Hafif Celik Tez

ISTANBUL TECHNICAL UNIVERSITY  GRADUATE SCHOOL OF SCIENCE
ENGINEERING AND TECHNOLOGY
STRENGTH AND STABILITY OF

THIN WALLED STEEL COLUMNS
IN STORAGE RACK STRUCTURES
M.Sc. THESIS
Atakan MANGIR
Department of Civil Engineering
Structural Engineering Programme
MAY 2014
ISTANBUL TECHNICAL UNIVERSITY  GRADUATE SCHOOL OF SCIENCE
ENGINEERING AND TECHNOLOGY
STRENGTH AND STABILITY OF

THIN WALLED STEEL COLUMNS
IN STORAGE RACK STRUCTURES
M.Sc. THESIS
Atakan MANGIR
(501111007)
Department of Civil Engineering
Structural Engineering Programme
Thesis Advisor: Prof. Dr. Ünal ALDEMİR

Anabilim Dalı : Herhangi Mühendislik, Bilim
Programı : Herhangi Program
MAY 2014
İSTANBUL TEKNİK ÜNİVERSİTESİ  FEN BİLİMLERİ ENSTİTÜSÜ
İNCE CİDARLI ÇELİK DEPO RAF SİSTEMLERİ KOLONLARININ

DAYANIM VE STABİLİTESİ
YÜKSEK LİSANS TEZİ
Atakan MANGIR
(501111007)
İnşaat Mühendisliği Anabilim Dalı
Yapı Mühendisliği Programı
Tez Danışmanı: Prof. Dr. Ünal ALDEMİR

Anabilim Dalı : Herhangi Mühendislik, Bilim
Programı : Herhangi Program
MAY 2014
Atakan MANGIR, a M.Sc. student of ITU Institute of Science Engineering and
Technology student ID 501111007 successfully defended the thesis/dissertation
entitled “ STRENGTH AND STABILITY OF THIN WALLED STEEL
COLUMNS IN STORAGE RACK STRUCTURES ”, which he prepared after
fulfilling the requirements specified in the associated legislations, before the jury
whose signatures are below.
Thesis Advisor : Prof. Dr. Ünal ALDEMİR ..............................

Istanbul Technical University
Co-advisor : Assoc. Prof. Dr. Güven KIYMAZ ..............................

Fatih University
Jury Members : Prof. Dr. Abdül HAYIR .............................

Prof. Dr. Bülent AKBAŞ ..............................

Gebze Institute of Technology
Assis. Prof. Dr. Cüneyt VATANSEVER ..............................

Date of Submission : 5 May 2014

Date of Defense : 29 May 2014
v
vi
In memory of my grandfather, to my family
vii
viii
FOREWORD
I would like to express my deepest gratitudes and thanks to my supervisors Prof. Dr.
Ünal ALDEMİR and Assoc. Prof. Dr. Güven KIYMAZ who gave me their
suggestions, supports and encouragements which made it possible for me to
accomplish this dissertation. This dissertation could not have been completed without
their guidance.
During the construction and completion of this thesis, my family has also a major
role. I believe that any appreciation word won’t be enough for their presence and
support. Their help to me to get over my medical inabilities during this study is
inestimable.
Specially, my sincere appreciations and gratitudes go to my collegues Ertuğrul Emre
ÇALIK and Nilgün Merve ÇAĞLAR for their supports during the construction of
this study.
I would like to thank Prof. Dr. Teoman PEKÖZ who is emeritus professor of Cornell
University and author of major portions of the specifications and recommendations
on cold-formed steel and rack structures for his precious suggestions.
May 2014 Atakan MANGIR

(Civil Engineer)
ix
x
TABLE OF CONTENTS
Page
FOREWORD ............................................................................................................. ix
TABLE OF CONTENTS.......................................................................................... xi
ABBREVIATIONS ................................................................................................. xiii
NOTATIONS ............................................................................................................ xv
LIST OF TABLES .................................................................................................. xix
LIST OF FIGURES ................................................................................................ xxi
SUMMARY ............................................................................................................ xxv
ÖZET..................................................................................................................... xxvii
1. INTRODUCTION .................................................................................................. 1
1.1 Overview of Cold Formed Steel ......................................................................... 1
1.1.1 Production ................................................................................................... 2
1.1.2 Structural applications................................................................................. 3
1.1.3 Advantages and disadvantages of cold formed steel ................................ 10
1.2 Aim and Layout of the Thesis .......................................................................... 11
2. LITERATURE REVIEW ON COLD FORMED STEEL COLUMNS .......... 13
2.1 General Remarks on Cold Formed Steel Column Behaviour .......................... 13
2.1.1 Yielding ..................................................................................................... 14
2.1.2 Global (overall) column buckling ............................................................. 14
2.1.2.1 Flexural column buckling .................................................................. 15
Elastic buckling .......................................................................................... 15
Inelastic buckling ....................................................................................... 15
2.1.2.2 Torsional buckling and flexural-torsional buckling ........................... 17
2.1.3 Local buckling........................................................................................... 20
2.1.4 Distortional buckling................................................................................. 26
2.2 Studies on Direct Strength Method .................................................................. 27
2.3 Studies on Perforated Cold Formed Steel Storage Rack Columns .................. 36
3. CODIFIED DESIGN OF COLD FORMED STEEL COLUMNS UNDER
PURE COMPRESSION ...................................................................................... 45
3.1 Review of American Iron and Steel Institute S100-2007 Standart .................. 45
3.1.1 Effective width method for concentric axially loaded compression
members .................................................................................................... 45
3.1.1.1 The nominal axial strength calculations ............................................. 46
3.1.1.2 Effective area calculations ................................................................. 48
3.1.1.3 Distortional buckling strength calculations ........................................ 52
3.1.2 Example solution of a c-channel column with a simple lip edge stiffener
via effective width method ........................................................................ 54
3.1.3 Direct strength method for concentric axially loaded compression
members .................................................................................................... 61
3.1.3.1 Determination of elastic buckling loads............................................. 63
3.1.3.2 Column analysis ................................................................................. 63
xi
Flexural, torsional, or flexural-torsional buckling ..................................... 63
Local buckling ............................................................................................ 64
Distortional buckling .................................................................................. 64
3.1.3.3. Summary and design curves of direct strength method .................... 65
3.1.4 Example solution of a c-channel column with a simple lip edge stiffener
via direct strength method ......................................................................... 67
4. INVESTIGATION OF RACK COLUMN BEHAVIOUR THROUGH
DIRECT STRENGTH METHOD ...................................................................... 75
4.1 Analysis of Unperforated Column Models....................................................... 77
4.1.1 Analytical study on unperforated members using direct strength
method ....................................................................................................... 79
4.1.2 Nominal strength - lip length relationships of unperforated column
models ....................................................................................................... 84
4.1.2.1 General comparison of results for unperforated column models ....... 88
4.2 Analysis of Perforated Specimens .................................................................... 92
4.2.1 Analytical study on perforated members using direct strength method .... 95
4.2.2. Nominal strength - lip length relationships of perforated column
models .................................................................................................... 102
4.2.2.1 General comparison of results for perforated column models ......... 106
4.3 Comparison Between the Strength of Unperforated Rack Column Models with
Perforated Rack Column Models ................................................................... 111
5. CONCLUSION ................................................................................................... 117
REFERENCES ....................................................................................................... 119
CURRICULUM VITAE ........................................................................................ 125
xii
ABBREVIATIONS
AISI : American Iron and Steel Institute

AS/NZS : The Australian/New Zealand Standard
ASD : Allowable Stress Design
CFS : Cold Formed Steel
CUFSM : Cornell University Finite Strip Method
D : Distortional Controlling Buckling Mode
DSM : Direct Strength Method
EC : Eurocode
EN : European Standart
EWM : Effective Width Method
FSM : Finite Strip Method
G : Global Controlling Buckling Mode
IS : I - Sections
LG : Local-global Controlling Buckling Mode
LRFD : Load and Resistance Factor Design
NAS : North American Standart
PCS : Plain Channel Sections
RHS : Rectangular Hollow Sections
RMI : Rack Manufacturers Institute
S.S. : Simply Supported
xiii
xiv
NOTATIONS
A : Full unreduced cross-sectional area

Ae : Effective area at the stress Fn, effective net area.
Ag : Gross area of element including stiffeners
b : Effective design width of compression element
b0 : Out-to-out flange width
b1 , b2 : Effective widths
Bnp : Length of nonperforated sheet between perforations in a row
perpendicular to column length
Bp : Total length of perforations in a row perpendicular to column length
Cw : Warping constant
D : Term determined by Et3/12(1 − μ2)
Overall depth of first lip
D2 : Overall depth of second lip
D3 : Overall depth of third lip
ds : Reduced effective width of stiffener
d’s : Effective width of stiffener
E : Modulus of elasticity
Er : Reduced modulus
Et : Tangent modulus
f : Stress in compression element computed on basis of effective design
width
fcr : Critical buckling stress for local buckling of plates and tubes
Fd : Elastic distortional buckling stress
Fe : The least of the applicable elastic flexural, torsional and flexural
torsional buckling stress
Fn : Nominal buckling stress
(Fn)e : Nominal elastic buckling stress
(Fn)I : Nominal inelastic buckling stress
fx : Compression stress in x direction
Fy : Yield stress of steel
G : Shear modulus
H : Total column width including complex lip lengths
h0 , B : Out-to-out web depth
I : Moment of innertia
Ia : Adequate moment of innertia of stiffener, so that each component
element will behave as a stiffened element
Is : Actual moment of innertia of full stiffener about its own centroidal
axis parallel to element to be stiffened
J : St. Venant torsion constant of cross section
k : Effective length factor
kd : Plate buckling coefficient for distortional buckling
xv
Kt : Effective length factor for torsion
L : Column length,
Pitch length,
Laterally unbraced length of member
Lcr : Critical unbraced length of distortional buckling
Lm : Distance between discrete restraints that restrict distortional buckling
Lnp : Length of nonperforated sheet between perforations in a row parallel
to column length
Lp : Total length of perforations in a row parallel to column length
Lt : Unbraced length of compression member for torsion
m : Number of half sine waves
n : Coefficient
: Applied axial load
: Euler buckling load
: The nominal axial strength
: Nominal axial strength for global(overall) buckling
: Critical elastic distortional column buckling load
: Critical elastic overall(global) column buckling load
: Critical elastic local column buckling load
: Nominal axial strength for distortional buckling
: Nominal axial strength for local buckling
: Reduced modulus load
: Yielding stress
: Nominal values of yielding stress
: Tangent modulus load
: Maximum axial load gathered from experimental test
: Ultimate axial load of experimental results
: Euler flexural buckling load about x axis
: Yield load
: Torsional buckling load about z axis
Q : Stress and/or area factor to modify allowable axial stress
r : Radius of gyration of full unreduced cross-section about axis of
buckling
r0 : Polar radius of gyration of cross section about shear center
RI : Is/Ia
rx , ry : Radii of gyration of cross-section about centrodial principal axes
S : √
t : Thickness of plate
Base steel thickness
tr : Reduced thickness
trD : Reduced thickness for distortional buckling
trG : Reduced thickness for global buckling
trL : Reduced thickness for local buckling
w : Flat width of element exclusive of radii
x0 : The distance from shear center to centroid along principal x-axis,
taken as negative
α : Coefficient accounts for the benefit of an unbraced length
β : 1 − (x0/r0)2
θ : Lip angle
xvi
: Length of half sine wave (Half wavelength)
: Column slenderness parameter. Slenderness factor for global
buckling
: Slenderness factor for distortional buckling
: Slenderness factor for local buckling
μ : Poisson’s ratio for steel, = 0.30
ρ : Local reduction factor
σe : Euler stress of elastic column buckling
σex : Px / A
σR : Critical buckling stress based on reduced modulus method
σT : Critical buckling stress based on tangent modulus method
σt : Pz / A
σTFO : Elastic flexural–torsional buckling stress
σTFT : Inelastic flexural–torsional buckling stress
ω : Deflection of plate perpendicular to surface
,w : Width of the plate, respectively
: Shear stress
xvii
xviii
LIST OF TABLES
Page
Table 3.1 : Determination of plate buckling coefficient, k ..................................... 51

Table 3.2 : Section properties of example solution ................................................. 55
Table 3.3 : Effective widths of section’s elements.................................................. 59
Table 4.1 : Geometric properties of unperforated specimens having 500 mm
column length ........................................................................................ 78
column length ........................................................................................ 78
column length ........................................................................................ 79
Table 4.4 : Critical elastic buckling loads obtained from CUFSM analysis for
unperforated specimens having 500 mm column length ....................... 80
unperforated specimens having 1000 mm column length ..................... 81
Table 4.7 : Nominal axial strengths of unperforated specimens having 500 mm
column length ........................................................................................ 83
column length ........................................................................................ 83
column length ........................................................................................ 84
Table 4.10 : Limits of the model of reduced thickness ............................................. 92
Table 4.11 : Geometric properties of perforated specimens having 500 mm
column length ........................................................................................ 94
column length ........................................................................................ 94
column length ........................................................................................ 95
perforated specimens having 500 mm column length ........................... 98
perforated specimens having 1000 mm column length ......................... 99
perforated specimens having 1500 mm column length ....................... 100
column length ...................................................................................... 101
column length ...................................................................................... 101
column length ...................................................................................... 102
xix
xx
LIST OF FIGURES
Page
Figure 1.1 : Cold formed steel member shapes. ........................................................ 1

Figure 1.2 : Cold roll-forming process (description). ............................................... 2
Figure 1.3 : Cold roll-forming process (real). ........................................................... 3
Figure 1.4 : Cold formed steel building example...................................................... 4
Figure 1.5 : Cold formed steel structure example. .................................................... 4
Figure 1.6 : A cold formed steel hangar.................................................................... 5
Figure 1.7 : Some CFS section geometries. .............................................................. 5
Figure 1.8 : Cold formed steel roof deck sheet example. ......................................... 6
Figure 1.9 : Cold formed steel roof deck example. ................................................... 6
Figure 1.10 : An example of cold formed steel wall panel. ........................................ 7
Figure 1.11 : A basic schematic sketch of a CFS storage rack structure. ................... 8
Figure 1.12 : Pallet racking example from Turkey ..................................................... 9
Figure 1.13 : A drive-in model of rack structures. ...................................................... 9
Figure 1.14 : Pallet racking structure example. ........................................................ 10
Figure 2.1 : Flexural column buckling stress. ......................................................... 17
Figure 2.2 : An exaggerated flexural-torsional column buckling example. ........... 18
Figure 2.3 : Flexural-torsional column buckling of an I section. ............................ 18
Figure 2.4 : Maximum stress for flexural-torsional buckling. ................................ 20
Figure 2.5 : Example of different local buckling modes......................................... 21
Figure 2.6 : Local buckling of stiffened compression flange of a hat-shaped
section. ................................................................................................. 22
Figure 2.7 : Square plate subjected to compression stress. .................................... 22
Figure 2.8 : Rectangular plate subjected to compression stress. ............................ 22
Figure 2.9 : Buckling coefficient for flat rectangular plates. ................................. 23
Figure 2.10 : Values of k for determining critical buckling stress. .......................... 24
Figure 2.11 : Distortional buckling of a C channel with stiffened lips. ................... 26
Figure 2.12 : Rack section column buckling stress versus half wavelength for
concentric compression. ...................................................................... 27
Figure 2.13 : Approach of finite element method and finite strip method............... 28
Figure 2.14 : An example of buckling curve. .......................................................... 28
Figure 2.15 : Local buckling mode of a Z section. .................................................. 29
Figure 2.16 : Comparison of test strengths with design strengths. .......................... 30
Figure 2.17 : Buckling stress versus half-wavelength for a C-Section in
compression. ........................................................................................ 31
Figure 2.18 : Comparison of the direct strength method predictor curves with test
data for columns. ................................................................................. 33
Figure 2.19 : Comparison design methods with test results. ................................... 34
Figure 2.20 : Comparison between the current DSM distortional curve and
ultimate loads with different end conditions........................................ 36
xxi
Figure 2.21 : Comparison between the modified DSM distortional curve and
column ultimate loads with different end conditions. .......................... 36
Figure 2.22 : Load versus displacement curves for fy=250 MPa. ............................ 38
Figure 2.23 : Load versus displacement curves for fy=320.23 MPa. ....................... 38
Figure 2.24 : Experimental collapsed specimens and deformed shapes from FE
analysis. ............................................................................................... 39
Figure 2.25 : Effective area and centre of gravity comparisons .............................. 40
Figure 2.26 : Main geometric parameters of the column ......................................... 42
Figure 3.1 : Stiffened elements. ............................................................................. 50
Figure 3.2 : Elements with simple lip edge stiffener. ............................................ 51
Figure 3.3 : Out-to-out dimensions of webs and stiffened elements...................... 53
Figure 3.4 : C-channel column section with simple lip edge example .................. 54
Figure 3.5 : Limits for pre-qualified columns ........................................................ 62
Figure 3.6 : DSM global buckling failure design curve and equations .................. 66
Figure 3.7 : DSM local buckling failure design curve and equations .................... 66
Figure 3.8 : DSM distortional buckling failure design curve and equations ......... 67
Figure 3.9 : Centerline dimensions in CUFSM ...................................................... 68
Figure 3.10 : Yield load calculation ......................................................................... 69
Figure 3.11 : Configuration of material properties .................................................. 69
Figure 3.12 : Boundary condition selection ............................................................. 70
Figure 3.13 : Result of local Buckling in CUFSM................................................... 71
Figure 3.14 : Result of distortional buckling in CUFSM ......................................... 71
Figure 3.15 : Result of flexural-torsional buckling in CUFSM ............................... 72
Figure 4.1 : General geometric details of specimen cross-sections ....................... 76
Figure 4.2 : Explanation of specimen’s notation through its properties ................ 77
Figure 4.3 : Nominal strength – lip length relationship of the unperforated
specimens having 500 mm column length (lip continues outward) .... 85
specimens having 500 mm column length (lip continues inward) ...... 85
specimens having 1000 mm column length (lip continues outward) .. 86
specimens having 1000 mm column length (lip continues inward) .... 86
specimens having 1500 mm column length (lip continues outward) .. 87
specimens having 1500 mm column length (lip continues inward) .... 87
Figure 4.9 : Comparison of nominal strength – lip length relationship of the
unperforated specimens having 1000 mm column length ................... 89
unperforated specimens having 1500 mm column length ................... 89
Figure 4.12 : Comparison between nominal strength – lip length realtionships
of the unperforated specimens having lips continuing outwards ........ 90
of the unperforated specimens having lips continuing inwards .......... 91
Figure 4.14 : Typical model for perforation configuration of all perforated
specimens............................................................................................. 93
Figure 4.15 : Typical model for reduced thickness approach from CUFSM ........... 97
xxii
Figure 4.16 : Nominal strength – lip length relationship of the perforated specimens
having 500 mm column length (lip continues outward).................... 103
having 500 mm column length (lip continues inward)...................... 103
having 1000 mm column length (lip continues outward).................. 104
having 1000 mm column length (lip continues inward).................... 104
having 1500 mm column length (lip continues outward).................. 105
having 1500 mm column length (lip continues inward).................... 105
perforated specimens having 500 mm column length ....................... 106
perforated specimens having 1000 mm column length ..................... 107
perforated specimens having 1500 mm column length ..................... 107
Figure 4.25 : Controlling buckling modes for inward and outward cases of
perforated 1500 mm long columns with maximum lip length
configuration ..................................................................................... 108
of the perforated specimens having lips continuing outwards .......... 109
of the perforated specimens having lips continuing inwards ............ 110
Figure 4.28 : Comparison between nominal strength – lip length relationships of the
perforated and unperforated specimens having lips continuing outward
and column length 500 mm ............................................................... 111
perforated and unperforated specimens having lips continuing inward
and column length 500 mm ............................................................... 112
and column length 1000 mm ............................................................. 112
xxiii
xxiv
STRENGTH AND STABILITY OF THIN WALLED STEEL COLUMNS IN
STORAGE RACK STRUCTURES
SUMMARY
Steel Storage Rack Systems play a vital role in logistics and storing goods.This has
particularly increased over the last decade. In our country, importance of these
systems is increasing through the demand of insufficient space to storage goods from
various industries and fields. The structural strength and stability of steel storage rack
structures became more important than before because the usage of these structures
and value of the goods carried by these structures have increased.
Safety of columns used as part of the structral system of storage rack frames is
particulary vital for safe use of the systems. Steel Storage Rack Columns can be
assumed to be a specific version of general cold formed steel (CFS) columns.
However, the behaviour of steel storage rack columns is different from general
behavior of the CFS columns due to their thickness, lip configuration and the
presence of the perforations. Steel storage rack columns may consist various lip
types, lengths and perforations for the intended use. These two major variables affect
the capacity of rack columns and hence the stability and strength of whole rack
structure.
Determining the structural capacity of steel storage rack columns has been currently
regulated with several standarts and methods. American Iron and Steel Institute
S100-2007 Standart is one of these few standarts to regulate the method of
calculation to determine the strength of cold formed steel columns. The standart
presents Effective Width Method and Direct Strength Method for calculations.
Direct Strength Method, presented in Appendix 1 part of AISI S100-2007 standart
has been newly developed method for determination of the strength of concentrically
axially loaded cold formed steel members. The effects of perforations and lip
configuration including lip type, lip forming direction and column length on nominal
column strength are investigated using Direct Strength Method on a range of various
specimens in this thesis.
The main parts of the present study can be summarized as follows;

