Structural behaviour of bolted moment connections in cold-formed steel beam-column sub-frames

Journal of Constructional Steel Research 58 (2002) 253274www.elsevier.com/locate/jcsr

Structural behaviour of bolted momentconnections in cold-formed steel beam-column

sub-framesM.F. Wong, K.F. Chung*

Department of Civil and Structural Engineering, The Hong Kong Polytechnic University,Hung Hom,Kowloon, Hong Kong, PR China

Abstract

This paper presents an experimental investigation on bolted moment connections betweencold-formed steel sections. A total of 20 column base connection tests and beam-column sub-frame tests with different connection configurations were carried out to assess the strengthand stiffness of bolted moment connections between cold-formed steel sections. Among thetests, four different modes of failure were identified:

Mode BFcsw: Bearing failure in section web around bolt hole Mode LTBgp: Lateral torsional buckling of gusset plate Mode FFgp: Flexural failure of gusset plate Mode FFcs: Flexural failure of connected cold-formed steel section

For those connections failed in Mode BFcsw, the moment resistances of the connections weretypically found to be below 50 % of the moment capacities of the connected sections. Forthose connections failed in Modes LTBgp and FFgp, the moment resistances of the connectionswere found to be about 60% and 75% of the moment capacities of the connected sections.Among all, the moment resistances of those connections failed in Mode FFcs were the highestwith a minimum of 85% of the moment capacities of the connected sections. Consequently,it is demonstrated that through rational design and construction, effective moment connectionsbetween cold-formed steel sections may be readily achieved. Engineers are encouraged tobuild light-weight low to medium rise moment frames with cold-formed steel sections. 2002Elsevier Science Ltd. All rights reserved.

* Corresponding author.E-mail address: [email protected] (K.F. Chung).

0143-974X/02/$ - see front matter 2002 Elsevier Science Ltd. All rights reserved.PII: S0143 -974X(01)00 04 4- X

254 M.F. Wong, K.F. Chung / Journal of Constructional Steel Research 58 (2002) 253274

Keywords: Bolted moment connections; Column base connection test; Failure modes

1. Introduction

Cold-formed steel sections are lightweight materials, suitable for building con-struction owing to their high structural performance and durability. They are widelyused as secondary members, such as purlins in roofs, joists of medium span in floors,studs in wall panels, storage racking in warehouses, and hoarding structures in con-struction sites. Since 1990, there has been a growing trend to use cold-formed steelsections as primary structural members in building construction, such as low tomedium rise residential houses and portal frames of modest span.

The most common cold-formed steel sections are lipped C-sections and lipped Z-sections, and the thickness typically ranges from 1.2 to 3.2 mm. Common yieldstrengths are 280 and 350 N/mm2. Moreover, there are a whole range of variants ofthese basic shapes, including sections with single and double lips, and sections withinternal stiffeners. Due to the thinness of cold-formed steel sections, local bucklingis a predominant consideration in assessing their section capacities. Furthermore, asthey are very weak in torsion, torsional flexural buckling in columns and lateraltorsional buckling in beams may be critical. There are a number of codes of practice[14] on the design of cold-formed steel structures together with complementarydesign guides and worked examples [58] to assist practising engineers.

In building construction, cold-formed steel sections are usually bolted to hot rolledsteel plates or sections to form simple and moment connections. However, despitetheir simplicity, simple connections between cold-formed steel sections have receivedrelatively little attention. An experimental experiment [9,10] was reported recentlyto study the structural performance on simple connections between cold-formed steelsections using web cleats of folded cold-formed steel strips. A complementary setof design rules was also provided in accordance with BS5950: Part 5 and Eurocode3: Part 1.3.

Much research work has been reported in the literature on the development ofmoment connections between cold-formed steel purlins in modern roof systems. Anumber of different connection configurations [11] with sleeves or overlaps werefound in various proprietary systems which offer partial continuity along the purlins.Cold-formed steel moment connections in column bases and also in beam-columnconnections were also tested and the proposed connection configurations were suit-able for portal frame construction [1216]. Besides experimental investigations onbolted connections between cold-formed steel strips [17,18], advanced finite elementmodelling using three-dimensional solid elements with material, geometrical andboundary non-linearities were also reported in the literature [1922].

