS EN 10143: 2006 apply. However, it is
common practice to specify material with different positive and negative
tolerances. Where this is done, the value of tnom should be taken as the midpoint of the extreme values, rather than the specified nominal value. The
negative tolerance tol should then be taken as the difference between the
modified nominal thickness and the extreme minimum thickness.
Consider the above example but with a positive tolerance of 0.0 mm and
a negative tolerance of 0.1 mm. The minimum and maximum thicknesses to
this specification are thus 1.70 mm and 1.80 mm, respectively, giving tnom =
1.75 mm. The core thickness is thus given by:
tcor = tnom tmetallic coatings = 1.75 0.04 = 1.71 mm
The negative tolerance from the mid-range of nominal thickness is 0.05 mm,
which is equivalent to 2.9% of the nominal thickness (i.e. tol = 2.9). Therefore,
the design thickness is given by expression (3.3a) as :
t = tcor = 1.71 mm

Example
Consider a cold formed member manufactured from S350 strip steel with a
nominal thickness of 1.8 mm and with a Z275 coating (for which the metallic
coating thickness is 0.04 mm).

Contact: Andrew Way


Tel:
01344 636525
Email:
advisory@steel-sci.com

Design thickness for normal tolerance


The normal tolerance of 0.15 mm is equivalent to 8.3% of the nominal
thickness (i.e. tol = 8.3). Therefore, the design thickness is given by expression
(3.3b) as:
t = tcor (100 tol) = 1.76 (100 8.3) / 95 = 1.70 mm
Design thickness for special tolerance
It might be noted that the negative tolerance is equivalent to 5.0% of the
nominal thickness (i.e. tol = 5.0) but, irrespective of its value, the design
thickness is given in this case by expression (3.3a) as:
t = tcor = 1.76 mm

NSC
May 11

37

You might also like