- Review of previous studies conducted on steel storage rack columns and direct
strength method.
- Assessment of current practise (AISI S100-700 provisions) with respect to
calculation of cold formed steel column capacity with Effective Width Method and
Direct Strength method.
- The analytical study carried out on 120 column models having different column
lengths, lip directions, lip lengths and perforated and unperforated cases.
In AISI S100-700 standart, currently the effect of perforations on column strength is

accounted for by physical testing. No design rules are given to consider the effect of
perforations on the three major buckling modes including local, distortional and
global. In this study, a recently developed analytical approach which integrates the
xxv
Direct Strength Method by a method called “Reduced Thickness Approach” was
used. Comparisons were made between nominal strength values of columns with
varying cross-section geometries (lip configuration) for both unperforated and
perforated cases with three different column lengths.
xxvi
İNCE CİDARLI ÇELİK DEPO RAF SİSTEMLERİ KOLONLARININ
DAYANIM VE STABİLİTESİ
ÖZET
Çelik Depo Raf Sistemleri ürün depolama ve lojistik sektörlerinde hayati bir rol
oynamaktadır. Son on yıllık periyotta bu sistemlerin öneminin bilhassa arttığı
görülmüştür. Ülkemizde, çeşitli endüstriyel sektörler ve alanlardaki depolama
alanlarının yetersizliği ile bu sistemlerin önemi gittikçe artmaktadır. Depo raf
sistemlerinin kullanımının yaygınlaşması ve bu sistemlerde taşınan ürünlerin
değerlerinin artması ile bu sistemlerin yapısal dayanımları önem kazanmaktadır.
Çelik depo raf sistemleri, ince cidarlı çelik elemanlardan oluşmaktadır. İnce cidarlı
çelik elemanlar, çelik plakaların oda sıcaklığında işlenmesi ile elde edilirler. İşleme
metodu, çelik plakalara soğuk merdaneli biçimleme veya sıkıştırma yapılmasıdır.
İşlenen çelik plakaların kalınlıkları genellikle 0.10 mm ile 7.7 mm arasında
değişmektedir. Plakaların akma gerilmeleri ise 230 MPa ile 380 MPa arasında
değişmektedir. İnce cidarlı çelik elemanların, taşıyıcı sistemlerde kullanımının bazı
avantaj ve dezavantajları bulunmaktadır. Hafiflik, toplu üretim, yüksek mukavemet,
kolay kurulum, kalıp işçiliğinin gerekmemesi, paslanma korunumu, düşük taşıma
giderleri, kesit çeşitliliği ve geri kazanılabilirlik gibi özellikler, ince cidarlı çelik
elemanların kullanım avantajları arasında yer almaktadır. Eleman kalınlıklarının
düşük olması sebebiyle burkulma problemleri ve yangın dayanımları ise ince cidarlı
çelik elemanlar kullanımının dezavantajları arasında yer alabilmektedir.
İnce cidarlı elemanların taşıyıcı sistemi oluşturduğu iki tip depo raf sistemi yaygın
olarak kullanılmaktadır. Bunlardan ilki ve en yaygını paletli depo raf sistemleridir.
Paletli depo raf sistemlerinde, raflara konacak ürünler çeşitli yüksekliklerdeki raf
kirişlerine paletler üzerinde yerleştirilirler. Sistemin bileşenlerini, kolon ve
çaprazların oluşturduğu çerçeveler ve bu çerçeveleri birbirine bağlayan, paletlerin
üzerlerine yerleştirildiği yatay kirişler oluşturmaktadır. Çapraz elemanlar ve kirişler,
kolonlara genellikle bulonlu veya kaynaklı birleşimler ile bağlanırlar. Avrupa’da ve
ülkemizde yaygın olarak bulonlu birleşimler kullanılmaktadır. İkinci tip depo raf
sistemleri ise içine girilebilir raf sistemleridir. Bu sistemlerde kolonlara bağlanan
yatay kirişler bulunmaz ve konulacak yükler forklift gibi araçlar ile sistem üzerine
sürülürler. Paletli depo raf sistemlerinin daha yaygın kullanılmasının sebebi, ürünlere
her iki yönden kolayca ulaşılabilmesidir.
Bu çalışmada öncelikle soğukta şekillendirilmiş genel çelik kolonların yapısal
davranışı incelenmiştir. İnce cidarlı çelik kolonların eksenel yük altında taşıma
kapasitelerini tayin edebilmek için dikkat edilmesi gereken başlıca unsurlar, akma,
global burkulma, yerel burkulma ve çarpılmalı burkulmadır. Akma için kapasite
hesabında kesit alanı ile malzemenin akma dayanımı kullanılır. Kesit kalınlıkları
düşük olduğu için genelde kolon kapasitesi, akma dayanımına ulaşmadan burkulma
problemleri oluşur. Global burkulmanın, eğilmeli burkulma, burulmalı burkulma ve
bu iki burkulma modunun aynı anda olduğu durum olarak üç farklı türü mevcuttur.
xxvii
İnce cidarlı çelik kolonları oluşturan elemanlar, şekil verilmiş çelik plak elemanlar
olduğundan dolayı yerel burkulmayı basınca maruz kalan plakların davranışı
belirlemektedir. Yerel burkulma için yapılan hesaplarda yaygın olarak efektif alan
yöntemi kullanılmaktadır. Çarpılmalı burkulma, özellikle açık ağızlı kesitlerde
karşılaşılan, kolonun kesit ağızlarının içe yada dışa doğru açılarak deforme olması ile
oluşan burkulma durumudur. İnce cidarlı çelik kolonların kapasite hesabında
kolonun taşıyabileceği maksimum yük, tüm bu etmenler dikkate alınarak hesaplanan
kapasitelerin minimum olanı alınarak hesaplanmaktadır.
Depo raf sistemleri taşıyıcı çerçevelerinde kullanılan kolon elemanların güvenliği
sistem güvenliği açısından özel ve önemli bir yer teşkil etmektedir. Çelik depo raf
kolonları genel soğukta biçimlendirilmiş ince cidarlı çelik kolonların özel bir tipi
olarak kabul edilebilir. Fakat, çelik depo raf kolonlarının yapısal davranışı;
kalınlıkları, kolon ağzı konfigürasyonları ve delikli yapılarından dolayı soğukta
şekillendirilmiş genel çelik kolonlardan farklılık göstermektedir. Çelik depo raf
kolonları, çeşitli kolon ağzı tipleri ve uzunlukları ve kullanım amacına göre çeşitli
delik konfigürasyonlarına sahip olabilirler. Bu iki ana değişken, raf kolonlarının
taşıma kapasitesini ve bundan dolayı tüm raf yapısının dayanımı ve stabilitesini
etkileyen başlıca etmenlerdir.
Çelik raf kolonlarının kapasitesinin hesabı şuan çeşitli yönetmelikler ve metotlar ile
düzenlenmektedir. “Amerikan Demir ve Çelik Enstitüsü S100-2007” standardı,
soğukta biçimlendirilmiş çelik kolonların dayanımlarının hesaplama yöntemlerini
düzenleyen az sayıda yönetmelikten biridir. Bu standartta hesap yöntemleri olarak
“Efektif Genişlik Yöntemi” ve “Doğrudan Dayanım Yöntemi” tarif edilmektedir.
Efektif genişlik yöntemi, eşmerkezli eksenel yüklenmiş ince cidarlı çelik elemanların
dayanımları için kullanılan klasik yöntemdir. Yapılan çalışmalarda, basınç altında
ince cidarlı çelik kolon kesitlerinin tüm alanları ile çalışmadıkları görülmüştür.
Dolayısı ile kapasite hesabında öncelikle kesit alanı, gelen basınca dayanan kısma
yaklaşacak şekilde yönetmelikte tarif edilen plakların yerel burkulma durumu dikkate
alınarak azaltılır. Global burkulma gerilmesi, akma gerilmesi ile ilişkilendirilerek
kolon narinliği hesaplanır. Bulunan kolon narinliği ve akma gerilmesi kullanılarak
kolon kapasitesi tayin edilir.
Doğrudan dayanım yöntemi, eşmerkezli eksenel yüklenmiş ince cidarlı çelik
elemanların dayanımlarının belirlenmesi için son zamanlarda geliştirilen bir yöntem
olup AISI S100-2007 standardının EK-1 kısmında tarif edilmektedir. Yapılan
çalışmalarda, birçok çelik kolon kapasitesinin test sonuçları biraraya getirilerek
direk dayanım yöntemi için öngörü eğrileri çıkarılmıştır. Bu eğrilerden elde edilen
analitik formüller ile her bir burkulma modu için kolon kapasitesi hesap
edilebilmektedir. Elemanların kritik elastik burkulma yükleri, her bir burkulma modu
için sonlu şerit yöntemi kullanan “CUFSM” paket programı veya sonlu eleman
yöntemi ile hesaplanır. Kritik elastik burkulma yükleri, direk dayanım yöntemi
formüllerinde kullanılarak her bir burkulma modu için kolon dayanımı hesap edilir.
Elde edilen dayanımların en küçüğü alınarak kolon kapasitesi tayin edilir. Direk
dayanım yönteminde yerel burkulma ve global burkulmanın etkileşimi dikkate
alınmaktadır. Çarpılmalı burkulma, akma dayanımı ile ilişkilendirilmektedir. Direk
dayanım yöntemi kullanılarak yapılan kapasite hesabında doğru sonuç elde
edilebilmesi için ilgili yönetmelikteki kesit geometrisi limitleri dikkate alınmalıdır.
xxviii
Bu tezde, çeşitli özelliklerde ve aralıklarda örnek kolonların dayanım hesabında
doğrudan dayanım yöntemi kullanılarak kolon ağzı tipi, yönü ve uzunluğu, kolon
boyu ve delik varlığının kolon dayanımı üzerindeki etkileri incelenmiştir.
Mevcut çalışma aşağıdaki şekilde özetlenebilir;

- Doğrudan dayanım yöntemi ve çelik depo raf kolonları ile ilgili yapılan önceki
çalışmaların incelenmesi ve özetlenmesi.
- Soğukta biçimlendirilmiş çelik kolonların taşıma kapasitelerinin mevcut
yönetmelikteki (AISI S100-700) Efektif genişlik yöntemi ve doğrudan dayanım
yöntemi ile örnek çözümler verilerek değerlendirilmesi.
- Kolon boyları, kolon ağzı yönleri ve uzunlukları, delikli olup olamama gibi
durumları bakımından çeşitlilik gösteren 120 adet kolon modelinin dayanım hesabı
üzerinde analitik bir çalışma yapılması.
AISI S100-700 standardında deliklerin kolon dayanımı üzerindeki etkisinin, fiziksel
test yapılarak bulunması tarif edilmektedir. Üç ana burkulma modu olan yerel
burkulma, çarpılmalı burkulma ve global burkulma üzerinde delik etkisinin
incelenebilmesi için bir tasarım kuralı verilmemiştir. Bu çalışmada, doğrudan
dayanım yöntemi ile bütünleştirilerek son zamanlarda geliştirilen bir yöntem olan
“Azaltılmış Kalınlık Yaklaşımı” kullanılmıştır. Üç farklı kolon boyunda, delilikli
olup olmama durumları için kolon ağzı konfigürasyonundan dolayı değişken
geometrili kesitlere sahip kolon modelleri dayanımları hesaplanarak karşılaştırmalar
yapılmıştır.
xxix
xxx
1. INTRODUCTION
1.1 Overview of Cold Formed Steel
Cold-formed steel (CFS) is the term used for products which are made by rolling or
pressing thin gauges of sheet steel. Usable products of cold-formed steel material can
be achieved by stamping, rolling or pressing the steel sheets. These products are
commonly used in various areas of manufacturing. Automobiles, industrial products,
railway coaches, storage racks, highway products, bridges, types of equipment etc.
However, cold-formed steel phrase is generally used to describe as a construction
material. There are two main structural member families in steel construction. Hot-
rolled shapes and cold-formed steel members. The use of cold-formed steel
construction materials’ popularity has increased since 1950s. In general expression,
structural and non-structural elements in the construction industry are created from
cold-formed members. These building materials form columns, beams, joists, studs,
floor decking, built-up sections and other components. Cold formed steel members
are also known as thin-walled steel members in construction industry. Some CFS
member shapes are given in figure 1.1. The material thicknesses for these members
usually range from 0.4 mm to about 6.4 mm [1-2].
Figure 1.1 : Cold formed steel member shapes [3].
1
1.1.1 Production
The hot-rolled steel shapes are formed at elevated temperatures while the cold-
formed steel shapes are formed at room temperature. Cold-formed steel structural
members are shapes commonly manufactured from steel plate, sheet or strip material.
The producing and manufacturing process includes forming the material by either
cold roll-forming or press-breaking to achieve the desired shape. Cold roll-forming is
the most widely used method for production of floor and wall panels, roof members
and structural components like C, Z and hat sections. Sections can be made from
sheet up to 1.5 m wide and from coils more than 1000 m long. Press-braking is often
used for production of simple shapes.
Cold roll-forming process is shown in the figures 1.2 and 1.3. Flat sheets are given
the desired shapes by cold-forming of the sheets between logitudinal rolls. Complex
shapes can require 24 to 30 rolls whereas a simple section may require as few as six
pairs of roll. Generally, the thickness of material that can be formed ranges between
0.10 mm up to 7.7mm, although heavy duty cold forming mills can handle steel up to
19 mm thick.
The manufacturing of cold-formed steel products occurs at room temperature. Design

considerations for the strength of these elements are usually deals with the effects of
buckling. The yield stress of the steels commonly used for cold-forming ranges from
230 to 380 MPa and higher. With regards to for the formability, tensile strength and
ductility are the other distinctive properties. The ratio of tensile strength to yield
strength for cold-formed steels commonly ranges from 1.2 to 1.8 [4].
Figure 1.2 : Cold roll-forming process (description) [2].
2
Figure 1.3 : Cold roll-forming process (real) [5].
1.1.2 Structural applications
There are many applications around the world which are related with the cold-
formed elements. Applications for structural purposes cover around %50 of general
usage around the world with cold formed steel. Industrialized housing, standardized
metal buildings, steel storage rack structures, pre-engineered buildings are general
structural areas related with cold formed steel materials.
Construction where cold-formed steel is finding wider application is in steel framing

systems for residental, commercial, agricultural and industrial applications. Small
buildings can be made entirely with cold formed steel sections as shown in Figure
1.4 Pre-engineered metal building manufacturers provide custom designed structures
from small residental houses to sophisticated structures such as schools, hangars, and
complex manufacturing facilities. Three adjacent building blocks shown in Figure
1.5 raised in Paris is very attractive due to a playful and optimized use of the cold
formed steel structure as a good example. A cold-formed steel hangar can be seen in
Figure 1.6.
3
Figure 1.4 : Cold formed steel building example [6].
Figure 1.5 : Cold formed steel structure example [7].
4
Figure 1.6 : A cold formed steel hangar [8].
More specifically, cold formed steel elements can be grouped in two major headline
according to their role in structural applications; individual structural members,
panels and decks.
Individual structural members consist of various shapes and sections for cold-formed
construction. Channels (C-sections), I sections, T sections, tubular members, Z-
sections, hat sections, angles are usual shapes used in cold-formed structural framing.
Generally, depth of members ranges from 50 to 305 mm and the thickness ranges
from 1.20 to 6.35 mm. In some cases, depth may be up to 457mm and thickness also
maybe up to 12.7 mm or thicker in building construction and transportation.
Some of the sections are shown in Figure 1.7.
Figure 1.7 : Some CFS section geometries [1].
5
Stiffness, structural strength and load carrying properties are the main design
considerations of an individual framing member.
The second major headlines of cold-formed sections are panels and decks. These
sections are generally used for floor decks, roof decks, bridge forms, wall panels and
siding materials. In general, the depth of panels ranges from 40 to 200 mm and
thickness ranges 0.5 to 2 mm. Besides their role to carry loads, these section also
provide a surface on which roofing, flooring or concrete fill can be applied. Some
perforations can be made on these sections combining with sound absorption
material to form an acoustically conditioned ceiling. As a result of their geometry,
they provide space for electrical conduits, communications and data cable
distribution as well as heating and air conditioning ducts. In Figures 1.8 and 1.9,
typical roof decks are shown. In figure 1.10, an example of cold formed wall panel is
given.
Figure 1.8 : Cold formed steel roof deck sheet example [9].
Figure 1.9 : Cold formed steel roof deck example [10].
6
Figure 1.10 : An example of cold formed wall panel [11].
Cold formed steel storage rack systems are another important category of cold form
steel structures. Rack systems are widely used throughout the world for storing
materials. Due to their vertical character these systems provide high storage density,
allowing the storage of a great amount of products in reduced areas. They also allow
great accessibility to the stored materials. These systems are used by many
companies to store their products in large scale. There are various models which fit
to the conditions demanded accessing each product to be stored in the available
room. These models vary from simple shelves to automated structures of more than
30 m height. Within these various models, Pallet and Drive-in Systems are widely-
used ones [12].
Pallet Racking Systems allow for the storage of palletized materials in horizontal
rows with multiple levels. Main parts of these systems are upright frames, load
carrying beams, braces and base plates. A basic schematic sketch is given in Figure
1.11.
7
Figure 1.11 : A basic schematic sketch of a CFS storage rack structure [13].
Upright frames are in blue color which include cold formed columns and braces,
beams are in orange color. The most widely used beam sections are box sections. On
the other hand, open C sections and many different configurations of perforated cold
formed sections are used as columns forming uprights. Braces are the members of
upright frame lacing can be welded or bolted between two upright columns to form
upright frames. One of the peculiar structural concepts is Drive-In models. In these
models the absence of transversal beams allows fork-lifts to move inside the
structure and the presence of perforations on columns also facilitate the assemblage.
Drive-in and pallet racking examples are shown in Figures 1.12 to 1.14.
Pallet racking systems are used more often with regard to Drive-in models.
Accessibility to goods from both sides makes pallet-racking systems more preferable.
In Europe, the designers prefer bolted connection types more than the welded
connection types on connections between beam-columns and beam-braces. In US,
the welded connections are preferred mostly.
8
Figure 1.12 : Pallet racking example from Turkey [Personal Photo].
Figure 1.13 : A drive in model of rack stuctures [14].
9
Figure 1.14 : Pallet racking structure example [15].
1.1.3 Advantages and disadvantages of cold formed steel
Cold-formed steel offers many advantages. If compared with other materials, like
timber or concrete; cold formed steel structural members have following
characteristics;
- Lightness
- Mass production and prefabrication ease
- High strength and stiffness
- Easy, quick and facilitated assembly
- Delays due to weather conditions
- Accurate detailing
- Formwork is not needed
- Rot and termite proof
- Low transportation and handling expenses
- Recyclability
- Shape variability
10
- Uniform of quality
Also, cold-formed steel structural members provide structural and constructional