It should be noted that most of the design recommendations on connections amongcold-formed steel sections concern the load carrying capacities of individual fastenerssuch as bolts, screws, rivets and spot welds. Little information on the structuralperformance of the bolted moment connections among cold-formed steel sections

255M.F. Wong, K.F. Chung / Journal of Constructional Steel Research 58 (2002) 253274

may be found in the literature. It is important to carry out physical tests to establishthe use of moment connections between cold-formed steel sections so that efficientmoment framing may be designed and built in building construction.

This paper presents the findings of an experimental investigation [2325] on thestructural performance of bolted moment connections between cold-formed steel sec-tions. A preliminary test series on four column base connections was first carriedout, and then a total of 16 beam-column sub-frame tests in two test series withdifferent connection configurations under lateral loads were performed. The investi-gation aims to assess the strength and the stiffness of connections with differentconfigurations in practical member orientations in order to establish the structuralefficiency of bolted moment connections between cold-formed steel sections for gen-eral building applications.

2. Basic configuration

The basic configuration of bolted moment connections proposed for cold-formedsteel lipped C-sections in general building applications is:

All structural members such as beams and columns are formed with two lippedC-sections back-to-back with interconnections at regular intervals.

Moment connections between beams and columns are formed with hot-rolled steelgusset plates. In general, only the column members are continuous over the con-nections.

Only the webs of lipped C-sections are bolted onto gusset plates for ease of build-ability.

Four bolts per member are used as a minimum configuration.

The connection details are rationalized after considering ease of fabrication andinstallation. In general, the proposed moment connections are not able to developfull moment capacity of the connected sections due to discontinuity of load pathsalong section flanges in the sections. Consequently, it is aimed to develop boltedmoment connections with high structural efficiency and to mobilize at least 75% ofthe moment capacities of the connected sections.

3. Scope of investigation

The research project may be divided into two main parts of investigation:

Bolted moment connections in column base testsTest Series AThis was a preliminary test series devised to establish the feasibility of the pro-posed basic configuration for bolted moment connections between cold-formedsteel lipped C-sections. A total of four column base connections were carried outwith different bolt pitches in order to compare the effectiveness of the connections.


For simplicity, thick gusset plates were used in all tests in order to ensure failurein the connected sections.

Bolted moment connections in beam-column sub-framesTest Series B1 & B2In order to investigate bolted moment connections in typical building applications,test series B1 was devised in which five internal and three external beam-columnsub-frames under lateral loads were carried out. Among the eight tests, the connec-tion configurations were similar, but the thicknesses of the gusset plates and alsothe bolt pitches were varied systemically. It was envisaged that a number of failuremodes in the gusset plates and also in the connected sections would be obtainedfor comparison.

In order to improve the structural behaviour of the bolted moment connections inbeam-column sub-frames, another test series B2 was carried out with engineeredgusset plates. Four internal and four external beam-column sub-frames were tested.

Two different section sizes of lipped C-sections in double sections back-to-backwere used in the tests, namely, C15016DS and C15020DS; the nominal yield strengthis 450 N/mm2. The moment capacities of the sections measured from four point loadtests were 16.95, and 21.36 kNm for C15016DS and C15020DS respectively. Allbolts are 16 mm in diameter and of Grade 8.8, they are all installed with a torqueof 50 Nm.

4. Test program

The general arrangement of the test set-up for column base connections is shownin Fig. 1 while that for beam-column sub-frames is shown in Fig. 2. The connectionconfigurations of all the tests are also illustrated in Fig. 1 and 2 while key parametersof the connections are summarized in Table 1. All these test specimens were con-structed according to the basic configuration with systematic variations in the connec-tion details, i.e. the bolt pitch, and the shape and thickness of gusset plates. Detailsof the test series are presented as follows.

4.1. Test Series A

A total of four column base connection tests were carried out as follows:

1. Three long columns in which the bolt pitches, S, of the connections are equal to90, 180 and 240 mm, and thus the tests are referred as Tests BC090L, BC180Land BC240L respectively.

2. One short column in which the bolt pitch of the connection is 180 mm, and thusthe test is referred as Test BC180S.

The gusset plates are 20 mm thick and the nominal yield strength is 350 N/mm2.Refer to Fig.1 for details of the connection configurations.


Fig. 1. General test setup of cold-formed steel column-base connections.

Fig. 2. General test setup of cold-formed steel beam-column sub-frames.