advantages. For short spans and relatively light loads, cold formed light members can
be manufactured in comparison with thicker hot-rolled sections. Some of cold
formed sections can be nested which allows compact packaging and transportation.
Various sectional configurations can be produced with cold-forming operations
economically. A useful surface can be provided for floor, roof and wall construction
by load carrying cold-formed panels and decks. Load carrying panels act like a shear
diaphragm to resist force in their own planes if they are adequately interconnected
with each other and supports beside withstanding loads normal to their surfaces.
Favorable strength to weight ratios can be obtained by variation of cold formed steel
sections [1].
Designing with cold-formed steel shapes and with hot-rolled structural shapes
differs.With the hot-rolled, primarily concerns are about two types of instability:
lateral buckling of unbraced beams and column buckling. Local buckling of
individual constituent elements generally will not occur before yielding due to the
dimensions of hot-rolled shapes. However, this is not the case with cold-formed
members. In most cases, the material used is thin relative to its width which causes to
local and distortional buckling hence these failure modes must also be considered.
This can result in premature buckling of individual flat, or plate, elements at stresses
well below the yield stress point [4].
Sometimes, high unit-prices, low fire resistance, more careful requirement of

treatment during assembly and difficult connections may be considered as
disadvantages.
1.2 Aim and Layout of the Thesis
This study mainly focuses on the approaches to predict the strength and stability of
thin walled steel columns in storage rack structures. Direct Strength Method, which
is a recently proposed approach on this topic is investigated in detail. The major
differences between columns in steel storage rack structures and general cold formed
steel columns are lip configuration and perforations. The effects of perforations and
lip configuration including lip type, lip forming direction and column length on
11
nominal column strength are investigated using Direct Strength Method on a range of
various specimens in this thesis.
Chapter 1 includes an overwiev on Cold-Formed Steel. This chapter discusses

production phases, structural applications, advantages and disadvantages of CFS.
The thesis moves on with a detailed literature review on cold-formed steel columns.
Under the “General Remarks” title in Chapter 2, the behaviour of cold formed steel
columns with respect to phenomenas like yielding, buckling and other buckling types
is summarized. Second part of Chapter 2 focuses on the studies conducted using
Direct Strength Method. The third part of this chapter summarizes various studies
conducted on perforated cold-formed steel storage rack columns.
Chapter 3 consists of an overview of the design rules given in American Iron and
Steel Institute S100-2007 Standart on concentrically loaded compression CFS
members. These design rules include the widely used Effective Width Method and
recently proposed Direct Strength Method. (Appendix 1 of AISI S100-2007)
Example Solution of a C-Channel Column with a Simple Lip Edge Stiffener is
presented via Effective Width Method and Direct Strength Method.
Main focus of Chapter 4 is on the investigation of nominal strength of various rack

columns through Direct Strength Method. This investigation includes the analysis of
120 column models having different lip configurations and column lengths. Both
perforated and unperforated columns were considered.
Chapter 5 presents the final conclusions drawn from the present study and
suggestions for future work.
12
2. LITERATURE REVIEW ON COLD FORMED STEEL COLUMNS
There have been many studies made on cold-formed steel columns since 1970s.
Design methods are improved due to many experimental results and numerical
studies. Effective width method is traditional method for determining strength of
cold formed steel sections. Direct strength method can also be considered as new
method for determining specifically cold formed steel column sections. This method
has been validated by many experiments and studies that will be summerized under
this topic. Besides, many studies have been made on perforated storage rack columns
which needs a different approach. It is necessary to realize the importance of
perforations on strength and behaviour of the columns. In this chapter, literature
survey is grouped into three parts. General Remarks on Cold-Formed Steel Column
Behaviour, Studies on Direct Strength Method and Studies on Perforated Cold-
formed Steel Storage Rack Columns.
2.1 General Remarks on Cold Formed Steel Column Behaviour
Thin walled cold formed steel compression members are used to carry a compressive
load through centroid of their cross-section. Cross-sections may be composed of
stiffened elements, unstiffened elements or combination of stiffened and unstiffened
elements. Most of the time, shear center of cross-sections may not coincide with the
centroid of the section. Thus; during the design phase; following limit states must be
considered depending on the configuration of the section, column lengths used and
thickness of the material;
- Yielding
- Global (overall) Column Buckling including;
Flexural Buckling (bending about a principal axis)
Torsional Buckling (twisting about shear center)
Flexural-Torsional Buckling (bending and twisting simultaneously)
13
- Local Buckling
- Distortional buckling (buckling of open sections with edge stiffened flanges)
Design provisions of global (overall) buckling and effect of local buckling on

column strength have been included in the AISI Specification [16]. These provisions
were added to Specification in 1968 by detailed investigations of George Winter,
Alexander Chajes, Pen Jeng Fang and Teoman Peköz at Cornell University [17-19].
Unified approach of design provisions developed in 1986 and discussed by Peköz in

[20]. In this approach, following steps for the design of axially loaded compression
members are given;
- Calculate the elastic column stress (flexural, torsional, or flexural-torsional)

for the full unreduced section
- Determine the nominal failure stress (elastic buckling, inelastic buckling or

yielding)
- Calculate the nominal column load based on the governing failure stress and
the effective area
- Determine the design column load from the nominal column load using the
specified safety factor or the resistance factor
North American Specification had the design provisions for determining the
distortional buckling strength of I, Z, C, hat and other open sections having edge-
stiffened flanges in 2007.
2.1.1 Yielding
Yield load can simply be described as;
(2.1)
Where “A” is the cross-sectional are of column and “Fy” is the yield stress of steel.
Very short, compact columns under axial load may fail by yielding.
2.1.2 Global (overall) column buckling
Global buckling of steel columns can be inspected into three parts. Flexural Column
Buckling, Torsional Column Buckling and Flexural-Torsional Column Buckling.
14
2.1.2.1 Flexural column buckling
Elastic buckling
Slender axially loaded cold formed steel column sections such as doubly symmetric
shapes, closed shapes, point-symmetric shapes or cylindrical shapes can easily be
subjected to fail by overall flexural buckling. The elastic critical buckling load for a
long steel column can be calculated by the Euler formula given in (2.2).
(2.2)
Where;
Pe is the Euler Buckling Load
E is the Modulus of Elasticity
I is the Moment of Inertia
L is the column length
K is the effective length factor
If substitution made on (2.2) by I = Ar2 , where A is the cross-section area, Euler

stress for elastic column buckling is obtained and given in (2.3);
(2.3)
Where KL/r is the effective slenderness ratio and r is the least radius of gyration.
Inelastic buckling
There are two different approach in the literature for inelastic buckling concept. The
tangent modulus method, firstly proposed by Engesser in 1889. Tangent modulus
load equation is given in (2.4) according to this method:
(2.4)
Where Et is the tangent modulus. Critical buckling stress is;
15
(2.5)
As the second method, reduced or double modulus concept including effect of elastic
unloading has been developed by Janisky in 1895. Equations used by this method are
as follows;
(2.6)
And
(2.7)
Where Er is the reduced modulus which can be defined as E (I1 / I) + Et (I2 / I)

I1 is moment of inertia about neutral axis of the area on unloading side after
buckling. I2 is moment of inertia about neutral axis of the area on loading side after
buckling.
Francis R. Shanley’s contributions on this topic are revealed in [21];
- Tangent modulus concept gives the maximum load up to which an initially

straight column remains straight.
- The actual maximum load exceeds the tangent modulus load, but it cannot
reach the reduced modulus load.
Other investigations and experiments improved Shanley’s concepts and Galambos

stated in [22] that the maximum load is usually higher than the tangent modulus load
by %5 or less.
The Structural Stability Research Council has pointed that if the effect of residual
stress is considered and the effective proportional limit is assumed to be equal to
one-half the yield stress, Eq. (2.5) can be approximated by the following equation:
( ) ( )( ) (2.8)
16
Where Fy is the minimum yield stress. Figure 2.1 shows that the value of √
is the limiting KL/r ratio corresponding to a stress equal to Fy/2. Elastic buckling
governs when this ratio is greater than limiting ratio. Inelastic buckling governs when
this ratio is smaller than limiting ratio.
Figure 2.1 : Flexural column buckling stress [1].
These basic equations have changed during years due to lots of investigations and
tests. Finally, AISI Specification states that the design equations calculating the
nominal inelastic and elastic flexural buckling stresses are calculated as follows:
( ) when (2.9)
[ ] when (2.10)
Where (Fn)I is the nominal inelastic buckling stress, (Fn)e is the nominal elastic
buckling stress, √ is the column slenderness parameter, is the
theoretical elastic buckling stress of the column which can be calculated by Eq.
(2.3).
2.1.2.2 Torsional buckling and flexural-torsional buckling
Due to their large torsional rigidity, closed sections are not always vulnerable for
torsional buckling. However, for open thin-walled members, three main buckling
modes should be considered during the analysis. These are flexural, torsional and
17
flexural-torsional buckling modes. As a definition of open-section columns in
flexural-torsional buckling mode, bending and twisting occur simultaneously. The
first investigations known on this problem were made by Goodier, Timoshenko and
Bleich [23-25]. Further studies conducted by George Winter and Alexander Chajes
in [18] and the same authors with Pen Jeng Fang in [19]. These further studies
formed a basis for the development of the AISI design criteria. In Figures 2.2 and 2.3
, flexural-torsional buckling examples on different sections are given.
Figure 2.2 : An exaggerated flexural-torsional column buckling example [26].
Figure 2.3 : Flexural-torsional column buckling of an I section [27].
Besides the calculation of the critical stress for flexural buckling discussed in
previous section, torsional buckling stress can ben calculated by the following
equation;
18
[ ] (2.11)
Where A is the area of the section, r0 is the polar radius of gyration of cross section
about shear center, G is the shear modulus, J is St. Venant torsion constant of cross
section, E is the modulus of elasticity, Cw is the warping constant of torsion of cross
section, Kt is effective length factor for torsion and Lt is unbraced length of
compression member for torsion. GJ describes the torsional rigidity and also ECw
describes the warping rigidity.
For the doubly symmetric sections like I sections, H sections etc. the shear center
coincides with the centroid of the section. Usually column fails with pure bending or
pure torsion. So, formulas given above can be used to calculate the desired stresses.
However, mostly in singly-symmetric sections like C sections, hat sections etc.,
flexural-torsional buckling possibility is higher than pure failure modes.
Elastic Flexural-Torsional buckling stress can be calculated by the following

equation;
[ √ ] (2.12)
Where; and
Px is the Euler flexural buckling load about x axis and Pz is the torsional buckling
load about z axis. Equations (2.13) and (2.14) are about the calculation of Px and Pz.
(2.13)
[ ]( ) (2.14)
In case of inelastic flexural-torsional buckling, following formula is used for

calculating inelastic flexural-torsional buckling stress:
( ) (2.15)
19
Equations (2.12) and (2.15) are graphically plotted in Figure 2.4.
Figure 2.4 : Maximum stress for flexural-torsional buckling [1].
Eq. (2.15) has been used in AISI specificaion up to 1996, similar to the case for
flexural column buckling. After 1996, the nominal inelastic flexural-torsional
buckling stress was computed by equation (2.9) in which;
√ (2.16)
2.1.3 Local buckling
Local buckling of individual component plates usually occurs before the applied load
passes the limit of the overall collapse load limit of the column in cold formed steel
compression members. Local and global column buckling interaction is very
important and may cause a reduction of the overall column strength. Following
factors form a basis for the influence of local buckling on column strength;
- Slenderness ratio of column
- Cross section shape
- Governing global column buckling type (flexural, torsional or flexural-

torsional)
- Steel material’s mechanical properties and type
20
- Imperfections
- Welding
- Perforations
- Residual Stress
- Plane components’ interaction
- Effect of cold work
Investigations on the interaction of local and global buckling on steel columns have
been conducted by Amos Henry Chilver, V. Kalyanaraman, Teoman Peköz, Paulus
Pieter Bijlaard, G. P. Fisher, Victor Gioncu and more during past 60 years [1].
Local buckling can be understood in more detail by determining the structural

behaviour of plates. Because in general, cold formed steel column sections are
formed with plates which are in interaction.
If a stiffened compression element’s width / thickness ratio is relatively small, the

strength of that element may be governed by yielding. If its width / thickness ratio is
relatively large, it may be governed by local buckling at a stress level less than yield
stress. In Figure 2.5, different local buckling modes of a box section are shown.
Figure 2.5 : Example of different local buckling modes [28].
If a simply supported square plate is subjected to a uniform compression stress in one

direction, it will buckle in a single curvature in both directions as shown in Figure
2.7. The length of the element is usually much larger than the width for individual
elements. Figure 2.6 shows the local buckling of stiffened compression flange of a
hat-shaped section.
21
Figure 2.6 : Local buckling of stiffened compression flange of a hat-shaped section
[17].
The critical buckling stress of a plate can be calculated by solving Bryan’s

differential equation based on small deflection theory. Figures 2.7 , 2.8 and Eq.
(2.17) describe the situation of a plate subjected to compression stress.
Figure 2.7 : Square plate subjected to compression stress [1].
Figure 2.8 : Rectangular plate subjected to compression stress [1].
(2.17)
Where;
(2.18)
22
E is the modulus of elasticity of steel, t is the thickness of plate, is the Poisson’s
ratio = 0.3 for steel, is the deflection of plate perpendicular to surface, is the
compression stress in x direction, is the halfwave length , a and w is the width of
the plate, respectively and s.s. meaning the simply supported.
Solution of this differential equation represents a general equation for critical load
buckling stress for a rectangular plate subjected to compression stress in one
direction which is given in Eq. (2.19).
(2.19)
Investigations on k values conducted with George Gerrard and Herbert Becker in

1957 and given in [29]. Figure 2.9 shows the k values used in eq. (2.19) for different
ratios where m is the numbers of half sine waves occur during local buckling
and it can be assumed that length of the half sine waves, for
respectively for long plates.
Figure 2.9 : Buckling coefficient for flat rectangular plates [29].
It should be notified that ratio is an integer the value of k equals 4. K can be

taken as 4 for relatively large ratios. The values of k for a long rectangular plate
subjected different types of stresses and within different kind of boundary conditions
have been investigated with Stephen P. Timoshenko and Stephen M. Gere and given
in [30]. Figure 2.10 shows the results of studies on values of k for determining
critical buckling stress.
23
Figure 2.10 : Values of k for determining critical buckling stress [30].
For the analysis of the buckling plates in the inelastic range, equations used were
modified by Bleich [25]. Basically, plate becomes an anisotropic plate which means
different properties in different directions of plate when the compression stresss in
only one direction exceeds the proportional limit of the steel. For this inelastic case,
Friedrich Bleich proposed following differential equation, (2.20), in [25].
( √ ) (2.20)
Where and Et is the tangent modulus of steel. After applying the modified
boundary conditions and solving this differential equation, following critical
buckling stress for plastic buckling of plate formula is obtained;
√ √ (2.21)
The wavelength for a long plate can be obtained by √

24
AISI specification used Q-factor method during the period from 1946 to 1986 for the
effects of local buckling on column strength. This method developed by
contributions of Winter and Fahy [17]. Q-factor method basically aims to assume a
reduced stress in yield stress equation previously given in Eq. (2.1). Because in cold-
formed steel members, if w/t ratios of compressions are relatively large, local
buckling may occur before yielding.
( ) (2.22)
Where Q is a factor which weakens the yield strength and less than unity due to the
effects of local buckling. This factor depends on the shape and form of the section.
Even though this method has been used for years, the recent stuides and
developments have shown that this method needs improvement. With the
contributions of DeWolf, Winter, Peköz, Kalyanaraman and Loh; test results and
analytical studies have shown that for compression members having stiffened
elements with large width/thickness ratios and for those members having
slenderness ratios around 100, Q-factor method can give unconservative results [20].
Q factor method was really precise on especially I sections which have unstiffened
flanges with small slenderness ratios. After 1986, Q-factor method has been
eliminated in AISI specification and Effective Area approach has been accepted.
This area basically focus on reducing the total cross-section area of cold formed steel
compression members for the influence of local buckling in yielding equation
calculating the nominal column load given in Eq. (2.23).
(2.23)
Where, the effective area of section is indicated by . In case of effective area can
not be calculated by generalized effective width calculations of AISI specification, it
can be calculated by stub column tests which is described by Peköz in [20]. Kim
Rasmussen and Gregory Hancock conducted studies on channel sections given in
[31] and also Rasmussen conducted another study on unstiffened flanges given in
[32] so that all of these studies guided building up the latest version of AISI
Specification and North American Design Provisions. These provisions of AISI
Specification on cold formed steel sections under uniform compression will also be
25
discussed in the “Codified Design of Cold-Formed Steel Columns” chapter in this
thesis.
2.1.4 Distortional buckling
The distortional buckling mode involves the rotation of each flange and lip about the
flange-web junction of cold-formed steel columns under compression stress. This
mode is usually seen in open sections like Z and C channels as their geometry is
suitable for this kind of failure. In Figure 2.11, distortional buckling of a C channel
with stiffened lips is given.
Figure 2.11 : Distortional buckling of a C channel with stiffened lips [33].
Distortional buckling is one of the important failure modes for open cross-sectional
members having stiffened flanges which should be precisely considered in column
design. Gregory Hancock have conducted many studies and research on distortional
buckling mode. Figure 2.12 shows an example of rack section column buckling
stress versus half wavelength for concentric compression from Hancock’s
investigations [34].
Many other researchers like Kwon, Schafer, Merrick, Tovar, Sputo, Camotim,
Davies, Jiang and Silvestre have been made contributions on this specific buckling
mode and all summerized in [34]. North American Specification included a new
section 4.2, for determining the distortional buckling strength of open sections
compression members like I, Z, C and hat shapes. In this thesis, distortional buckling
stregth section of AISI Specification will also be discussed in the “Codified Design
of Cold-Formed Steel Columns” chapter.
26
Figure 2.12 : Rack section column buckling stress versus half wavelength for
concentric compression [34].
2.2 Studies on Direct Strength Method
The direct strength method has been adopted as an alternative design methodology in
Appendix 1 of the North American Specification including the design of cold-formed
steel structural members [16]. Practicaly, DSM (Direct Strength Method) brings a
new perspective and a relatively different approach for calculating the nominal
strength of cold formed steel structural sections. DSM needs numerical methods such
as finite strip method and generalized beam theory in calculation of member elastic
behavior primarily for accurate solution. Computer software “Cornell University
Finite Strip Method, CUFSM“ using finite strip method is programmed by Benjamin
W. Schafer, Zhanjie Li and Sándor Ádány to explore elastic buckling behaviour and
calculate the buckling stress buckling mode of arbitrarily shaped, simply supported,
thin-walled members. Outputs of these calculations are used in DSM equations for
prediction of nominal strengths of cold formed steel sections. Some important
headlines related with DSM will be defined under this topic. Detailed information on
AISI Specification about DSM will be described under “Codified Design of Cold-
formed Steel Columns” chapter in this thesis.
Finite strip method

Finite Strip Method is a specialized finite element method used by CUFSM. Element
shape functions use polynomials in the transverse direction. However, they use
27
trigonometric functions in the longitudinal direction. Judicious choice of the
longitudinal shape function allows a single element which is called “strip” to be
used. Finite strip method uses a single half sine wave (sin(πx/a)) for the longitudinal
direction. In Figure 2.13, approach difference between an example sketch of finite
element method and finite strip method is shown.
Figure 2.13 : Approach of finite element method and finite strip method [35].
Buckling curve
The buckling curve is a graphical representation showing a section’s elastic buckling
behaviour and also the primary result from a finite strip analysis. A typical buckling
curve is shown in Figure 2.14. The minima of curvatures on this curve indicate the
critical halfwavelength and load factor for a given buckling mode.
Figure 2.14 : An example of buckling curve [35].
Buckling mode
Basically, the buckling mode means the shape that a member buckles into. A
buckling mode represents a secondary deformed shape that has the same potential
energy as the primary deformation. As an example, local buckling mode shape of a Z
section is given in Figure 2.15.
28
Figure 2.15 : Local buckling mode of a Z section [35].
Half-Wavelength
Half-Wavelength is 1/2 of the sine waves assumed during the finite strip analysis
which is performed for systematically increasing halfwavelengths to determine the
buckling behavior of a member.
Load factor
During the finite strip analysis, the member assumed to be loaded with a reference
stress distribution. Buckling stress is found by multiplying the load factor with the
reference stress distribution. In more detail, the load factor is the eigenvalue of the
relevant eigenvalue buckling problem where the buckling mode is the eigenvector.
Reference stresses are generally chosen as yield stresses during the analysis of cold
formed steel members which are under compression.
There have been several studies on DSM which will also be summarized under this
chapter.
Design of Cold Formed Steel Channel Columns with Complex Edge Stiffeners by
Direct Strength Method;
Ben Young, Jintang Yan (2004)
Mainly, this paper focuses on the comparisons between the failure modes observed
in the tests with the failure modes predicted by the direct strength method [36]. The
test program and results are taken from a previous study of Ben Young and Jintang
Yan [37], has a total of 30 column specimens were tested. Comparison of test
strengths with design strength for a serie of specimens shown in Figure 2.16. Design
Strengths have been found by DSM (Direct Strength Method), The Australian/New
Zealand Standard (AS/NZS) and North American Standart (NAS).
29
Figure 2.16 : Comparison of test strengths with design strengths [36].
The test specimens were tested between fixed end supports. It is shown that in the
prediction of column strengths of the fixed-ended cold-formed steel channels with
complex edge stiffeners having slender flanges, accurate results are achieved by
using the direct strength method. For sections having less slender flanges, direct
strength method conversatively predicts the column strengths. As another point
shown in this paper is for the failure modes. For the tests of long columns, the failure
modes predicted by direct strength method are generally in parallel with the failure
modes observed in the tests. However, for short columns and intermediate columns,
this agreement of predictions is not valid.
Direct Strength Design of Hot Rolled and Cold-Formed Steel Compression

Members;
Gregory J Hancock, Daniel O Cook, Robert E. Moisy and Andy Yen (2005)
This paper focuses on the analysis of hot rolled sections like I-sections, square
hollow sections and fabricated sections with Direct Strength method due to put its
simplicity forward than Effective Width Method [38]. Paper begins with a
computation that shows the analysis results of an opened channel section under
compression given in Figure 2.17. Finite Strip Buckling Analysis given in Figure
2.17, conducted by Papangelis and Hancock in 1995 [39].
30
Figure 2.17 : Buckling stress versus half-wavelength for a C-Section in compression
[39].
Figure 2.17 clearly shows the the three basic buckling modes for design of thin-
walled C-section used under compression which are the local, distortional and
overall modes. Paper indicates that these three modes are characteristic of thin-
walled open-section members in compression or bending. Than, studies on hot rolled
sections, welded series and unlipped channel sections have been conducted. In
conclusion; the DSM was found to give better design accuracy than the Effective
Width Method (EWM) in all cases except for the very slender unlipped channel
section. The local buckling stress was much lower than the yield stress and the
effective centroid of the locally buckled section moved creating a premature failure
compared with the DSM in the case of the very slender unlipped channel. DSM
design is suggested for all other sections investigated except this type of sections.
Buckling Analysis of Cold-formed Steel Members Using CUFSM: Conventional

and Constrained Finite Strip Methods;
B.W. Schafer & S. Ádány (2006)
This paper provide technical background and examples for stability analysis of cold-
formed steel members using the conventional and constrained finite strip methods
[40]. CUFSM software uses finite strip method. Conventional Finite Strip Method is
used to examine all the possible instabilities in a cold-formed steel member under
longitudinal stresses like axial, bending, or combinations). The constratinted finite
strip method provides stability solutions to be focused only a given buckling mode
31
(modal decomposition). Modal decomposition is completed by forming a series of
constraint equations that describe a particular buckling class. For exploring cross-
section stability in cold-formed steel members, the conventional finite strip method
combined with the constrained finite strip method provide a powerful tool. Elastic
and geometric stiffness matrices are formed from a summation of cross-section strips
and employed in an eigenvalue stability analysis in the conventional finite strip
method. In this paper the stiffness matrices are explicitly derived and can readily be
used in the software. The provided solution is identical to that employed in the
CUFSM software. The strength of this new extension to finite strip solutions is the
ability to decompose and identify conventional finite strip solutions as related to
buckling classes of interest like global, distortional, or local buckling. The potential
use of the constrained finite strip method is shown in this paper with the examples
provided, and the algorithms discussed are implemented in CUFSM [40].
Review: The Direct Strength Method of Cold-Formed Steel Member Design;

Benjamin W. Schafer (2008)
This paper of Benjamin W. Schafer summarizes general remarks of DSM studies and
developments [41]. Especially, DSM for columns development part of this paper will
be described due to the relation with this thesis. The beginning of the Direct Strength
Method for columns most clearly be traced to research into distortional buckling of
rack postsections at the University of Sydney. Lau S. C. W. and Hancock G. J.
conducted studies on columns published in distortional buckling formulas for
channel columns [42]. Also, Kwon and Hancock conducted Strength tests of cold-
formed channel sections undergoing local and distortional buckling and published
their studies in Journal of Structural Engineering [43]. Hancock et al. collected the
research and demonstrated that for a large variety of cross-sections the measured
compressive strength in a distortional failure correlated well with the slenderness in
the elastic distortional mode. Hancock attributes his methodology to Trahair’s work
on the strength prediction of columns undergoing flexural–torsional buckling [44].
As a final regard, it seems clear that the Direct Strength Method is not a new idea,
but rather the extension of an old one to new instability limit states. Development of
the Direct Strength Method beyond distortional buckling was completed using a
much wider set of cold-formed steel cross-sections and tests that included failures in
local, distortional, and global flexural or flexural–torsional modes in the papers by
Schafer [45-46]. Hand solutions and numerical (finite strip) solutions for the elastic
32
buckling of 187 columns were gathered in these papers. Also lots of test has been
conducted and graphically compared with the Direct Strength Method predictor
curves given in Figure 2.18.
Figure 2.18 : Comparison of the direct strength method predictor curves with test
data for columns (equation numbers refer to those used in the North
American Specification) [41].
This figure indicates that DSM is a reasonable predictor of strength over a wide
range of slenderness. Interaction of the buckling modes were systematically studied
for local–global, distortional–global, and local–distortional buckling of the columns.
Based on overall test-to-predicted ratios, and when available the failure modes noted
by the researchers in their testing, it was concluded that local–global interaction
should be included, but not distortional–global, or local–distortional interaction. As a
result, it was recommended to only include local–global interaction in the Direct
Strength Method. Recent work by Hancock, Camotim, Yang and Silvestre is focused
on the investigations for the effect of local-distortional interactions [47-48]. Work is
ongoing to determine the most appropriate way to identify and predict the strength
for the small number of columns that do have potential local–distortional interaction.
In conclusion, Direct Strength Method is on path to be a completely viable alterative
design procedure for cold-formed steel member design according to many research
groups’ efforts around the world [41].
33
Evaluation of Direct Strength Method for CFS Compression Members without
Stiffeners;
M. V. Anil Kumar, V. Kalyanaraman (2010)
This paper includes the results of a study on the suitability of Direct Strength Method
to evaluate the compressive strength of plain channel, I and rectangular tubular
members [49]. For calculating the strength of compression members such as tubular,
plain channel, and I-sections consisting stiffened and unstiffened elements, without
either lip stiffener or intermediate stiffener it is known that the Direct Strength
Method has not been adequately evaluated. These members do not have edge or
intermediate stiffeners. Hence, these sections are not vulnerable to
distortional/stiffener buckling and experience interaction of only local and overall
buckling before failure. Authors used the test results data from following studies
previously conducted: Test results on fixed plain channel compression members [50-
52, 20], hinged ended rectangular tubular members [53-55], hinged ended I-Sections
[56]. Calculations have been made via Effective Width Method with CUFSM and
Direct Strength Method. The graphs in Figure 2.19 shows the comparison of design
methods with test results for Plain Channel Sections (PCS), I-Sections (IS), and
Rectangular Hollow Sections (RHS).
Figure 2.19 : Comparison design methods with test results [49].
In conlusion, authors observed the applicability of Direct Strength Method on

sections mentioned above. DSM estimates the strength of these compression
members within acceptable accuracy, for practical purposes according to the
comparative study with test results and EWM-CUFSM results.
On the Direct Strength Method (DSM) Design of Cold-formed Steel Columns

Against Distortional Failure;
A. Landesmann, D. Camotim (2013)
This study includes many numerical investigations have been made to determine the
influence of the cross-section geometry and end support conditions on the post-
34
buckling behaviour and Direct Strength Method (DSM) design of cold-formed steel
columns buckling and failing in distortional modes [57]. ANSYS computer software
has been used for elastic and elastic-plastic shell finite element analysis of the
columns which are analyzed exhibit, four end support conditions (fixed, pinned-
pinned, pinned and fixed-free end supports) , different cross-section shapes,
dimensions and lengths and several yield stresses. During the investigations, studies
are carefully conducted to ensure that all the columns selected have been failed in
“pure” distortional buckling mode cover a wide distortional slenderness range. From
the performance of a parametric study involving 648 columns the ultimate strength
data acquired and then used to show that, the current DSM distortional design curve
is only able to predict adequately the ultimate loads of fixed columns, regardless of
the column geometry. This design curve was shown to clearly over estimate the
numerical ultimate loads in the moderate-to-high slenderness range for the columns
with the remaining end support conditions (pinned-fixed, pinned and fixed-free).
Fig.2.20 shows a comparison between these studies and current DSM distortional
curve where Pu is ultimate load, Py is column squash load, Pcrd is critical distortional
buckling load and ( ) is the distortional slenderness.
As shown in the Figure 2.20, analyzes other than fixed end condition points a
modification on current DSM distortional curve which may be an alternative and
shown in Figure 2.21.
These studies show that the proposed preliminary DSM-based distortional strength
curves require further validation, in order to assess their reliability and universality to
find out whether they remain valid for any column cross-section shape.
35
Figure 2.20 : Comparison between the current DSM distortional curve and column
ultimate loads with different end conditions [57].
Figure 2.21 : Comparison between the modified DSM distortional curve and column
ultimate loads with different end conditions [57].
2.3 Studies on Perforated Cold Formed Steel Storage Rack Columns
There have been many studies on this specific topic of cold formed steel columns.
Rack columns are a specific version of CFS columns having perforations which have
36
different shapes and sizes. Placement of these holes vary depending on configuration
of beams, rack type and purpose of usage.
Analysis of Steel Storage Rack Columns,

A.M.S Freitas, M.S.R. Freitas, F.T. Souza (2005)
In this article, authors used experimental and numerical methods to quantify the
influence of the perforations on the load capacity of the steel storage rack columns
[12]. Finite element analysis has been done on the column types that tested before by
using “ANSYS” software and comparisons have been made. Shell, solid and contact
finite elements were used to represent the stub column analyzed in the model. 4
specimens tested by “stub column test” and ultimate compressive strengths of these
specimens were evaluated. Different cross-sectional areas have been used for the
numerical simulation of perforation influence in the load capacity. Four different
area-types are used which can be described as;
Nominal column: Actual cross section of the column, considering the perforations
with nominal dimensions.
Actual column: Actual cross section of the column, considering the perforations
with actual dimensions.
Gross column: Full cross section of the column, without perforations with nominal
dimensions.
Net column: Cross section of the column without perforations considering the area as
the minimum net area of the column.
Nominal and experimental steel properties used during the numerical analysis.
Comparisons have been made with numerical results, experimental output and the
RMI prescriptions values using experimental results (Pua - ultimate axial load of
experimental results) to yielding stress (PRMI) and using nominal values to yielding
stress (PRMI-n). Result graphs are given in Figures 2.22 and 2.23
37
Figure 2.22 : Load versus displacement curves for fy = 250 MPa [12].
Figure 2.23 : Load versus displacement curves for fy = 320.23 MPa [12].
The authors have mentioned that the representative curves of the nominal, net and
actual columns indicate similar values to experimental data, varying from 8% to 17%
from the average experimental ultimate loads. Actual column and net column
indicated good agreement with experimental results when nominal yielding stress is
used in finite element analysis of the nominal column. Another observation have
been made on the failure mode. When measured imperfections are not included, the
38
collapse mode was caused by the perforations of the cross section in the middle
height of the column in region of minimum area. When measured imperfections are
included, similar collapse mode like in the experimental test have been observed
showing the strong influence of imperfections on stub columns.
Stub Column Tests for Racking Design: Experimental Testing, FE Analysis and
EC3;
F. Roure, M. M. Pastor, M. Casafont, M. R. Somalo (2010)
This research has been conducted in the Laboratory of Elasticity and Strength of
Materials , in the School of Engineering of Barcelona [58]. Twenty pallet rack
column specimens have been tested aiming to evaluate alternative methods for
testing as a design option. An analytical and numerical method have been followed
to compare the results with the ones gained from experimental tests. Analytical
method have been followed by applying the European Standard EN 1993-1-3
involvilng the effective width determination for each part of the section subject to
compression [59]. The numerical method have been followed by applying finite
element analysis, including non-linear material and geometrical behaviour. The test
specimens were formed with a piece of an upright, whose length was more than three
times greater than the longest dimension of the upright section, and more than five
times the distance between perforations. As an example, experimental collapsed
specimens and deformed shapes from finite element analysis have been given in
Figure 2.24.
Figure 2.24 : Experimental collapsed specimens and deformed shapes from FE

analysis [58].
After experimental, numerical and analytical testing a comparison have been made
between the effective areas and effective centre of gravities. Results of these
comparisons are also shown in Figure 2.25.
39
Figure 2.25 : Effective area and centre of gravity comparisons [58].
In conclusion, two methods are suitable determining the effective area and centre of
gravity with some precautions. When working with the effective net area and
effective net centre of gravity, the analytical method used with Euro Code 3 gives
good results [59]. Finite Element Analysis also gives good results when the
following precautions are considered. The model must reproduce the upright
faithfully and the increments in displacement must be small, and also the force-
displacement curve must show that the point of the maximum force is clearly
reached. Authors elaborate that at the present time, none of the methods can
completely replace experimental testing. Factors such as actual boundary conditions,
geometric imperfections, or residual stresses, for instance, are not usually totally
known; therefore physical testing is necessary. However, the analytical and the
simulation by finite element methods come rather close to the experimental results
which is confirmed by these investigations. Finite Element Analysis seems like most
advantageous method in the design and optimization phase. Using Finite Element
Analysis can cause a decrease in number of tests mandatory in Euro Code by
reproducing most of the factors involved in the problems [58].
An Experimental Investigation of Distortional Buckling of Steel Storage Rack

Columns;
Miquel Casafont, Maria Magdalena, Francesc Roure, Teoman Peköz (2011)
This study includes experimental investigation of steel storage columns under
compression for distortional buckling failure mode [60]. The study mainly focused in
moderately long specimens which are longer than the stub columns used for the
determination of the local buckling strength and short enough to avoid the effects of
global buckling. The objective of the study is to formulate experimental and
analytical approaches to the design of industrial rack columns against distortional
buckling. Four different type of rack cross-sections having medium load carrying
40
capacity produced by European manufacturers are used in analysis and testing. These
investigations made possible to study failure modes combining distortional buckling
and global buckling. Determination of distortional buckling strength from test results
have been gathered using a modification to Rack Manufacturers Institute (RMI)
specification, European Code and Direct Strength Method. Better results are obtained
if the interaction between distortional buckling and global modes is considered when
the strength of a member is determined from experimental tests according to the
American RMI specification. Another point is coming from the results; if this
interaction is not considered, the strength overestimation can be up to about 15% for
the cross-sections analyzed. Another keypoint from paper is an evaluation has been
done for Direct Strenght Method calculations for distortional buckling. The method
includes two important changes for the effect of perforations; usage of the net are of
cross-section and Finite Element Analysis for calculation of determining elastic
buckling loads. The calculated results are in good agreement with experimental
results and showed that DSM equations provide interaction between distortional
buckling and global modes. This paper also forms a start for ongoing and new studies
like the Research on the behavior of upright frames aimed at checking whether the
experimental tests on individual columns are useful for assessing the strength of
complete frames.
Design of Steel Storage Rack Columns via the Direct Strength Method;
Miquel Casafont, Maria Magdalena Pastor, Francesc Roure, Jorde Bonada,
Teoman Peköz (2013)
This paper presents an attempt for prediction of the load carrying capacity of
perforated rack columns by the direct strength method [61]. Investigations are
focused on the prediction of the elastic buckling loads of members with multiple
perforations and the accuracy of the current DSM buckling curves on rack columns.
Currently there is no calculation procedure is adopted by the main design codes of
rack structures in Eurocode. Design is simply based on expreimental physical tests.
Non-linear finite element analysis is most probably the best alternative to
experimental approach. Reduced Thickness formulas given below form another
alternative for the effect of perforations on calculating strength of a cold formed
column. Casafont et al. found these formulas in 2012 [62]. Reduced Thickness can
be used in DSM or EWM calculations instead of thickness for the effect of
perforations on Rack Columns.
41
Reduced Thickness for local buckling;
(2.24)
Reduced Thickness for distortional buckling;
( ) (2.25)
Reduced Thickness for torsional-flexural buckling;
( ) (2.26)
Where the terms are the geometric parameters on a rack column given in Figure 2.26.
Figure 2.26 : Main geometric parameters of the column [61].
Five different rack column types have been used in this study. Elastic buckling load
have been calculated via Finite Strip Method. For the results, the performance of the
distortional buckling and global buckling reduced thickness equations are correct.
Concerning the global buckling loads of columns with large flange holes, some small
42
problems have only arisen. For the accuracy of DSM, authors have tried different
methods and parameters like minimum net area, Finite Element Analysis and
CUFSM for comparison. In determining elastic buckling loads using Finite Element
Method, the results were not good if the failure is governed by local buckling.
Predictions were accurate on the columns have failed in the modes of distortional
buckling and torsional-flexural buckling. In determining elastic buckling loads using
CUFSM, the results are quite parallel and even better for those cross sections with
large web stiffeners.
43
44
3. CODIFIED DESIGN OF COLD FORMED STEEL COLUMNS UNDER
PURE COMPRESSION
The present study is based on the design rules given in American Iron and Steel
Institute S100-2007 standart [16]. The same standart is also called as “North
American Specification for the Design of Cold-Formed Steel Structural Members”.
This chapter consists of an overview of this standart on Concentrically loaded
compression CFS members. Approaches with Effective Width Method and Direct
Strength Method are discussed by example solutions.
3.1 Review of American Iron and Steel Institute S100-2007 Standart
Section-C of American Iron and Steel Institute S100-2007 (AISI S100-2007)