258 M.F. Wong, K.F. Chung / Journal of Constructional Steel Research 58 (2002) 253274Ta

ble

1Su

mm

ary

of

test

serie

sa Max

imum

Failu

reM

axim

umM

om

ent

Initi

alM

embe

rG

usse

tpla

teFa

ilure

posit

ion

of

appl

ied

mo

dem

om

ent

resis

tanc

ere

sista

nce

stiff

ness

con

nec

tion

forc

eat

0.05

(kNm/

rad)

Test

Sect

ion

(kN)

Mom

emt

Rot

atio

nra

dTh

ickn

ess

Yie

ldTh

ickn

ess

Yie

ldN

orm

alise

d

(kNm)

(rad)

(kNm/

rad)

(mm)

stre

ngth

(mm)

stre

ngth

mo

men

t(N

/mm2

)(N

/mm

2 )(kN

m)

(CFS

)(HR

S)

BC0

90L

C150

16D

S10

.60

BFc

sw/F

Fcs

13.8

30.

077

13.2

120

001.

6245

115

.89

411

12.6

90.

75*

BC1

80L

C150

16D

S11

.93

FFcs

14.4

90.

046

20

001.

6145

115

.89

411

14.0

50.

83*

BC1

80S

C150

16D

S28

.33

FFcs

13.1

70.

106

14.9

215

001.

6145

115

.89

411

12.1

10.

71*

BC2

40L

C150

16D

S13

.87

FFcs

15.9

80.

039

20

001.

6045

115

.89

411

15.5

90.

92*

S090

A1

C150

16D

S6.

08B

Fcsw

6.72

0.04

0

750

1.62

471

9.87

388

6.18

0.36

S090

A2

C150

20D

S16

.81

BFc

sw18

.59

0.07

512

.00

750

1.94

459

15.8

941

112

.13

0.57

S180

A1

C150

20D

S15

.14

BFc

sw/F

Fgp

16.7

40.

040

10

001.

9845

19.

8738

814

.68

0.69+

0.78

E180

C1C1

5020

DS

9.10

FFgp

20.1

20.

030

13

001.

9545

49.

8738

817

.91

0.84+

0.93

S180

A2

C150

20D

S20

.94

FFcs

23.1

60.

031

16

001.

9945

415

.89

411

19.7

20.

92*

E180

C2C1

5020

DS

11.0

5FF

cs24

.44

0.01

2

1750

1.94

468

15.8

941

120

.71

0.97

*

S240

A1

C150

16D

S13

.57

FFgp

15.0

10.

010

15

001.

6346

59.

8738

812

.47

0.74+

0.70

E240

C1C1

5020

DS

9.48

FFgp

20.9

70.

041

16

001.

9445

69.

8738

818

.77

0.88+

0.97

S180

D1

C150

20D

S21

.47

FFcs

23.7

50.

040

15

002.

0248

69.

9630

218

.61

0.87

*

E180

D1

C150

20D

S11

.35

FFcs

25.1

10.

043

16

002.

0647

69.

9630

219

.71

0.92

*

S180

D4

C150

20D

S17

.82

LTB

gp19

.71

0.09

917

.30

1300

2.02

490

5.96

329

13.7

80.

650.

88E1

80D

4C1

5020

DS

9.35

LTB

gp20

.68

0.13

017

.14

1300

2.02

477

5.96

329

13.6

50.

640.

87S2

40D

1C1

5020

DS

22.9

1FF

cs25

.34

0.03

6

2000

2.01

478

9.96

302

19.5

90.

92*

E240

D1

C150

20D

S10

.43

FFcs

23.0

70.

038

17

502.

0445

09.

9630

218

.66

0.87

*

S240

D4

C150

20D

S16

.82

LTB

gp18

.60

0.05

818

.08

1500

2.03

468

5.96

329

14.3

30.

670.

92E2

40D

4C1

5020

DS

7.55

LTB

gp16

.70

0.02

0

1200

2.03

482

5.96

329

13.2

30.

620.

85

aB

C:de

note

sa

colu

mn-

base

con

nec

tion

with

ate

esh

aped

guss

etpl

ate

un

derl

ater

allo

ad;S

:den

otes

anin

tern

albe

am-c

olum

nsu

b-fra

me

with

acr

oss

shap

edgu

sset

plat

eu

nde

rla

tera

llo

ad;E

:de

note

san

exte

rnal

beam

-col

umn

sub-

fram

ew

itha

te

esh

aped

guss

etpl

ate

un

der

late

ral

load

;90,

180,

240:

deno

tebo

ltpi

tche

so

f90

,180

and

240

mm

resp

ectiv

ely.

A,C

:den

ote

four

bolts

per

mem

ber

with

outc

ham

fers

ingu

sset

plat

e;D

:de

note

sfo

urbo

ltspe

rm

embe

rwith

cham

fers

ingu

sset

plat

e;1:

deno

tes

the

thic

knes

so

fgus

setp

late

at10

mm

;2:

deno

tes

the

thic

knes

so

fgus

setp

late

at16

mm

;4:

deno

tes

the

thic

knes

so

fgus

setp

late

at6

mm

;L,S

:den

ote

alo

ng

colu

mn

ofh

eigh

t150

0m

man

da

shor

tco

lum

no

fhei

ght7

50m

mre

spec

tivel

y.Th

em

easu

red

mo

men

tca

paci

ties

of

C150

16D

SG

450

and

C150

20D

SG

450

are

16.9

5kN

man

d21

.36

kNm

resp

ectiv

ely

afte

rn

orm

alise

dto

resp

ectiv

ede

sign

thic

knes

san

dyi

eld

stre

ngth

.*

Den

otes

con

nec

tions

with

high

stru

ctur

alef

ficie

ncy.

+D

enot

espr

e-m

atur

efle

xu

ralf

ailu

reat

sect

ions

with

defe

ctiv

ecr

acks

.


4.2. Test Series B1

A total of eight tests with three different connection configurations in both internaland external beam-column sub-frames were carried out as follows:

1. Three internal beam-column sub-frame tests in which the bolt pitches, S, of theconnections are equal to 90, 180 and 240 mm, and thus the tests are referred toas Tests S090A1, S180A1 and S240A1 respectively. The gusset plates are crossshaped and 10 mm thick.

2. Two external beam-column sub-frame tests in which the bolt pitches, S, of theconnections are equal to 180 and 240 mm, and thus the tests are referred to asTests E180C1 and E240C1 respectively. The gusset plates are tee shaped and10 mm thick.

3. Two internal beam-column sub-frames in which the bolt pitches, S, of the connec-tions are equal to 90 and 180 mm, and thus the tests are referred to as TestsS090A2 and S180A2 respectively. The gusset plates are cross shaped and 16mm thick.

4. One external beam-column sub-frame test in which the bolt pitch, S, of the con-nection is equal to 180 mm, thus the test is referred to as Test E180C2. The gussetplate is teeshaped and 16 mm thick.

The nominal yield strength of the gusset plates is 350 N/mm2. Refer to Fig. 2 fordetails of the connection configurations.

4.3. Test Series B2

In order to enhance the structural performance of the connections, chamfers of 50mm in depth are provided in all the gusset plates for improved structural efficiency.Furthermore, for gusset plates with different shapes and thicknesses, two bolt pitchesof 180 and 240 mm are used for comparison. Consequently, another test series witha total of eight beam-column sub-frame tests in four different connection configur-ations with engineered gusset plates were carried out as follows:

1. Two internal beam-column sub-frame tests with cross shaped gusset plates 10mm thick, namely, Tests S180D1 and S240D1 respectively.

2. Two external beam-column sub-frame tests with tee shaped gusset plates 10 mmthick, namely, Tests E180D1 and E240D1 respectively.

3. Two internal beam-column sub-frame tests with cross shaped gusset plates 6mm thick, namely, Tests S180D4 and S240D4 respectively.

4. Two external beam-column sub-frame tests with tee shaped gusset plates 6 mmthick, namely, Tests E180D4 and E240D4 respectively.

The nominal yield strength of the gusset plates is 275 N/mm2. Refer to Fig. 2 fordetails of the connection configurations.


4.4. Test procedures

In all tests, the applied load and the displacements of each member of the testspecimens were measured during the entire deformation history. In order to assessend rotations of the members, two transducers were used at each location of interest,and member end rotations were obtained as the relative displacements divided bythe separation between the transducers. For beam-column tests, any out-of-planedeformation of the test specimens was prohibited by two sets of restraining rollers.

In general, the tests were terminated when excessive deformation, section failureor member buckling occurred. In most cases, a pre-load of 2 kN was applied to thetest specimens before testing in order to ensure that all bolts were in contact withthe section webs of connected sections despite all the bolt holes were perfect-fittedto 16 mm diameter bolts.