Standart describes the limitations and calculations for Members of Cold-Formed
Structural Steel family[16]. C4-Concentrically Loaded Compression Members part is
related with the scope of this thesis. Allowable Stress Design (ASD) and Load and
Resistance Factor Design (LRFD) considerations and methods are also two different
approaches proposed for the design of these members which are described in the
standart. In this thesis, nominal column capacities are considered instead of design
strength. There are two main methods to calculate ultimate capacity of the
concentrically loaded compression members which are Effective Width Method and
Direct Strength Method. The standart and its main sections describe the limitations
and calculations for traditional method (EWM). In Appendix 1, the standart describes
DSM method and its provisions. Members and Column Design part in Appendix 1 is
an alternative approach for traditional method.
3.1.1. Effective width method for concentric axially loaded compression

members
Studies have shown that cross-section of a thin walled steel member may not be fully
effective under compression. Basically, some parts of the cross-section resists the
load instead 100% of the member cross-sectional area. The reduced area is used
45
during the calculation of maximum strength of the section which is also called
Effective Area; can be calculated by multiplying the thickness of the member with
the Effective Width. The standart aims to calculate the available axial strength which
shall be smaller of the values obtained from section C4. For usage of the effective
widths in the calculations; standart also describes the calculation of effective widths
of elements under B Section.
3.1.1.1 The nominal axial strength calculations
The nominal axial strength which is also called compressive resistance of a

compression member, Pn , can be calculated in accordance with C4.1 section of
related standart.
(3.1)
Where Ae is the effective area calculated at stress Fn. Effective area shall be
calculated in accordance with section B of the related standart. Fn shall be calculated
as follows depending on slenderness factor, :
For ;
( ) (3.2)
For ;
[ ] (3.3)
Where;
(3.4)
√
And Fe is the least of the applicable elastic flexural, torsional and flexural-torsional
buckling stress determined in accordance with section C4.1.1-5 of the related
standart.
For any sections that can be shown not to be subjected to torsional or flexural-
torsional buckling, the elastic flexural buckling stress, Fe , shall be found as follows;
46
(3.5)
Where;
E is the modulus of elasticity of steel. K is the effective length factor. L is laterally

unbraced length of member and r is the radius of gyration of full unreduced cross-
section about axis of buckling. In frames where lateral stability is provided by
diagonal bracing, the effective length factor, K, for compression members that do not
depend upon their own bending stiffness for lateral stability of the frame shall be
taken as unity, unless analysis shows that a smaller value is suitable. Otherwise, the
effective length, KL, of the compression members shall be found by rational method
and can not be less than the actual unbraced length.
For singly-symmetric sections subject to flexural-torsional buckling, Fe can be taken

as the smaller value obtained from Eq. (3.5) or (3.6).
[ √ ] (3.6)
Where;
(3.7)
x0 is the distance from shear center to centroid along principal x-axis, taken as
negative. r0 is the polar radius of gyration of cross section about shear center
calculated from E.q.(3.8). rx and ry are the radii of gyration of cross-section about
centrodial principal axes.
√ (3.8)
(3.9)
Torsional buckling stress can be calculated as follows:
47
[ ] (3.10)
G is the shear modulus, J is Saint-Venant torsional constant of cross-section, Cw is

torsional warping constant of cross-section, Kt is effective length factors for twisting,
Lt is the unbraced length of member for twisting, Kx is effective length factor for
buckling about x-axis, Lx is the unbraced length of compression member for for
bending about x-axis and A is full unreduced cross-sectional area of member.
An alternative also conservative estimation Fe shall be permitted to be calculated as

in Eq. (3.11).
(3.11)
For singly-symmetric sections, the x-axis shall be selected as the axis of symmetry.
For doubly-symmetric sections subject to torsional buckling, Fe shall be taken as the
smaller value obtained from eq. (3.5) and . For non-symmeytic sections, Fe
shall be determined by rational analysis.
3.1.1.2 Effective area calculations
For the calculation of nominal strength of a member given in eq. (3.1), effective area,
Ae , should also be calculated. Following sections of related standart describes the
methods for calculating effective areas:
B2 – Effective Widths of Stiffened Elements
B3 – Effective Widths of Unstiffened Elements
B4 – Effective Width of Uniformly Compressed Elements with a Simple Lip Edge

Stiffener
B5 – Effective Widths of Stiffened Elements with Single or Multiple Intermediate

Stiffeners or Edge Stiffened Elements with Intermediate Stiffeners
Since this thesis mainly focuses on DSM calculations, only B2 and B4 sections will
be discussed covering general remarks and methods for effective area calculations.
Example solution is also a suitable member for this section.
48
Section B2 describes the effective width calculations for stiffened element which is
also used in general. b, the effective width shall be calculated as follows:
b =w when (3.12)
b = when (3.13)
Where;
w = Flat width as shown in Figure 3.1.

ρ = Local Reduction factor, given in eq. (3.14).
(3.14)
= Slenderness factor, given in eq. (3.15).
(3.15)
√
f = Stress in compression element. In flexural members, when the initial

yielding is in compression element is considered, f = fy.
Fcr = Plate elastic buckling stress
( ) (3.16)
k = Plate buckling coefficient. It can be taken as “4” for stiffened elements

supported by a web on each longitudinal edge. It’s also shown as Case (a) in Figure
2.10.
E = Modulus of elasticity of steel
t = Thickness of the section
μ = Poisson’s ratio of steel
For compression members, f is also taken equal to nominal buckling stress, Fn.
49
Figure 3.1 : Stiffened elements [16].
The effective widths of uniformly compressed elements with a simple edge stiffener
shall be calculated in accordance with for Strength Determination and for
Serviceability Determination calculations according to strength determination shall
be as follows;
For w/t ≤ 0.328S:
Ia =0 (no edge stiffener is needed. (3.17)
b =w (effective design width) (3.18)
b1 = b2 = w/2 (portions of effective design widths as shown in fig.3.1) (3.19)
ds = d’s (reduced effective width of stiffener shown in fig.3.1) (3.20)
For w/t > 0.328S:
b1 = (b/2) (RI) (3.21)

b2 = b2 = w/2 (3.22)
ds = d’s (RI) (3.23)
where;
S = √ (3.24)
w = Flat dimensions defined in Figure 3.1 and Figure 3.2.
Ia = Adequate moment of inertia of stiffener, so that each component element
will behave as a stiffened element, given in eq. (3.25).
[ ] [ ]
(3.25)
d’s = Effective width of stiffener seen in Figure 3.1.

(RI) = (3.26)
Is = Moment of inertia of full section of stiffener about its own centrodial axis
50
parallel to element to be stiffened. For edge stiffeners, the round corner between
stiffener and element to be stiffened is not considered as a part of the stiffener, given
in eq. (3.27).
(3.27)
The effective width, b in equations (3.21) and (3.22) shall be calculated in

accordance with previously described section B2 with eq. (3.16), plate buckling
coefficient, k, as given in table 3.1.
Table 3.1 : Determination of plate buckling coefficient, k [16].

Simple Lip Edge Stiffener ( )
( )
Where;
( ) (3.28)
Figure 3.2 shows the definitions of dimensional variables on elements with a simple
lip edge stiffener;
Figure 3.2 : Elements with simple lip edge stiffener [16].

51
3.1.1.3 Distortional buckling strength calculations
C4.2 section of AISI S100-2007 standart describes the calculations of distortional

buckling strength for concentrically loaded compression members [16]. The
provisions of this section shall apply to I, Z, C, hat and any other open-sections
which will be affected by distortional buckling. Nominal axial strength, Pn will be
calculated for distortional buckling as one found for general in previous section and
will be compared with previous strength.
For λd ≤ 0.561;
Pn = Py (3.29)
For λd > 0.561;
( ) (3.30)
( )( )
Where;
= Slenderness factor for distortional buckling, given in eq. (3.31).
√ (3.31)
Pn is the nominal axial strength and Py is the member yield strength given in Eq.
(3.32).
(3.32)
Ag is the gross area of element including stiffeners. Fy is the yield stress of steel. Pcrd
is the distortional buckling load given in eq. (3.33).
(3.33)
Fd is elastic distortional buckling stress. For C and Z sections that have no rotational
restraint of flange and that are within the dimensional limits provided below, a
conservative prediction of distortional buckling stress can be made with eq. (3.34).
Geometric definitions are shown on Figure 3.2 and Figure 3.3.
52
Figure 3.3 : Out-to-out dimensions of webs and stiffened elements [16].
h0 is out-to-out web depth, b0 is out-to-out flange width, D is out-to-out lip

dimension, t is base steel thickness and is the lip angle.
Following dimensional limits shall apply:
- ,
- ,
- ,
- ,
- , and
( ) (3.34)
Where;
α is a value that accounts for the benefit of an unbraced length, Lm , shorter than Lcr ,
but can be conservatively taken as “1”.
For (3.35)
For (3.36)
Lm is the distance between discrete restraints that restrict distortional buckling,

basically unbraced length. Lcr is the critical unbraced length of distortional buckling,
given in eq. (3.37).
53
( ) (3.37)
kd is the plate buckling coefficient for distortional buckling, given in eq. (3.38).
( ) (3.38)
A rational elastic buckling analysis that considers distortional buckling shall be

permitted to be used with these expressions given.
3.1.2. Example solution of a c-channel column with a simple lip edge stiffener
via effective width method
Figure 3.4 : C-channel column section with simple lip edge example.
54
Table 3.2 : Section properties of example solution.
Section Properties Calculated Section Properties
t = 1.5 mm A = 588.21 mm2
b = 80 mm rx = 79.03 mm
h = 200 mm ry = 30.02 mm
fy = 235 MPa J = 441.16 mm4
E = 200000 MPa x0 = -61.96 mm
G = 80000 MPa r0 = 104.81 mm
inner curvature radius = 6 mm Cw = 4487970571.77 mm6
Section is pin connected from both ends, concentrically loaded and effective length
factors of this column are;
Kx = Ky = Kt = 1.0
Nominal Capacity of this cold-formed steel column will be calculated using Effective
Width Method and distortional buckling calculations according to AISI S100-2007
standart [16].
Nominal Section Strength

- Check for flexural buckling;
Since , flexural buckling will control about y-axis
Using eq. (3.5);
(3.5)
( )
- Check for flexural- torsional buckling, eq. (3.6) will be used. For the terms used in
eq. (3.6); equations (3.7) , (3.9) and (3.10) will be used.
( ) ( ) (3.7)
(3.9)
55
(3.10)
[ ]
[ ]
(3.6)
[ √ ]
676.49 MPa < 907.43 MPa therefore, flexural-torsional buckling governs.
Fe = 676.49 Mpa will be used in eq. (3.4) to calculate the slenderness factor:
(3.4)
√ √
Therefore in Eq. (3.2) nominal buckling stress, Fn , will be calculated as follows for
;
( ) ( ) (3.2)
For calculating the nominal section capacity Pn , Effective area of the section should
be calculated and multiplied by nominal buckling stress, Fn according to section C4.1
of related standart.
Effective Area Calculations (for Fn = 203.14 MPa)

- Calculations for compression flanges
f = 203.14 MPa, E = 200000 Mpa, t = 1.5 mm

w = 80 – 2 (6+1.5) = 65 mm (clear width of flange)
From Eq. (3.24);
S= √ = √
56
0.328 x S = 0.328 x 40.16 = 13.17 and w / t = 65 / 1.5 = 43.33
Since , check on effective width of flange:
[ ] [ ] (3.25)
[ ] [ ]
so should be used.
Calculating Is of the stiffener with eq. (3.27) (d is the clear width of stiffener =
25 - 6 - 1.5 = 17.5 mm)
(3.27)
From Eq. (3.26); (RI) =

so RI = 1
From Eq. (3.24);
( ) (3.24)
( )
So n = 1/3. From table 3.1 plate buckling coefficient can be calculated as;
( )
( ) OK!
So k = 3.33
Calculating plate elastic buckling stress using Eq. (3.16)
57
(3.16)
( )
( )
Slenderness factor will be found using Eq. (3.15)
(3.15)
√ √
According to Eq. (3.13) b= when (3.13)
Flange is subjected to local buckling.
Effective reduction factor will be found using Eq. (3.14)
(3.14)
Effective width of the flange is
- Calculations for stiffener lips
f = 203.14 Mpa k = 0.43 (from fig.2.10 case c)

clear width w = d = 17.5 mm From Eq. (3.16);
( )
Using Eq. (3.15);
√ √
According to Eq. (3.13) b= when (3.13)
Web is subjected to local buckling.

Effective reduction factor will be found using Eq. (3.14);
(3.14)
58
Effective width of the flange is
- Calculations for Total Effective Area
Table 3.3 : Effective widths of section’s elements.

Element Effective Widths (mm)
Top Flange 59.09
Bottom Flange 59.09
Web 79.92
Top Inside Corner 10.6
Bottom Inside Corner 10.6
Top Outside Corner 10.6
Bottom Outside Corner 10.6
Top Lip 17.5
Bottom Lip 17.5
Total Length 275.5
Effective area of the section can be obtained by multiplying total effective width with
the thickness:
Computation of Pn for local buckling

Effective area, Ae , of the section is calculated 413.25 mm2 under 203.14 MPa
nominal buckling stress, Fn.
From Eq. (3.1) nominal capacity of the section is;
(3.1)
Calculations for distortional buckling strength

Distortional buckling strength is calculated in accordance with AISI 100-2007
Section C4.2
Controlling dimensional limits for calculations
- , OK!
- , OK!
- , OK!
- , OK!
- , and OK!
- OK!
59
Controlling critical unbraced length of distortional buckling using Eq. (3.37)
(3.37)
( )
( )
OK!
In Eq. (3.35) ;
For (3.35)
so
Calculating the distortional plate buckling coefficient by using Eq. (3.38);
( ) (3.38)
( ) (3.33)
OK!
Calculating the elastic distortional buckling stress using Eq. (3.29);
( ) (3.34)
( )
Calculating member yield strength using Eq. (3.27);
(3.32)
Calculating distortional buckling load using Eq. (3.33);
(3.33)
60
Calculating slenderness factor for distortional buckling using Eq. (3.31);
√
(3.31)
√
Since Eq. (3.30) will be used for calculating nominal distortional

buckling strength;
( ) (3.30)
( )( )
( ( ) )( )
Since Pn obtained from distortional buckling strength calculations is critical than the
one obtained from local buckling strength calculations, the nominal capacity of this
section is:
From Distortional buckling strength
From Local buckling strength
3.1.3 Direct strength method for concentric axially loaded compression

members
In Appendix 1 of AISI 100-2007 standart, provisions, descriptions and details of
Direct Strength Method is given. Since this thesis is focused on this method, codified
view for DSM will be summarized in more detail.
The provisions of this appendix shall be permitted to be used to determine the

nominal axial (Pn) and flexural (Mn) strengths of cold formed steel members. Section
1.2.1 of the standart presents a method applicable to all cold-formed steel columns
meeting the geometric and material limitations.
These limitations are given under “pre-qualified columns” chart in the standart [16].
Basically, this means that DSM method is applicable for unperforated columns
falling within these limits as a result of many studies and investigations. However,
more studies and investigations are continuing to improve this pre-qualified column
61
chart for DSM method. This thesis also aims to bring a different perspective to this
pre-qualified column chart by trial of different sections standing outside of this
geometrical limitations. This pre-qualified column chart is given below in Figure 3.5.
Figure 3.5 : Limits for pre-qualified columns [16].
62
3.1.3.1 Determination of elastic buckling loads
Section 1.1.2 of Appendix 1 describes the determination of elastic buckling loads in
AISI 100-2007 standart. For columns, this includes the local, distortional and overall
(global) buckling loads which are; Pcrℓ Pcrd and Pcre . It should be noted that in some
cases, for a given column all three modes may not exist. These non-existent mode
shall be ignored in such cases. These elastic buckling loads shall be calculated using
software programs like CUFSM which uses Finite Strip Method in analysis. Finite
element analysis is also another option to obtain these critical loads. Software
programs like ABAQUS, ANSYS etc. can be used for finite element analysis.
3.1.3.2 Column analysis

The nominal axial strength, Pn, shall be the minimum of nominal axial strengths
obtained from three major checks. These checks are for;
- Flexural, Torsional, or Flexural-Torsional Buckling
- Local Buckling
- Distortional Buckling
Nominal axial strength values should be calculated from mentioned checks are Pne ,
Pnℓ , Pnd respectively.
Flexural, torsional, or flexural-torsional buckling

The nominal axial strength, Pne , for flexural, torsional, or flexural-torsional buckling
shall be calculated in accordance with the following formulas and method;
For
( ) (3.39)
For
( ) (3.40)
Where , is the global slenderness factor and can be calculated with Eq. (3.41);
√ (3.41)
63
Py is the member yield strength previously given in Eq. (3.32);
(3.32)
Ag is the gross area of the section. Fy is the yield stress of steel.
Pcre is the minimum of the critical elastic column buckling load in flexural, torsional
or flexural-torsional buckling determined by analysis in accordance with section
3.1.3.1.
Local buckling
The nominal axial strength, Pnℓ , for local buckling shall be calculated in accordance
with the following formulas and method;
For
(3.42)
For
[ ( ) ]( ) (3.43)
Where , is the local slenderness factor and can be calculated with Eq. (3.44);
√ (3.44)
Pne is the nominal axial strength value obtained from Flexural, Torsional, or Flexural-
Torsional Buckling section using eq. 3.39 or 3.40
Pcrℓ is the critical elastic local column buckling load determined by analysis in
accordance with section 3.1.3.1.
Distortional buckling
The nominal axial strength, Pnd , for distortional buckling shall be calculated in
accordance with the following formulas and method;
For
(3.45)
64
For
( ) (3.46)
[ ]( )
Where , is the distortional slenderness factor and can be calculated with Eq.
(3.47);
√ (3.47)
Py is the member yield strength given in Eq. (3.32)
Pcrd is the critical elastic distortional column buckling load determined by analysis in
accordance with section 3.1.3.1.
3.1.3.3 Summary and design curves of direct strength method

As a summary, DSM method will be explained shortly in this section. Also Design
Curves for Direct Strength Method will be given.
To calculate the capacity of a cold-formed steel member with DSM, the elastic
buckling properties of a CFS cross-section are obtained from an elastic buckling
curve. The curve should be generated using a software progra like CUFSM
performing eigenbuckling analyses over a range of buckled half-wavelengths with
finite strip method. Local buckling occurs as plate buckling of individual slender
elements in whole cross-section. Only open cross-sections such as C and hat section
will be faced to existence of distortional buckling where the compressed flanges
buckle inward or outward along the length of the member. Global buckling defines
buckling of the full member at long half-wavelengths which is also known as Euler
buckling. Global buckling includes both flexural and flexural-torsional effects. For
columns, the critical elastic buckling loads related with global, local, and distortional
buckling (Pcre , Pcrℓ , Pcrd) can be obtained directly from the elastic buckling curve of
that member. Then, these values are used to find that member’s slenderness factors
( ) defining the sensitivity to each type of buckling at failure. Then the
nominal strengths (Pne , Pnℓ , Pnd) shall be obtained by inserting the slenderness
factors into DSM equations. The minimum of the global, local and distortional
65
nominal strengths will be the actual nominal strength of the member. For design of
steel columns using Direct Strength Method and as an explanation of DSM
calculations, three empirical design curves are given in Figures 3.6 to 3.8.
Figure 3.6 : DSM global buckling failure design curve and equations [63].
Figure 3.7 : DSM local buckling failure design curve and equations [63].
66
Figure 3.8 : DSM distortional buckling failure design curve and equations [63].
3.1.4 Example solution of a c-channel column with a simple lip edge stiffener via
direct strength method
Same example will be used with DSM which have been previously solved with
Effective Width Method. Geometric details of the example can be gathered from
Figure 3.4 and Table 3.2.
Section is pin connected from both ends, concentrically loaded and effective length
factors of this column are; Kx = Ky = Kt = 1.0
Dimension controls for DSM using “Limits for Pre-Qualified Columns Chart” given
in Figure 3.5;
For all C Sections:
OK!
OK!
OK!
OK!
OK!
OK!
67
[ ]
OK!
Example column is in the Pre-qualified column limits. This means that nominal
section strength can be calculated using Direct Strength Method.
Determination of Critical Elastic Buckling Loads Using CUFSM Software