5. Test results

Table 1 summaries the results of Test Series A, B1 and B2 together with themeasured thickness and yield strength of both the cold-formed steel sections and thegusset plates. Furthermore, all the measured moment resistances are evaluated atfailure positions and normalized with the ratio of design yield strength and designthickness to measured yield strength and thickness of the test specimens. For testspecimens with excessive deformation under testing, the moment resistances of theconnections are restricted to be the applied moment at a connection rotation of0.05 rad.

Among all 20 tests, four different modes of failure were identified as follows:

Mode BFcsw: Bearing failure in section web around bolt hole, as shown in Fig. 3

Fig. 3. Mode BFcsw. Bearing failure in section web around holes.


Fig. 4. Mode FFcs. (a) Flexural failure of connected section in a column-base connection test. (b) Flex-ural failure of connected cold-formed steel section in an external beam-column sub-frame test.

Mode LTBgp: Lateral torsional buckling of gusset plate, as shown in Fig. 4 Mode FFgp: Flexural failure of connected gusset plate Mode FFcs: Flexural failure of connected cold-formed steel section, as shown in

Fig. 5

Fig. 6 illustrates the locations of critical sections for various failure modes in the


Fig. 5. Mode LTBgp. (a) Lateral torsional buckling of connected gusset plate in an (a) internal bean-column sub-frame test (b) external beam-column sub-frame test.

connections. The measured loaddeflection curves of test specimens in Test SeriesA, B1 and B2 are presented in Fig. 7, 8 and 9 respectively, and they are plotted onthe same graph for each test series for easy comparison.

In order to allow direct comparison, the applied moments for all of the test speci-mens are normalized with respect to the measured moment capacities of the connec-


Fig. 6. Locations of critical sections in Test Series B1 and B2.

ted sections. After evaluation of member rotations, joint rotations are then obtainedas the difference of the member end rotations of those connected members. Forinternal beam-column sub-frames, four different joint rotations may also be evaluatedfrom the connections, depending on the calculation. Similarly, two different jointrotations may be obtained for external beam-column sub-frames. Typical momentrotation curves of all the test specimens in Test Series A, B1 and B2 are presentedin Fig. 10, 11 and 12 respectively.


Fig. 7. Loaddeflection curves of column-base connections.

5.1. Moment resistance ratios

In order to assess the effectiveness of the bolted moment connections, a momentresistance ratio, , is established which is defined as follows:

Measured moment resistance of a connection

Measured moment capacity of connected section

For connections with equal to unity, the connections are shown to be capableof mobilizing the full moment resistances of the connected sections. For connectionswith less than 0.50, the connections are regarded to be inefficient in transmittingmoment across. All the moment resistances of the connections are evaluated at thefailure positions of the connections, and thus they may be used directly in connectiondesign. The results of all the tests are compared among each other and the findingsare presented in the following sub-sections.

5.2. Test Series A

1. Tests BC090L, BC180L and BC240LIn test BC090L, significant bearing deformation was observed in the bolt holesof the section webs due to high moment acting at small lever arms. In both testsBC180L and BC240L, flexural failure in connected sections was apparent. Themoment resistance ratios, at the failure positions of the connections were foundto be 0.75, 0.83 and 0.92 in tests BC090L, BC180L and BC240L respectively.

2. Tests BC180L and BC180SIn both tests, flexural failure in connected sections was apparent. The moment


Fig. 8. Loaddeflection curves of (a) internal and (b) external beam-column sub-frames

resistance ratios, , at the failure positions of the connections were found to be0.83 and 0.71 in tests BC180L and BC180S respectively.

3. Consequently, it is shown that for connections with small bolt pitches, the struc-tural efficiency of the connections is low with bearing failure in section webaround bolt holes. For connections with large bolt pitches, flexural failure of con-nected cold-formed steel sections is critical, and the corresponding moment resist-ance ratio of the connections, , is at least 0.83. This is regarded as a favourablemode of failure with high structural efficiency.


Fig. 9. Loaddeflection curves of beam-column sub-frame tests with (a) 10 mm thick and (b) 6 mmthick gusset plates.