Centerline dimensions of the section of this example column has been built in
CUFSM v4.5. shown in Figure 3.9. Simple support boundary conditions and the
yield load which is based on maximum yield stress (235 MPa) have been set on the
section shown in Figure 3.10.
Figure 3.9 : Centerline dimensions in CUFSM.
Material properties and boundary conditions which is simply supported has been set
in CUFSM which are shown in Figures 3.11 and 3.12.
68
Figure 3.10 : Yield load calculation.
Figure 3.11 : Configuration of material properties.
69
Figure 3.12 : Boundary condition selection.
The load factor on the elastic buckling curve of the section represents the ratios of
Pcrℓ/Py , Pcrd/Py , and Pcre/Py according to the type of the buckled shape (local,
distortional and global) on the minimas of the curve. Results of the finite strip
analysis conducted in CUFSM are given in Figures 3.13 to 3.15. Load factors for
local buckling, distortional buckling and flexural-torsional (global) buckling are
obtained in order to calculate critical elastic column buckling loads.
It has been seen that load factor for flexural-torsional buckling mode caught on the
2nd mode in 1400 mm of half-wave length for this section. Member yield strength
will be calculated using Eq. (3.32);
(3.32)
According to the results of CUFSM;
70
Figure 3.13 : Result of local buckling in CUFSM.
Figure 3.14 : Result of distortional buckling in CUFSM.
71
Figure 3.15 : Result of flexural-torsional buckling in CUFSM.
Calculations for flexural-torsional buckling

The global slenderness factor is calculated with Eq. (3.41);
√ √ (3.41)
The nominal axial strength, Pne , for flexural-torsional buckling is calculated with eq.
(3.39);
For
( ) ( ) (3.39)
Pne = 114.22 kN
Calculations for local buckling

The local slenderness factor is calculated with eq. (3.44);
√ √ (3.44)
The nominal axial strength, Pnℓ , for local buckling is calculated with eq. (3.43);
72
For
[ ( ) ]( ) (3.43)
[ ( ) ]( )
Pnℓ = 66.70 kN
Calculations for distortional buckling

The distortional slenderness factor is calculated with eq. (3.47);
√ √ (3.47)
The nominal axial strength, Pnd , for distortional buckling is calculated with eq.
(3.46);
For
( ) (3.46)
[ ]( )
[ ( ) ]( )
Pnd = 88.53 kN
Nominal axial strength of section

From the calculated nominal axial strengths, it can be seen that the minimum value
is Pnℓ among others. Nominal axial strength of this column is found Pn = 66.70 kN
using Direct Strength Method.
73
74
4. INVESTIGATION OF RACK COLUMN BEHAVIOUR THROUGH
DIRECT STRENGTH METHOD
This study consists of investigation on nominal section strength of various rack

columns through a newly developed analytical method which is known as Direct
Strength Method.
120 column models having different lip configurations and column lengths are used
in this study. Half of these specimens have web perforations with same cross-section
geometries.
In Figure 4.1, general specimen geometric cross-sectional details are shown. All
cross-sections have same b0 and h0 values which are 50 mm and 140 mm. Lip length
starts from 0 and increases with 5 mm intervals. After reaching the 20 mm length of
first lip (as simple lip section) second lip continues inwards or outwards which
makes the section a complex lipped section. After reaching the 30 mm length of
second lip, third lip starts and ends at 10 mm length. Cross-sections are generated
using this 5 mm interval lip length. These specimens’ geometric properties are all
within the prequalified column cross-section limits for DSM given in Figure 3.5.
Specimens have been investigated having three different column lengths with same
cross-section properties which are L1 = 500 mm, L2 = 1000 mm, L3 = 1500 mm.
Yield strength of the material is assumed to be 235 MPa and thickness is assumed 2
mm for all column models. Elastic Modulus of steel (E) has been taken as 200000
MPa and Shear Modulus (G) has been taken as 80000 MPa. Columns are assumed to
be compressed between pin supports which makes the chosen boundary conditions
simply supported. Notation explanation of specimen types with three different
specimen examples are given in Figure 4.2.
75
Figure 4.1 : General geometric details of specimen cross-sections.
76
Figure 4.2 : Explanation of specimen’s notation through its properties.
As a final explanation of these chosen specimens, Lip starts and continues as D with
an interval of 5 mm . Then moves inwards or outwards as D2 again in 5mm intervals.
Finally after last turn, it continues as D3 again in 5 mm interval. Maximum lip length
limits have been set in accordance with manufactured sections in Turkey and also
within the limits of AISI Prequalified column chart for DSM given in Figure 3.5:
Maximum limit of D is 20 mm
Maximum limit of D2 is 30 mm
Maximum limit of D3 is 10 mm
For instance, if a specimen has D3.10 on its notation that means its D3 is 10 mm, D2
is 30 mm and D is 20 mm. In total, calculations of 60 perforated and 60 unperforated
column models have been done.
4.1 Analysis of Unperforated Column Models
60 unperforated column models have been analyzed in this study. Geometric details
of these specimens are given in tables 4.1 to 4.3 based on the length of the columns.
77
Table 4.1 : Geometric properties of unperforated specimens having 500 mm column
length.
t h0 b0 D D2 D3 Section Area Column Length
2
Column (mm) (mm) (mm) (mm) (mm) (mm) (mm ) (mm)
D.5.NP.L500 2 140 50 5 - - 484 500
D.10.NP.L500 2 140 50 10 - - 504 500
D.15.NP.L500 2 140 50 15 - - 524 500
D.20.NP.L500 2 140 50 20 - - 544 500
O.D2.5.NP.L500 2 140 50 20 5 - 556 500
O.D2.10.NP.L500 2 140 50 20 10 - 576 500
O.D2.15.NP.L500 2 140 50 20 15 - 596 500
O.D2.20.NP.L500 2 140 50 20 20 - 616 500
O.D2.25.NP.L500 2 140 50 20 25 - 636 500
O.D2.30.NP.L500 2 140 50 20 30 - 656 500
O.D3.5.NP.L500 2 140 50 20 30 5 668 500
O.D3.10.NP.L500 2 140 50 20 30 10 688 500
I.D2.5.NP.L500 2 140 50 20 5 - 556 500
I.D2.10.NP.L500 2 140 50 20 10 - 576 500
I.D2.15.NP.L500 2 140 50 20 15 - 596 500
I.D2.20.NP.L500 2 140 50 20 20 - 616 500
I.D2.25.NP.L500 2 140 50 20 25 - 636 500
I.D2.30.NP.L500 2 140 50 20 30 - 656 500
I.D3.5.NP.L500 2 140 50 20 30 5 668 500
I.D3.10.NP.L500 2 140 50 20 30 10 688 500

column length.
Column (mm) (mm) (mm) (mm) (mm) (mm) (mm2) (mm)
D.5.NP.L1000 2 140 50 5 - - 484 1000
D.10.NP.L1000 2 140 50 10 - - 504 1000
D.15.NP.L1000 2 140 50 15 - - 524 1000
D.20.NP.L1000 2 140 50 20 - - 544 1000
O.D2.5.NP.L1000 2 140 50 20 5 - 556 1000
O.D2.10.NP.L1000 2 140 50 20 10 - 576 1000
O.D2.15.NP.L1000 2 140 50 20 15 - 596 1000
O.D2.20.NP.L1000 2 140 50 20 20 - 616 1000
O.D2.25.NP.L1000 2 140 50 20 25 - 636 1000
O.D2.30.NP.L1000 2 140 50 20 30 - 656 1000
O.D3.5.NP.L1000 2 140 50 20 30 5 668 1000
O.D3.10.NP.L1000 2 140 50 20 30 10 688 1000
I.D2.5.NP.L1000 2 140 50 20 5 - 556 1000
I.D2.10.NP.L1000 2 140 50 20 10 - 576 1000
I.D2.15.NP.L1000 2 140 50 20 15 - 596 1000
I.D2.20.NP.L1000 2 140 50 20 20 - 616 1000
I.D2.25.NP.L1000 2 140 50 20 25 - 636 1000
I.D2.30.NP.L1000 2 140 50 20 30 - 656 1000
I.D3.5.NP.L1000 2 140 50 20 30 5 668 1000
I.D3.10.NP.L1000 2 140 50 20 30 10 688 1000
78
column length.

2
Column (mm) (mm) (mm) (mm) (mm) (mm) (mm ) (mm)
D.5.NP.L1500 2 140 50 5 - - 484 1500
D.10.NP.L1500 2 140 50 10 - - 504 1500
D.15.NP.L1500 2 140 50 15 - - 524 1500
D.20.NP.L1500 2 140 50 20 - - 544 1500
O.D2.5.NP.L1500 2 140 50 20 5 - 556 1500
O.D2.10.NP.L1500 2 140 50 20 10 - 576 1500
O.D2.15.NP.L1500 2 140 50 20 15 - 596 1500
O.D2.20.NP.L1500 2 140 50 20 20 - 616 1500
O.D2.25.NP.L1500 2 140 50 20 25 - 636 1500
O.D2.30.NP.L1500 2 140 50 20 30 - 656 1500
O.D3.5.NP.L1500 2 140 50 20 30 5 668 1500
O.D3.10.NP.L1500 2 140 50 20 30 10 688 1500
I.D2.5.NP.L1500 2 140 50 20 5 - 556 1500
I.D2.10.NP.L1500 2 140 50 20 10 - 576 1500
I.D2.15.NP.L1500 2 140 50 20 15 - 596 1500
I.D2.20.NP.L1500 2 140 50 20 20 - 616 1500
I.D2.25.NP.L1500 2 140 50 20 25 - 636 1500
I.D2.30.NP.L1500 2 140 50 20 30 - 656 1500
I.D3.5.NP.L1500 2 140 50 20 30 5 668 1500
I.D3.10.NP.L1500 2 140 50 20 30 10 688 1500
4.1.1 Analytical study on unperforated members using direct strength method

A study including construction of elastic buckling curves with CUFSM Software
and DSM calculations has been done to calculate the nominal section strength of
these specimens. Critical elastic column buckling loads of these various sections are
obtained from CUFSM software. Tables 4.4, 4.5 and 4.6 include the results of
critical elastic buckling loads obtained by CUFSM. For models with non-existent
buckling modes; a high value of elastic critical buckling load is intentionally
assumed to calculate the slenderness values. The non-existent buckling modes are
indicated with N.A in relevant cells of tables 4.4 to 4.6.
Direct Strength Method is used after obtaining the critical elastic buckling loads.
Nominal strengths for local, distortional and global buckling are calculated with
DSM equations. Then, the nominal strength for all sections is determined. Results are
given in tables 4.7, 4.8 and 4.9. Controlling mode indicates which buckling mode
governs the behaviour of the member.
79
Table 4.4 : Critical elastic buckling loads obtained from CUFSM analysis for unperforated specimens having 500 mm column length.
Yield Load (kN) CUFSM RESULTS (LOAD FACTORS) Critical Elastic Buckling Loads (kN)
Column Py Pcrℓ/Py hw. length (mm) Pcrd/Py hw. length (mm) Pcre/Py hw. length (mm) Pcrℓ Pcrd Pcre
D.5.NP.L500 113,74 1,651 50 0,737 200 N.A. - 187,78 83,83 N.A.
D.10.NP.L500 118,44 0,902 120 0,933 140 N.A. - 106,83 110,50 N.A.
D.15.NP.L500 123,14 0,918 100 1,349 400 N.A. - 113,04 166,12 N.A.
D.20.NP.L500 127,84 0,921 100 1,569 500 N.A. - 117,74 200,58 N.A.
O.D2.5.NP.L500 131,60 0,924 100 1,616 500 N.A. - 121,60 212,67 N.A.
O.D2.10.NP.L500 136,30 0,929 100 1,544 500 N.A. - 126,62 210,45 N.A.
O.D2.15.NP.L500 141,00 0,930 100 1,421 500 N.A. - 131,13 200,36 N.A.
O.D2.20.NP.L500 145,70 0,930 100 1,299 500 N.A. - 135,50 189,26 N.A.
O.D2.25.NP.L500 150,40 0,930 100 1,203 500 N.A. - 139,87 180,93 N.A.
O.D2.30.NP.L500 155,10 0,930 100 1,136 500 N.A. - 144,24 176,19 N.A.
O.D3.5.NP.L500 158,86 0,930 100 1,142 500 N.A. - 147,74 181,42 N.A.
O.D3.10.NP.L500 163,56 0,930 100 1,298 500 N.A. - 152,11 212,30 N.A.
I.D2.5.NP.L500 131,60 0,922 100 1,699 500 N.A. - 121,34 223,59 N.A.
I.D2.10.NP.L500 136,30 0,924 100 1,871 500 N.A. - 125,94 255,02 N.A.
I.D2.15.NP.L500 141,00 0,926 100 2,061 500 N.A. - 130,57 290,60 N.A.
I.D2.20.NP.L500 145,70 0,927 100 2,263 500 N.A. - 135,06 329,72 N.A.
I.D2.25.NP.L500 150,40 0,928 100 2,463 500 N.A. - 139,57 370,44 N.A.
I.D2.30.NP.L500 155,10 0,929 100 2,641 500 N.A. - 144,09 409,62 N.A.
I.D3.5.NP.L500 158,86 0,929 100 2,517 400 N.A. - 147,58 399,85 N.A.
I.D3.10.NP.L500 163,56 0,929 100 2,449 400 N.A. - 151,95 400,56 N.A.
* hw length = halfwave length
80
D.5.NP.L1000 113,74 1,651 50 0,737 200 1,843 1000 187,78 83,83 209,62
D.10.NP.L1000 118,44 0,902 120 0,933 140 2,001 1000 106,83 110,50 237,00
D.15.NP.L1000 123,14 0,918 100 1,349 400 2,031 1000 113,04 166,12 250,10
D.20.NP.L1000 127,84 0,921 100 1,569 500 1,971 1000 117,74 200,58 251,97
O.D2.5.NP.L1000 131,61 0,924 100 1,607 600 1,901 1000 121,61 211,50 250,19
O.D2.10.NP.L1000 136,31 0,929 100 1,502 600 1,817 1000 126,63 204,74 247,68
O.D2.15.NP.L1000 141,00 0,930 100 1,374 600 1,735 1000 131,13 193,73 244,64
O.D2.20.NP.L1000 145,70 0,931 100 1,255 600 N.A. - 135,65 182,85 N.A.
O.D2.25.NP.L1000 150,40 0,931 100 1,157 600 N.A. - 140,02 174,01 N.A.
O.D2.30.NP.L1000 155,10 0,930 100 1,084 600 N.A. - 144,24 168,13 N.A.
O.D3.5.NP.L1000 158,86 0,930 100 1,072 600 N.A. - 147,74 170,30 N.A.
O.D3.10.NP.L1000 163,56 0,930 100 1,155 700 N.A. - 152,11 188,91 N.A.
I.D2.5.NP.L1000 131,60 0,922 100 1,668 600 1,895 1000 121,34 219,51 249,38
I.D2.10.NP.L1000 136,30 0,924 100 1,729 700 1,805 1000 125,94 235,66 246,02
I.D2.15.NP.L1000 141,00 0,926 100 1,778 700 1,735 1000 130,57 250,70 244,64
I.D2.20.NP.L1000 145,70 0,927 100 1,693 1000 N.A. - 135,06 246,67 N.A.
I.D2.25.NP.L1000 150,40 0,928 100 1,678 1000 N.A. - 139,57 252,37 N.A.
I.D2.30.NP.L1000 155,10 0,929 100 1,692 1000 N.A. - 144,09 262,43 N.A.
I.D3.5.NP.L1000 158,86 0,929 100 1,708 1000 N.A. - 147,58 271,33 N.A.
I.D3.10.NP.L1000 163,56 0,929 100 1,715 1000 N.A. - 151,95 280,51 N.A.
81
D.5.NP.L1500 113,74 1,651 50 0,737 200 0,909 1500 187,78 83,83 103,39
D.10.NP.L1500 118,44 0,902 120 0,933 140 1,009 1500 106,83 110,50 119,51
D.15.NP.L1500 123,14 0,918 100 1,349 400 1,119 1500 113,04 166,12 137,79
D.20.NP.L1500 127,84 0,921 100 1,569 500 1,242 1500 117,74 200,58 158,78
O.D2.5.NP.L1500 131,61 0,924 100 1,607 600 1,294 1500 121,61 211,50 170,30
O.D2.10.NP.L1500 136,31 0,929 100 1,502 600 1,369 1500 126,63 204,74 186,61
O.D2.15.NP.L1500 141,00 0,930 100 1,374 600 1,457 1500 131,13 193,73 205,44
O.D2.20.NP.L1500 145,70 0,931 100 1,255 600 1,554 1500 135,65 182,85 226,42
O.D2.25.NP.L1500 150,40 0,931 100 1,157 600 1,646 1500 140,02 174,01 247,56
O.D2.30.NP.L1500 155,10 0,930 100 1,084 600 1,670 1500 144,24 168,13 259,02
O.D3.5.NP.L1500 158,86 0,930 100 1,072 600 1,655 1500 147,74 170,30 262,91
O.D3.10.NP.L1500 163,56 0,930 100 1,155 700 1,588 1500 152,11 188,91 259,73
I.D2.5.NP.L1500 131,60 0,922 100 1,668 600 1,271 1500 121,34 219,51 167,26
I.D2.10.NP.L1500 136,30 0,924 100 1,729 700 1,269 1500 125,94 235,66 172,96
I.D2.15.NP.L1500 141,00 0,926 100 1,778 700 1,242 1500 130,57 250,70 175,12
I.D2.20.NP.L1500 145,70 0,927 100 1,693 1000 1,204 1500 135,06 246,67 175,42
I.D2.25.NP.L1500 150,40 0,928 100 1,678 1000 1,165 1500 139,57 252,37 175,22
I.D2.30.NP.L1500 155,10 0,929 100 1,692 1000 1,130 1500 144,09 262,43 175,26
I.D3.5.NP.L1500 158,86 0,929 100 1,708 1000 1,106 1500 147,58 271,33 175,70
I.D3.10.NP.L1500 163,56 0,929 100 1,715 1000 1,079 1500 151,95 280,51 176,48
82
column length.
DSM Calculations Results (kN) Nominal Strength (kN)

Column Pne Pnℓ Pnd Pn Controlling Mode
D.5.NP.L500 113,69 113,50 74,99 74,99 Distortional controls
D.15.NP.L500 123,09 101,72 103,30 101,72 Local-Global controls
O.D2.5.NP.L500 131,55 108,94 117,00 108,94 Local-Global controls
O.D2.20.NP.L500 145,64 120,90 120,60 120,60 Distortional controls
I.D2.5.NP.L500 131,55 108,87 118,70 108,87 Local-Global controls

column length.