5.3. Test Series B1

1. Tests S090A1, S180A1 and S240A1In test S090A1, significant bearing deformation was observed in the bolt holes ofthe section webs due to high moment acting at small lever arms. In both testsS180A1 and S240A1, gross bending deformation was apparent in the hot-rolledsteel gusset plates. The gusset plates failed prematurely in flexure due to the pres-ence of defects at the corners of the gusset plates. The moment resistance ratios,, at the failure positions of the connections were found to be 0.36, 0.69 and0.74 in tests S090A1, S180A1 and S240A1 respectively.


Fig. 10. Typical moment rotation curves of column-base connections.

2. Tests E180C1 and E240C1In both tests, gross bending deformation was apparent in the hot-rolled steel gussetplates, and the gusset plates failed prematurely in flexure due to the presence ofdefects at the corners of the gusset plates. The moment resistance ratios, , atthe failure positions of the connections were found to be 0.84 and 0.88 in testsE180C1 and E240C1 respectively.

3. Tests S090A2, S180A2 and E180C2In test S090A2, significant deformation was observed in the bolt holes of sectionwebs due to high moment acting at small lever arms. For both tests S180A2 andS240A2, flexural failure in connected cold-formed steel sections was apparent.The moment resistance ratios, , at the failure positions of the connections werefound to be 0.57, 0.92 and 0.97 in tests S090A2, S180A2 and E180C2 respectively.

4. Comparison among tests S090A1, S180A1 and S240A1 shows that for internalbeam-column sub-frames with 10 mm thick gusset plates, the connection resist-ance ratios are increased from 0.36, 0.69 and 0.74 when the bolt pitch is increasedfrom 90 to 180 mm and then to 240 mm. Furthermore, comparison among testsE180C1 and E240C1 also shows that for external beam-column sub-frames with10 mm thick gusset plates, the connection resistance ratios are also increasedwhen larger bolt pitch is used.

It is thus demonstrated that connections with large bolt pitch always have highermoment resistances. Moreover, with an increase in the moment resistances, flex-ural failure in gusset plates becomes more critical than bearing failure in connectedsection web.

5. By increasing the thickness of gusset plates from 10 to 16 mm, the moment resist-ance ratios of the proposed connection configurations are found to be increasedfrom


Fig. 11. Typical moment rotation curves of (a) internal and (b) external beam-column sub-frames.

0.36 to 0.57 as shown in tests S090A1 and S090A2, 0.69 to 0.92 as shown in tests S180A1 and S180A2, and 0.84 to 0.97 as shown in tests E180C1 and E180C2.This shows that thicker gusset

plates always give higher moment resistances. Moreover, with an increase in themoment resistances of the gusset plates, flexural failure in the connected cold-formed steel sections becomes critical instead of flexural failure of the gusset plat-es.

The maximum moment resistance of the proposed connection configuration isfound to be over 90% of the moment capacities of the connected sections, demon-


Fig. 12. Typical moment rotation curves of beam-column sub-frame tests with (a) 10 mm and (b) 6 mmthick gusset plates.

strating that the proposed connection configuration is effective in transferringmoment across the connected members. As small bolt pitch is found to give connec-tions of low moment resistance and large member rotation, it is thus recommendedthat the minimum value of bolt pitch for moment connections should not be lessthan 150 mm and 1.0 times section depth.

5.4. Test Series B2

1. Tests S180D1 and E180D1For both tests, flexural failure in connected sections was apparent. The moment


resistance ratios, , at the failure positions of the connections were found to be0.87 and 0.92 in tests S180D1 and E180D1 respectively.

2. Tests S180D4 and E180D4In both tests, lateral torsional buckling in connected hot-rolled steel gusset platewas apparent. The moment resistance ratios, , at the failure positions of theconnections were found to be 0.65 and 0.64 in tests S180D4 and E180D4 respect-ively.

3. Tests S240D1 and E240D1For both tests, flexural failure in connected cold-formed steel sections was appar-ent. The moment resistance ratios, at the failure positions of the connectionswere found to be 0.92 and 0.87 in tests S240D1 and E240D1 respectively.

4. Tests S240D4 and E240D4In both tests, lateral torsional buckling in connected hot-rolled steel gusset platewas apparent. The moment resistance ratios, , at the failure positions of theconnections were found to be 0.67 and 0.62 in tests S240D4 and E240D4 respect-ively.