83
column length.
D.5.NP.L1500 71,77 71,77 74,99 71,77 Global controls
4.1.2 Nominal strength - lip length relationships of unperforated column models
Totally 6 different nominal strength graphs have been constructed for unperforated
models for different column lengths (L=500, L=1000, L=1500) and different lip
lengths and directions. Half of these graphs show the relationship for inward lip
direction cases. And the other half for the outward lip direction cases as described at
the begining of chapter 4. X axis of these graphs indicate the lip length and Y axis
indicate the nominal strengths of the sepcimens. 5 mm - 20 mm interval of the lip
length indicates the D length. D2 length covers the margin of 20 mm – 50 mm which
is also indicating the specimen’s lip configuration continues inwards or outwards. D3
length of specimen covers the final 50 mm – 60 mm interval. These regions are
shown on all the following graphs. In figures 4.3 to 4.8 ; these graphs are given for
unperforated specimens. Controlling buckling modes are marked as D for distortional
buckling, LG for Local-global buckling and G for global buckling.
84
Figure 4.3 : Nominal strength – lip length relationship of the unperforated specimens
having 500 mm column length (lip continues outward).
having 500 mm column length (lip continues inward).
85
86
87
4.1.2.1 General comparison of results for unperforated column models
Previous graphs are used to compare the nominal strengths of unperforated
specimens. Five different graphs have been constructed with two major feature. First
feature is the lip direction. Figures 4.9 to 4.11 include the comparison between the
sections having same column length different lip directions. Figures 4.12 and 4.13
include the comparison between the sections having same lip direction different
column lengths.

unperforated specimens having 500 mm column length.
On specimens having 500 mm and 1000 mm column lengths; lip direction starts to
effect the column strength after 40 mm lip length which means D2 lip is 20 mm long.
Comparing the outward and inward lip cases, it is noted that there is a 5% strength
difference the outward case being lower than the inward case. On specimens having
500 mm column length, it is observed that 2 specimens having maximum limit of lip
length (60 mm) but having different lip directions have the same column capacity.
This is shown by point A on Figure 4.9.
88

89
However, on specimens having 1000 mm column length, strength loss starts on the
same lip length as in 500 mm column length but 2 specimens having maximum limit
of lip length (60 mm) don’t have the same strength. Similar trend is observed for
1000 column lengths as in 500 mm column lengths.Between 40mm and 60mm lip
lengths (D2 = 20-40 mm) if lip continues outwards, column strength will be less than
in the situation that lip continues inwards. In this region; maximum strength
difference is noted as 7.1%
On specimens having 1500 mm column length, a more or less linear relationship is

observed both for outward and inward lip cases. The column strength of specimens
having lips continuing outwards is greater than the column strength of the specimens
having lips continuing inwards. Strength difference is up to 8.8% on sections having
maximum lip length. Figures 4.12 and 4.13 include the comparison of nominal
strengths for the models having same lip direction but different column lengths.
unperforated specimens having lips continuing outwards.
90
unperforated specimens having lips continuing inwards.
On specimens having lips continuing inwards; 1500 mm long specimens have the
same approximate linearity on its function between column strength and lip length.
An interesting point is noted on 500 mm and 1000 mm long specimens, column
strengths are the same on 44 mm of lip length (D2 lip = 24 mm). And the strength
gain and the column strength are the same on increasing lip lengths. That means D3
lip causes the same effect on both 500 mm and 1000 mm long specimens. It should
be noted that for this region of lip lengths for both 500 mm and 1000 mm specimens
local-global buckling is the controlling mode. On 1000 mm long specimens, between
the lip length of 34-42 mm (D2 lip = 14-22 m) there’s a significant increase in
column strength which is also similar with specimens with lips continuing inwards.
As noted earlier; this situation seems to be related with the change in the buckling
mode of corresponding cross-sections in this region (from distortional to local-global
and vice versa).
91
4.2 Analysis of Perforated Specimens
60 perforated column models have been used in the analysis. The models have same
hole configuration and different lip directions and lengths. Perforation configuration
for all perforated specimens are same and given in Figure 4.14. Major concern is the
effect of web perforations in rack columns. The flange perforations are not included
on models, only perforations on the webs are considered. Reduced Thickness
approach is used, which is previously explained in section 2.2 in the study titled
“Design of Steel Storage Rack Columns via the Direct Strength Method” [61].
Chosen perforation configuration of the model is within the geometric limits of
reduced thickness table given in Table 4.10.
Table 4.10 : Limits of the model of reduced thickness [61].
Main geometric parameters of a column having web perforations have already been
given in Figure 2.26. A check based on these limits has been made for the chosen
configuration of perforations on our typical model. Dimension notation in Figure
4.14 is adapted for the limits given in Table 4.10 and following calculations are
related with these limits.
in the limits of 24 to 88 OK!

in the limits of 26 to 83 OK!
in the limits of 0.48 to 1.87 OK!
in the limits of 50 mm to 75 mm OK!
OK!

92
Geometric details of all perforated specimens used are given in tables 4.11 to 4.13
for different column lengths. Specimen designation is the same as previously
described in Figure 4.2. In these tables, perforated section areas are calculated and
given for different models.
Figure 4.14 : Typical model for perforation configuration of all perforated

specimens.
93
column length.
t h0 b0 D D2 D3 Section Area Perforated Sect. Area Column Length

Column (mm) (mm) (mm) (mm) (mm) (mm) (mm2) (mm2) (mm)
D.5.P.L500 2 140 50 5 - - 484 424 500
D.10.P.L500 2 140 50 10 - - 504 444 500
D.15.P.L500 2 140 50 15 - - 524 464 500
D.20.P.L500 2 140 50 20 - - 544 484 500
O.D2.5.P.L500 2 140 50 20 5 - 556 496 500
O.D2.10.P.L500 2 140 50 20 10 - 576 516 500
O.D2.15.P.L500 2 140 50 20 15 - 596 536 500
O.D2.20.P.L500 2 140 50 20 20 - 616 556 500
O.D2.25.P.L500 2 140 50 20 25 - 636 576 500
O.D2.30.P.L500 2 140 50 20 30 - 656 596 500
O.D3.5.P.L500 2 140 50 20 30 5 668 608 500
O.D3.10.P.L500 2 140 50 20 30 10 688 628 500
I.D2.5.P.L500 2 140 50 20 5 - 556 496 500
I.D2.10.P.L500 2 140 50 20 10 - 576 516 500
I.D2.15.P.L500 2 140 50 20 15 - 596 536 500
I.D2.20.P.L500 2 140 50 20 20 - 616 556 500
I.D2.25.P.L500 2 140 50 20 25 - 636 576 500
I.D2.30.P.L500 2 140 50 20 30 - 656 596 500
I.D3.5.P.L500 2 140 50 20 30 5 668 608 500
I.D3.10.P.L500 2 140 50 20 30 10 688 628 500

column length.

D.5.P.L1000 2 140 50 5 - - 484 424 1000
D.10.P.L1000 2 140 50 10 - - 504 444 1000
D.15.P.L1000 2 140 50 15 - - 524 464 1000
D.20.P.L1000 2 140 50 20 - - 544 484 1000
O.D2.5.P.L1000 2 140 50 20 5 - 556 496 1000
O.D2.10.P.L1000 2 140 50 20 10 - 576 516 1000
O.D2.15.P.L1000 2 140 50 20 15 - 596 536 1000
O.D2.20.P.L1000 2 140 50 20 20 - 616 556 1000
O.D2.25.P.L1000 2 140 50 20 25 - 636 576 1000
O.D2.30.P.L1000 2 140 50 20 30 - 656 596 1000
O.D3.5.P.L1000 2 140 50 20 30 5 668 608 1000
O.D3.10.P.L1000 2 140 50 20 30 10 688 628 1000
I.D2.5.P.L1000 2 140 50 20 5 - 556 496 1000
I.D2.10.P.L1000 2 140 50 20 10 - 576 516 1000
I.D2.15.P.L1000 2 140 50 20 15 - 596 536 1000
I.D2.20.P.L1000 2 140 50 20 20 - 616 556 1000
I.D2.25.P.L1000 2 140 50 20 25 - 636 576 1000
I.D2.30.P.L1000 2 140 50 20 30 - 656 596 1000
I.D3.5.P.L1000 2 140 50 20 30 5 668 608 1000
I.D3.10.P.L1000 2 140 50 20 30 10 688 628 1000
94
column length.

D.5.P.L1500 2 140 50 5 - - 484 424 1500
D.10.P.L1500 2 140 50 10 - - 504 444 1500
D.15.P.L1500 2 140 50 15 - - 524 464 1500
D.20.P.L1500 2 140 50 20 - - 544 484 1500
O.D2.5.P.L1500 2 140 50 20 5 - 556 496 1500
O.D2.10.P.L1500 2 140 50 20 10 - 576 516 1500
O.D2.15.P.L1500 2 140 50 20 15 - 596 536 1500
O.D2.20.P.L1500 2 140 50 20 20 - 616 556 1500
O.D2.25.P.L1500 2 140 50 20 25 - 636 576 1500
O.D2.30.P.L1500 2 140 50 20 30 - 656 596 1500
O.D3.5.P.L1500 2 140 50 20 30 5 668 608 1500
O.D3.10.P.L1500 2 140 50 20 30 10 688 628 1500
I.D2.5.P.L1500 2 140 50 20 5 - 556 496 1500
I.D2.10.P.L1500 2 140 50 20 10 - 576 516 1500
I.D2.15.P.L1500 2 140 50 20 15 - 596 536 1500
I.D2.20.P.L1500 2 140 50 20 20 - 616 556 1500
I.D2.25.P.L1500 2 140 50 20 25 - 636 576 1500
I.D2.30.P.L1500 2 140 50 20 30 - 656 596 1500
I.D3.5.P.L1500 2 140 50 20 30 5 668 608 1500
I.D3.10.P.L1500 2 140 50 20 30 10 688 628 1500
4.2.1 Analytical study on perforated members using direct strength method

A study including construction of elastic buckling curves with CUFSM Software
and DSM Calculations has been done to calculate the nominal section strength of
these specimens. Critical elastic column buckling loads of these various sections are
obtained from CUFSM software. For the calculation of elastic buckling loads for all
perforated specimens reduced thickness method was used. In section 2.3 of this
thesis; reduced thickness formulas have been given. Details of this method was
presented in section 2.3 under the title “Design of Steel Storage Rack Columns via
the Direct Strength Method” [61].
Basically; 3 different reduced thickness values are calculated in the begining of

analysis. In CUFSM, perforation locations are represented by elements having these
reduced thickness values. A typical section constructed in CUFSM is given in
Figure 4.16 for reduced thickness approach where reduced thickness on the section is
shown. For the determination of each buckling load (local, distortional, global) a
seperate calculation was made by using a different reduced thickness formula e.g. for
the calculation of elastic buckling load for distortional buckling, the corresponding
formula is used to calculate the reduced thickness value. Following formulas are used
for determining reduced thickness values for local, distortional and global (torsional-
95
flexural) buckling modes for all perforated specimens. Formulas used are previously
given in section 2.3:
Reduced thickness for local buckling;
(2.24)
Reduced thickness for distortional buckling;
( ) (2.25)
( )
Reduced thickness for torsional-flexural buckling;
( ) (2.26)
( )
96
Figure 4.15 : Typical model for reduced thickness approach from CUFSM.
These thickness values are used where perforations are situated on cross-sections
constructed with CUFSM software. Tables 4.14 to 4.16 include the results of critical
elastic buckling loads obtained by CUFSM. Each critical elastic buckling load is
found using related reduced thickness value for the effects of perforations. For
models with non-existent buckling modes; a high value of elastic critical buckling
load is intentionally assumed to calculate the slenderness values. The non-existent
buckling modes are indicated with N.A in relevant cells of tables 4.14 to 4.16.
Direct Strength Method is used after obtaining the critical elastic buckling loads.
Nominal strengths for local, distortional and global buckling are calculated with
DSM equations. Then, the nominal strength for all sections are determined. Results
are given in tables 4.17 to 4.19. Controlling mode indicates which buckling mode
governs the behaviour of the member.
97
Table 4.14 : Critical elastic buckling loads obtained from CUFSM analysis for perforated specimens having 500 mm column length.
98
99
100
Table 4.17 : Nominal axial strengths of perforated specimens having 500 mm
column length.
D.5.P.L500 99,60 84,86 63,38 63,38 Distortional controls
D.15.P.L500 108,99 87,76 88,77 87,76 Local-Global controls
O.D2.5.P.L500 116,51 94,07 101,33 94,07 Local-Global controls
O.D2.25.P.L500 135,30 109,49 106,47 106,47 Distortional controls
I.D2.5.P.L500 116,51 93,96 103,02 93,96 Local-Global controls

column length.

101
column length.

4.2.2 Nominal strength - lip length relationships of perforated column models

Totally 6 different nominal strength graphs have been constructed for perforated
models for different column lengths (L=500, L=1000, L=1500) and different lip
lengths and directions. Half of these graphs show the relationship for inward lip
direction cases. And the other half for the outward lip direction cases as described at
the begining of chapter 4. X axis of these graphs indicate the lip length and Y axis
indicate the nominal strengths of the sepcimens. 5 mm - 20 mm interval of the lip
length indicates the D length. D2 length covers the margin of 20 mm – 50 mm which
is also indicating the specimen’s lip configuration continues inwards or outwards. D3
length of specimen covers 50 mm – 60 mm interval. These regions are shown on all
the following graphs. In figures 4.16 to 4.21 ; these graphs are given for perforated
specimens. Controlling buckling modes are marked as D for distortional buckling,
LG for Local-global buckling and G for global buckling.
102
103
104
105
4.2.2.1 General comparison of results for perforated column models
Previous graphs are used to compare the nominal strengths of perforated specimens.
Five different graphs have been constructed with two major feature. First feature is
the lip direction. Figures 4.22 to 4.24 include the comparison between the sections
having the same column length with different lip directions. Figures 4.26 and 4.27
include the comparison between the sections having same lip direction different
column lengths.

perforated specimens having 500 mm column length.
On specimens having 500 mm and 1000 mm column lengths; lip direction starts to
effect the column strength after 40 mm lip length which means D2 lip is 20 mm long,
similar as unperforated specimens. Comparing the outward and inward lip cases, it is
noted that there is a 4.7% strength difference the outward case being lower than the
inward case. On specimens having 500 mm column length, it is observed that 2
specimens having maximum limit of lip length but having different lip directions
have approximately the same column capacity. This is shown by point B on Figure
4.22.
106

107
However, on specimens having 1000 mm column length, strength loss starts on the
30 mm lip length (D2 = 10 mm) and two specimens having maximum limit of lip
length (60mm) don’t have the same strength. Similar trend is observed for specimens
having lips continuing outward and inward. If lip continues outwards, column
strength will be less than in the situation that lip continues inwards. This strength
difference increases after 40 mm of lip length (D2 = 20 mm). Maximum strength
difference is noted as 8.6%.
On specimens having 1500 mm column length, the specimens having lips continuing
inwards has a linear relationship up to 45 mm of lip length (D2 = 25mm). The
column strength of specimens having lips continuing outwards are greater than the
specimens having lips continuing inwards. This strength difference begins after 45
mm of lip length (D2 = 25mm) and is up to 30.4% on sections having maximum lip
length. It is interesting to note here that in this region, where there is high strength
difference controlling buckling modes are also different. Inward cases are dominated
by local-global buckling mode whereas the corresponding outward cases in this
region are dominated by distortional buckling (As shown in Figure 4.25). This
behaviour of buckling mode difference was not observed for unperforated cases
although there was still a strength difference between outward and inward cases
(8.8% more for outward).
Figure 4.25 : Controlling buckling modes for inward and outward cases of
perforated 1500 mm long columns with maximum lip length
configuration.
As a result of these comparisons, lip direction effects the same specimens differently
with respect to their column length. Especially, on sections having 1500 mm column
length, the strength difference is significantly considerable on sections having more
108
complicated lips. The column lengths between 1000 mm and 1500 mm are
distinctive for the effects of inward or outward direction of rack column specimens.
Figures 4.26 and 4.27 include the comparison of nominal strengths for the models
having same lip direction, different column lengths.
perforated specimens having lips continuing outwards.
On specimens having lips continuing outwards; in general, strength of 500 mm long

specimens are greater than strength of 1000mm long specimens. The strength of
1000 mm long specimens are generally greater than strength of 1500mm long
specimens except 25-35 mm (D2 = 5-15 mm) and 50-55mm (D3 = 0-5 mm) of lip
length intervals. Specimens having 40 mm lip length (D2 lip = 20 mm) with 500 and
1000 mm column lengths have the same strength. On 1000 mm long specimens,
between the lip length of 35-40 mm (D2 lip = 15-20 m) there’s a significant increase
in strength. This situation seems to be related with the change in the buckling mode
of corresponding cross-sections in this region (from distortional to local-global and
vice versa).
109
perforated specimens having lips continuing inwards.
On 500 and 1000 mm long specimens, between 40 – 50 mm of lip length (D2 lip =
20-30 m) not a major strength gain is seen but after 50 mm of lip length (start of D3
lip) , there is a sudden increase in strength which shows the importance of D3 lip on
column strength. On 1500 mm long specimens, there is also a sudden increase in
strength between 45 – 50 mm of lip length (D2 lip = 25-30 m). This situation seems
to be also related with the change in the buckling mode of corresponding cross-
sections in this region (from local-global to distortional).
On specimens having lips continuing inwards; 1500 mm long specimens have

approximate linearity on its function between column strength and lip length. An
interesting point is noted on 500 mm and 1000 mm long specimens, column
strengths are the same on 44 mm of lip length (D2 lip = 24 m). And the strength gain
and the column strength are the same on increasing lip lengths. That means D3 lip
causes the same effect on both 500 mm and 1000 mm long specimens. On 1500 mm
and 1000 mm long specimens, between 18-34 mm of lip length (begining of D2 lip
until 14 mm length) the strength values is approximately same. In this region;
buckling modes for both cases are also same (local-global controlling modes). On
110
1000 mm long specimens, between the lip length of 34-42 mm (D2 lip = 14-22 m)
there’s a significant increase in column strength which is also similar with specimens
having lips continuing outwards. As noted earlier this situation seems to be related
with the change in the buckling mode of corresponding cross-sections in this region
(from distortional to local-global and vice versa).
4.3 Comparison Between the Strength of Unperforated Rack Column Models

with Perforated Column Models
Comparison of relationships between nominal strength and lip length for

unperforated and perforated members is given on 6 different graphs for 3 different
lengths. X axis of these graphs indicate the lip length and Y axis indicate the nominal
strengths of the specimens. 5 mm - 20 mm interval of the lip length indicates the D
length. D2 length covers the margin of 20 mm – 50 mm which is also indicating the
specimen’s lip configuration continuing inwards or outwards. D3 length of specimen
covers 50 mm – 60 mm interval. In figures 4.28 to 4.33 ; these graphs are given
including comparisons for various column lengths and lip directions.
and column length 500 mm.
111
112
113
On specimens having lips continuing outward and column length 500 mm,
perforated and unperforated curves follow a similar trend for increasing lip length.
However, as expected perforated specimens’ strength values are less than
unperforated specimens. This strength difference is calculated as minimum 12.2%
maximum 14.1% . On specimens having lips continuing inward and column length
500 mm, again similar trend is observed for the curves of perforated and
unperforated specimens. Strength difference is calculated as minimum 12%
maximum 14%.
perforated and unperforated curves have similarity in shapes but differ in severeal
intervals of lip length. Between 35 and 40 mm of lip length (D2 = 15-20 mm) there is
a significant increase in strength on both perforated and unperforated specimens. The
amount of increase on perforated specimens is calculated as 30 kN however the
increase seen on unperforated specimens is calculated as 20 kN in related interval.
This situation shows D2 lip length affects perforated sections more than unperforated
sections on corresponding column length and lip direction. Generally, perforated
114
specimens’ strength values are less than unperforated specimens. This strength
difference is calculated as minimum 14.2% maximum 36.9% . On specimens having
lips continuing inward and column length 1000 mm, similar situation in same
interval is also observed. The amount of increase on perforated specimens is
calculated as 28 kN however the increase seen on unperforated specimens is
calculated as 20 kN in related interval. In general, perforated specimens’ strength
values are less than unperforated specimens and this strength difference is calculated
as minimum 14.3% maximum 29.8% . It has been noted that maximum strength
difference calculated between perforated and unperforated sections on specimens
having lips continuing outwards are 7.1% greater than specimens having lips
continuing inward.
perforated and unperforated curves follow similar trend up to 40 mm of lip length
(D2=20 mm). After 40 mm of lip length, there is a significant increase on strength of
perforated specimens and on 50 mm of lip length (end of D2 lip) perforated
specimens’ strength is approximately same as unperforated specimens. On
continuing lip lengths there is a drop on perforated specimens’ strength and at the
end of D3 lip, the strength values of perforated and unperforated specimens are
approximately same again. This situation might be related to the change in the
controlling mode of related perforated specimens within those lip length intervals.
Perforated specimens on related interval have distortional controlling buckling mode
and unperforated specimens have local-global controlling buckling mode. Maximum
strength difference calculated between perforated and unperforated specimens is
29.6% and minimum is 0%. On specimens having lips continuing inward and column
length 1500 mm, similar situation is not observed. A parallel trend is noted for the
curves of perforated and unperforated specimens. Generally, perforated specimens’
strength values are less than unperforated specimens and maximum strength
difference is calculated as 20.9% and minimum strength difference is calculated as
17.9%. It has been noted that maximum strength difference calculated between
perforated and unperforated sections on specimens having lips continuing outwards
are 8.7% greater than specimens having lips continuing inwards.
115
116
5. CONCLUSION
Steel Storage Rack Systems play a vital role in logistics and storing goods.This has
particularly increased the importance of these systems over the last decade. In our
country, importance of these systems is increasing through the demand of
insufficient space to storage goods from various industries and fields. The structural
strength and stability of steel storage rack structures have an increasing significance
since the usage of these structures and value of the goods carried by these structures
have increased.
Determining the structural capacity of steel storage rack columns has currently been
regulated by several standarts and methods. American Iron and Steel Institute S100-
2007 Standart is one of these few standarts to regulate the method of calculation to
determine the strength of cold formed steel columns. The standart presents Effective
Width Method and Direct Strength Method for calculations.
Direct Strength Method, presented in Appendix 1 part of AISI S100-2007 standart

has newly been developed method for determination of the strength of concentrically
axially loaded cold formed steel members. DSM calculations include finite strip
method which is previously summarized in detail.
Steel Storage Rack Columns can be assumed to be a specific type of the general Cold
formed steel columns. However, the behaviour of steel storage rack columns differs
from the general behaviour of the CFS columns due to their thickness, lip
configuration and the presence of perforations. Steel storage rack columns may
consist of various lip types, lengths and perforations for the intended use. These two
major variables affect the capacity of rack columns and hence the stability and
strength of whole rack structure. In this study, the effects of lip configuration and
perforations on the strength of steel storage rack columns is investigated using Direct
Strength Method.
The main parts of the present study can be summarized as follows;
117
 Review of previous studies conducted on steel storage rack columns and
direct strength method.
 Assessment of current practise (AISI S100-700 provisions) with respect to

calculation of cold formed steel column capacity with Effective Width
Method and Direct Strength method.
 The analytical study carried out on 120 column models having different
column lengths, lip directions, lip lengths and perforated and unperforated
cases.
In AISI S100-700 standart, currently the effect of perforations on column strength is

accounted for by physical testing. No design rules are given to consider the effect of
perforations on the three major buckling modes including local, distortional and
global. In this study, a recently developed analytical approach which integrates the
Direct Strength Method by a method called “Reduced Thickness Approach” was
used. Comparisons were made between nominal strength values of columns with
varying cross-section geometries (lip configuration) for both unperforated and
perforated cases. Also three different column lengths were considered. The results
were given in tables and graphs showing the effect of changing the above mentioned
variables. In these tables and graphs, it was shown that changing the lip direction
(inward or outward) effected the column strength. Significant differences were
observed for specific ranges of lip lengths and column lengths. For example, sudden
increases in strength were observed particularly for 1000 mm long columns with
having both inward and outward lip configurations within a specific range of lip
lengths. In this range, it was observed that buckling modes were also changing. As
expected, the presence of perforations resulted in strength reductions of up to 36.9%.
By using the reduced thickness approach, it was possible to calculate these
differences in strength due to presence of perforations. The details of these findings
were given within the thesis. It is recommended as a future study to support these
findings by experimental testing.
118
REFERENCES
[1] Yu, W. W., and LaBoube, R. A. (2010). Cold Formed Steel Design. Fourth Ed.,
John Wiley & Sons, Inc., Missouri University of Science and
Technology
[2] Cold-formed steel. (n.d.). In Wikipedia. Date retrieved: 20.03.2013, adress:
http://en.wikipedia.org/wiki/Cold_formed_steel
[3] Cold-formed steel sections. (n.d.). In Ruukki. Date retrieved: 20.03.2013, adress:
http://www.ruukki.com/~/media/Images/Steel-products/Cold-rolled-
steels/cold-formed-steel-sections.ashx
[4] Chen, H. (n.d.). An Overview of Cold-Formed Steel Structures, American Iron
and Steel Institute, AIA HSW, Washington DC.
[5] Roll forming. (n.d.). In Cmrp. Date retrieved: 20.03.2013, adress:
http://www.cmrp.com/blog/wp-content/uploads/2014/02/Roll-
Forming-with-a-Series-of-Progessive-Forming-Dies.jpg
[6] CFS building. (n.d.). In Alphaturn. Date retrieved: 20.03.2013, adress:
www.alphaturn.grlight_construction.aspx
[7] Three adjacent building blocks, Boulogne. (n.d.). In Nemetscheck-scia. Date
retrieved: 20.03.2013, adress: http://nemetschek-
scia.com/sites/nemetschek-scia.com/files/references/profil-du-
future_billancourt.jpg
[8] CFS hangar. (n.d.). In Elast. Date retrieved: 20.03.2013, adress:
http://www.elast.com.au/images/thiess%207.jpg
[9] CFS roof deck sheet. (n.d.). In D-macindustries. Date retrieved: 20.03.2013,
adress: http://www.d-
macindustries.com/images/type%20b%20roof%20deck.jpg
[10] CFS roof deck. (n.d.). In Structuremag. Date retrieved: 20.03.2013, adress:
http://www.structuremag.org/images/0108-is-1.jpg
[11] CFS wall panel. (n.d.). In Chinasteelstructure. Date retrieved: 20.03.2013,
adress: http://www.chinasteelstructure.en.made-in-china.com/
[12] Freitas, A. M. S., Freitas, M. S. R., Souza F. T. (2005). Analysis of steel
storage rack columns. Journal of Constructional Steel Research, 61,
1136-1137.
[13] CFS storage rack structure. (n.d.). In Weiku. Date retrieved: 20.03.2013,
adress: http://www.weiku.com/
[14] Pallet Racking. (n.d.). In Wikipedia. Date retrieved: 20.03.2013, adress:
http://en.wikipedia.org/wiki/File:Drive-In-Rack.jpg
119
[15] Pallet Racking Hull. (n.d.). In Premier-storage. Date retrieved: 20.03.2013,
adress: http://www.premier-storage.co.uk/projects/wp-
content/uploads/2011/08/pallet_racking_std_mainpic_large.jpg
[16] AISI S100-2007 (2007). North American Specification for the Design of Cold-
Formed Steel Structural Members, American Iron and Steel Institute,
USA.
[17] Winter, G. (1970). Commentary on the 1968 Edition of the Specification for
the Design of Cold-Formed Steel Structural Members, American Iron
and Steel Institute, Washington, DC.
[18] Chajes, A., and Winter, G. (1965). Torsional-Flexural Buckling of Thin
Walled Members. Journal of the Structural Division, ASCE
Proceedings, vol. 91.
[19] Chajes, A., Fang, P. J., and Winter, G. (1966). Torsional-Flexural Buckling,
Elastic and Inelastic, of Cold-Formed Thin-Walled Columns,
Engineering Research Bulletin, 66-1, Cornell University, Ithaca, NY.
[20] Peköz, T. B. (1986). Development of a Unified Approach to the Design of
Cold-Formed Steel Members. AISI Report SG-86-4, American Iron
and Steel Institute, Washington, DC.
[21] Shanley, F. R. (1947). Inelastic Column Theory. Journal of Aeronautical
Science, 14.
[22] Galambos, T. V. (1968). Structural Members and Frames. Prentice-Hall,
Englewood Cliffs, NJ.
[23] Goodier, J. N. (1941). The Buckling of Compressed Bars by Torsion and
Flexure. Cornell University Experiment Station, Ithaca, NY.
[24] Timoshenko, S. P. (1945). Theory of Bending, Torsion and Buckling of Thin-
Walled Members of Open Cross Section. Journal of the Franklin
Institute, 4,5.
[25] Bleich, F. (1952). Buckling Strength of Metal Structures. McGraw-Hill, NY.
[26] Flexural-torsional column buckling. (n.d.). In Emeraldinsight. Date retrieved:
22.04.2013, adress:
http://www.emeraldinsight.com/content_images/fig/1820250204024.p
ng
[27] Flexural-torsional column buckling of I beam. (n.d.). In Fhwa. Date
retrieved: 22.04.2013, adress:
http://www.fhwa.dot.gov/bridge/lrfd/images/us_ds3_img22785.gif
[28] Maljaars, J. (2008). Local Buckling of Slender Alminium Sections Exposed to
Fire, PhD Thesis, Eindhoven Technical University, Netherlands.
[29] Gerard, G., and Becker. H. (1957). Handbook of Structural Stability, Part –I
Buckling of Flat Plates. NACA Technical Note 3781, Washington, DC.
[30] Timoshenko, S. P., Gere, J. M. (1961). Theory of Elastic Stability. McGraw-
Hill, NY.
[31] Rasmussen, K. J. R., Hancock, G. J. (1992). Nonlinear Analyses of Thin-
Walled Channel Section Columns. Thin-Walled Structures, 13, 1-2
120
[32] Rasmussen, K. J. R. (1994). Design of Thin-Walled Column with Unstiffened
Flanges. Engineering Structures, 13, 5.
[33] Nagahama, K., Camotim, D., Batista, E. (2002). Local Buckling, Post-
Buckling and Mode Interaction Finite Element Analyses in Cold-
Formed Steel Members. Proceedings of the sixth conference on
Computational structures technology, 235-236.
[34] Hancock, G. J. (1998). Design of Cold-Formed Steel Structures. Third Ed.
Australian Institute of Steel Construction, Sydney.
[35] Schafer, B. W. (2012). In John Hopkins University Civil Engineering.
Professor’s personal web page Date retrieved: 05.05.2013, adress:
http://www.ce.jhu.edu/bschafer/cufsm/
[36] Young, B., Jintang Y. (2004). Column Tests of Cold-Formed Steel Channels
with Complex Stiffeners by Direct Strength Method. Journal of
Structural Engineering, 130, 1756-1763.
[37] Jintang Y., Young, B. (2002). Column Tests of Cold-Formed Steel Channels
with Complex Stiffeners. Journal of Structural Engineering, 128,
737–745.
[38] Hancock, G. J., Cook, D., Moisy, R. E., Yen A. (2005). Direct Strength
Design of Hot Rolled and Cold-Formed Steel Compression Members.
Advances in Steel Structures, I, 13-20.
[39] Papengelis, J. P. and Hancock, G. J. (1995), Computer Analysis of Thin-
Walled Structural Members. Computers and Structures, 56(1), 157-
176.
[40] Schafer, B. W., Ádány, S. (2006). Buckling analysis of cold-formed steel
members using CUFSM: conventional and constrained finite strip
methods. 18th International Specialty Conference on Cold-Formed
Steel Structures, Orlando.
[41] Schafer, B. W. (2008). Review: The Direct Strength Method of cold-formed
steel member design. Journal of Constructional Steel Research, 64,
766–778.
[42] Lau, S. C. W., Hancock, G. J. (1987). Distortional buckling formulas for
channel columns. Journal of Structural Engineering, 113(5).
[43] Kwon, Y. B., Hancock, G. J. (1992). Strength tests of cold-formed channel
sections undergoing local and distortional buckling. Journal of
Structural Engineering, 117(2), 803-1786.
[44] Hancock, G. J., Kwon, Y. B., Bernard, E.S. (1994). Strength design curves for
thin-walled sections undergoing distortional buckling. Journal of
Constructional Steel Research, 31, 2-3.
[45] Schafer, B. W. (2000). Distortional buckling of cold-formed steel columns.
Final Report to the American Iron and Steel Institute, Washington DC
[46] Schafer, B. W. (2002). Local, distortional, and Euler buckling in thin-walled
columns. Journal of Structural Engineering, 128(3), 289–99.
121
[47] Yang, D., Hancock, G. J. (2004). Compression tests of high strength steel
channel columns with interaction between local and distortional
buckling. Journal of Structural Engineering, 130(12), 1954–63.
[48] Silvestre, N., Camotim, D. (2006). Direct Strength Method for lipped channel
columns and beams affected by local-plate/distortional interaction.
Proceedings of the eighteenth international specialty conference on
cold formed steel structures, 17–37.
[49] Kumar, M. V. A., Kalyanaraman, V. (2010). Evaluation of Direct Strength
Method for CFS Compression Members without Stiffeners. Journal of
Structural Engineering, 136(7) , 879–885.
[50] Mulligan, G. P., and Peköz, T. B. (1984). Local buckling interaction in cold
formed columns. Journal of Structural Engineering, 110(11), 2635-
2654.
[51] Talja, A. (1990). Design of the buckling resistance of compressed HSS
channels. Research Note No. 1163, Technical Research Centre of
Finland.
[52] Young, B. (1998). The behaviour and design of the cold formed channel
columns, Ph.D. thesis, Univ. of Sydney, Australia.
[53] Key, P. W., Hasan, S. W., Hancock, G. J. (1988). Column behaviour of cold
formed hollow sections. Journal of Structural Engineering, 1142,
309-407.
[54] Wilkinson, T., and Hancock, G. J. (1997). Test for the compact web
slenderness of cold formed rectangular hollow section. Res. Rep.
No.R744, Dept. of Civil Engineering, Univ. of Sydney.
[55] Niemi, E., and Rinnevalli, J. (1990). Buckling tests on cold formed square
hollow sections of steel Fe510. J. Constr. Steel Res., 16(3), 221-230.
[56] Kalyanaraman, V., Pekoz, T. B., and Winter, G. (1979). Unstiffened
compression elements. ASCE Annual Convention and Exposition. 1-
38.
[57] Landesmann, A. and Camotim, D. (2013). On the Direct Strength Method
(DSM) design of cold-formed steel columns against distortional
failure. Thin-walled Structures, 67, 1.
[58] Roure, F., Pastor, M. M., Casafont, M., Somalo, M. R. (2010). Stub column
tests for racking design: Experimental testing, FE analysis and EC3.
Thin-walled Structures, 49, 167-184.
[59] European Standard EN1993-1-3 (2009). Eurocode 3 - Design of Steel
Structures, European Committee for Standardization, Brussels.
[60] Casafont, M., Pastor, M. M., M., Roure, F., Peköz, T. B. (2011). An
experimental investigation of distortional buckling of steel
storage rack columns. Thin-walled Structures, 49, 933-946.
[61] Casafont, M., Pastor, M. M. M., Roure, F., Bonada, J., Peköz, T. B. (2013).
Design of Steel Storage Rack Columns via the Direct Strength
Method. Journal of Structural Engineering, 139, 669-679.
122
[62] Casafont, M., Pastor, M. M., M., Bonada, J., Roure, F., Peköz, T. B. (2012).
Linear buckling of perforated steel storage rack columns with
the finite strip method. Thin-walled Structures, 61, 71-85.
[63] Moen, C. D., Schafer, B. W. (2009). Direct Strength Design of Cold-
Formed Steel Members With Perforations. Research Report, AISI
Committee on Specifications for the Design of Cold-Formed Steel
Structural Members, Washington, DC.
123
124
CURRICULUM VITAE
Name Surname:
Atakan MANGIR
Place and Date of Birth:
Istanbul – 09.04.1989
Address:
Seyitnizam mah. Akevler Sitesi, C-1 Blok, Kat:7, D:25, Zeytinburnu / İST.
E-Mail:
atakanmangir@gmail.com
B.Sc.:
Civil Engineering - İstanbul Kültür University (100% Sch.)
M.Sc.:
Structural Engineering - Istanbul Technical University (2011-2014)
Professional Experience and Rewards:
(2011) Istanbul Kültür University

1st Degree Reward in Hall of Fame between graduates of Civil
Engineering Dept.
(2011-2012) DEMARS Construction

Maintenance and restoration of structures and facilities owned by
Istanbul Metropolitan Municipality
Project Engineer
(2012 - ) Fatih University

Civil Engineering Dept. – Research Assistant.
(2013-2014) Istanbul Development Agency

Establishment of Industrial Storage Rack Systems (ISRS) Design
and Test Centre, Project Assistant - Fatih University.
125

Hafif Celik Tez

Uploaded by

Copyright:

Available Formats

Hafif Celik Tez

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Hafif Celik Tez

Uploaded by

Copyright:

Available Formats

ISTANBUL TECHNICAL UNIVERSITY  GRADUATE SCHOOL OF SCIENCE

ENGINEERING AND TECHNOLOGY

STRENGTH AND STABILITY OF

Department of Civil Engineering

Structural Engineering Programme

STRENGTH AND STABILITY OF

Department of Civil Engineering

Structural Engineering Programme

Thesis Advisor: Prof. Dr. Ünal ALDEMİR

İNCE CİDARLI ÇELİK DEPO RAF SİSTEMLERİ KOLONLARININ

YÜKSEK LİSANS TEZİ

İnşaat Mühendisliği Anabilim Dalı

Yapı Mühendisliği Programı

Tez Danışmanı: Prof. Dr. Ünal ALDEMİR

Thesis Advisor : Prof. Dr. Ünal ALDEMİR ..............................

Co-advisor : Assoc. Prof. Dr. Güven KIYMAZ ..............................

Jury Members : Prof. Dr. Abdül HAYIR .............................

Prof. Dr. Bülent AKBAŞ ..............................

Assis. Prof. Dr. Cüneyt VATANSEVER ..............................

Date of Submission : 5 May 2014

May 2014 Atakan MANGIR

AISI : American Iron and Steel Institute

A : Full unreduced cross-sectional area

Table 3.1 : Determination of plate buckling coefficient, k ..................................... 51

Figure 1.1 : Cold formed steel member shapes. ........................................................ 1

The main parts of the present study can be summarized as follows;

In AISI S100-700 standart, currently the effect of perforations on column strength is

Mevcut çalışma aşağıdaki şekilde özetlenebilir;

1.1 Overview of Cold Formed Steel

Figure 1.1 : Cold formed steel member shapes [3].

The manufacturing of cold-formed steel products occurs at room temperature. Design

Figure 1.2 : Cold roll-forming process (description) [2].

1.1.2 Structural applications

Construction where cold-formed steel is finding wider application is in steel framing

Figure 1.5 : Cold formed steel structure example [7].

Some of the sections are shown in Figure 1.7.

Figure 1.7 : Some CFS section geometries [1].

Figure 1.9 : Cold formed steel roof deck example [10].

Figure 1.13 : A drive in model of rack stuctures [14].

1.1.3 Advantages and disadvantages of cold formed steel

- Mass production and prefabrication ease

- High strength and stiffness

- Easy, quick and facilitated assembly

- Delays due to weather conditions

- Formwork is not needed

- Rot and termite proof

- Low transportation and handling expenses

Also, cold-formed steel structural members provide structural and constructional

Sometimes, high unit-prices, low fire resistance, more careful requirement of

1.2 Aim and Layout of the Thesis

Chapter 1 includes an overwiev on Cold-Formed Steel. This chapter discusses

Main focus of Chapter 4 is on the investigation of nominal strength of various rack

2.1 General Remarks on Cold Formed Steel Column Behaviour

- Global (overall) Column Buckling including;

Flexural Buckling (bending about a principal axis)

Torsional Buckling (twisting about shear center)

Flexural-Torsional Buckling (bending and twisting simultaneously)

- Distortional buckling (buckling of open sections with edge stiffened flanges)