5. For connections with thick gusset plates, it is shown that flexural failure of connec-ted cold-formed steel sections is always critical, and the corresponding momentresistance ratio of the connections, , is at least 0.87. This is regarded as a favour-able mode of failure with high structural efficiency.

For connections with thin gusset plates, lateral torsional buckling of hot rolledsteel gusset plates is always critical, and the corresponding moment resistanceratio, , of the connections is about 0.60. This failure mode is regarded to beinefficient with low structural efficiency.

6. For connections with similar configurations but with different bolt pitches, it isshown that there is little difference in the moment resistance ratios. Thus, the boltpitch of 180 mm may be considered to be effective in forming moment connec-tions for 150 mm deep lipped C-sections.

6. Relative performance on connections with different failure modes

1. The bearing failure in section web around bolt hole, or Mode BFcsw, alwaysoccurs at low applied load, and thus this failure mode is unfavourable as themoment resistances of the connections are relatively low with large memberrotation. Due to progressive yielding of material in section web under confinedbearing, this mode of failure may be considered to be ductile.

2. For connections with thin gusset plates, lateral torsional buckling, or ModeLTBgp, may be critical and thus the moment resistances of the connections arealso low. In practice, this failure mode depends largely on the unrestrained lengthof the gusset plates between the beam and the column members, and also theprovision of external restraints to the connections. It should be noted that thisfailure mode always takes place in a sudden manner with quick unloading.

3. For connections with thick and effectively restrained gusset plates, two differentfailure modes may occur, depending on the relative moment capacities of the


gusset plates and the connected cold-formed steel sections. For those gusset plates with limited moment capacities, flexural failure of theconnected gusset plates, or Mode FFgp, is critical at the location subjected to thelargest applied moment, i.e. at the corners of both cross and tee shaped gussetplates close to column members. In general, this failure mode is ductile withsignificant member rotation after failure. However, due to the presence of defectsaround the corners of gusset plates in some tests, premature failure with littleductility was observed. For those gusset plates with large moment capacities, flexural failure of the

connected cold-formed steel sections, or Mode FFcs, is critical at the end ofthe gusset plates where the beam members are connected. Due to limited post-buckling strength in cold-formed steel sections after yielding, there is no appar-ent rotation capacity in the connection after failure.

4. Among all the failure modes observed in the tests, flexural failure of connectedcold-formed steel sections is the most desirable one as the moment resistances ofthe connections are found to be very close to those of the connection sections.

7. Overall comments

It is interesting to compare the moment resistance ratios of all the twenty connec-tions and to group them together according to their critical failure modes; Table 2presents the summary of the comparison. It should be noted that:

1. A total of 10 connections with different configurations are able to develop momentresistances over 75% of the moment capacities of the connected sections withflexural failure in connected cold-formed steel sections; all of them are, thus,regarded as effective moment connections.

2. The presence of 50 mm deep chamfers is found to be effective in gusset plates

Table 2Structural performance of bolted moment connections

Failure Moment resistance ratio Number of tests Testmode

BFcsw 0.5 2 S090A1/S090A2

LTBgp 0.6 4 S180D4/E180D4S240D4/E240D4

FFgp 0.7 0.8 4 S180A1/E180C1S240A1/E240C1

FFcs 0.750.95 10 BC090L/BC180L/BC180S/BC240LS180A2/E180C2S180D1/E180D1S240D1/E240D1


10 mm thick as their structural behaviour is similar to those gusset plates 16 mmthick but without chamfers. In both cases, the moment resistance ratios of theconnections are found to be over 90% with flexural failure in connected cold-formed steel sections.

3. There is little difference in the moment resistances of connections with bolt pitchesof 180 and 240 mm. Thus, a bolt pitch of 180 mm may be considered to beeffective for 150 mm deep lipped C-sections.

4. While the member rotations are measured with high accuracy, caution should beused in the interpretation of the measured rotations. As the test specimens arerelatively flexible, deformations of the test specimens are always significant beforefailure. Thus, the reference points for member rotations are moved continuouslyaccording to the lateral displacements of the test specimens since the transducersare stationary. Consequently, the measured member rotations may only be con-sidered as an approximation to the actual member rotations of the test speci-mens.Moreover, it is difficult to define precisely the centres of rotation of themembers within the connections. Any measured rotations based on deflections ofthe section flanges may not necessarily correspond to rotations of the whole sec-tion in the presence of cross-sectional distortion, especially under high stress lev-els. In general, the error is considered to be small during the initial deformationstage, and the maximum discrepancy in joint rotations is envisaged to be within15% at large lateral displacements

8. Conclusions

In order to establish moment framing for cold-formed steel sections in buildingapplications, a bolted moment connection configuration for double lipped C sectionsback-to-back was proposed. A total of 20 column base connections and beam-columnsub-frames tests were executed and four different modes of failure were identified.The moment resistances of the proposed connection configurations were found torange from below 50% to over 85% of the moment capacities of connected sections.

Based on the findings of the experimental investigation, it is demonstrated thatbolted moment connections between cold-formed steel sections are readily achievedwith the proposed connection configurations. The bolted moment connections areshown to be effective in transmitting moment between the connected sections,enabling effective moment framing in cold-formed steel structures. Engineers areencouraged to build lightweight low to medium rise moment frames with cold-for-med steel sections.

Acknowledgements

The research project leading to the publication of this paper is supported by theResearch Grants Council of the Hong Kong Government of the Special Administrat-


ive Region (Project No. PolyU5031/98E), and also by the Research Committee ofthe Hong Kong Polytechnic University Research (Project No. G-V750).

All the test specimens were supplied by the P & Ls Engineering Co. Ltd. Thetests were carried out at the Heavy Structure Laboratory of the Department of Civiland Structural Engineering, the Hong Kong Polytechnic University. The authorswould like to express their gratitude to Mr W.K. Yu, Mr K.W. Chow, Mr S.Y. Liuand Mr Y.K. Wong, and also to the technicians of the Heavy Structure Laboratoryfor the assistance during the execution of the tests.

References

[1] AISI specification for the design of cold-formed steel structural members. Washington (DC): Amer-ican Iron and Steel Institute, 1996.

[2] Cold-formed steel structure code AS/NZ 4600: 1996. Sydney: Standards Australia/Standards NewZealand, 1996.

[3] Eurocode 3: design of steel structures: part 1.3: general rules supplementary rules for cold-formedthin gauge members and sheeting, ENV 1993-1-3. Brussels: European Committee for Standardiz-ation, 1996.

[4] BS5950: structural use of steelwork in buildings: part 5: code of practice for the design of cold-formed sections. London: British Standards Institution, 1998.

[5] Rhodes J. Design of cold-formed steel members. Oxford: Elsevier, 1991.[6] Chung KF. Building design using cold-formed steel sections: worked examples to BS5950: part 5:

1987. The Steel Construction Institute, 1993.[7] Hancock GJ. Design of cold-formed steel structures. 3rd ed. Australian Institute of Steel Construc-

tion, 1998.[8] Yu WW. Cold-formed steel design. 3rd ed. New York: John Wiley and Sons Inc, 2000.[9] Chung KF, Godwin MJ, Griffiths DA. Connections for structural cold-formed sections. Commission

of European Communities, ECSC Sponsored Research Project: Directorate General XII Science,Research and Development, British Steel Technical, FR W302-5 922, May 1993.

[10] Chung KF, Lawson RM. Structural performance of shear resisting connections among cold-formedsteel members using web cleats of cold-formed steel strips. Engng Struct 2000;22:135066.

[11] Trahair NS. A survey of tests on cold-formed purlins. In: International Workshop on Cold-formedSteel Structures, University of Sydney, Feb 1993.

[12] Baigent AH, Hancock GJ. The behaviour of portal frames composed of cold-formed members. In:Rhodes J, Walker AC, editors. Thin walled structures. Granada, 1980.

[13] Zhao XL, Hancock GJ. T-joints in rectangular hollow sections subject to combined actions. J StructEngng ASCE 1991;117(8).

[14] Wheeler AT, Clarke MJ, Hancock GJ. Tests of bolted flange plate connections joining square andrectangular hollow sections. In: Proceedings of the Fourth Pacific Structural Steel Conference, Singa-pore, 1995:9704.

[15] Wilkinson T, Hancock GJ. Tests of knee joint in cold-formed rectangular hollow sections. Researchreport no. R779, Department of Civil Engineering, the University of Sydney, Nov. 1998.

[16] Chung KF, Lau L. Experimental investigation on bolted moment connections among cold-formedsteel members. Engng Struct 1999;21:898911.

[17] Toma A, Sedlacek G, Weynard K. Connections in cold-formed steel. Thin Wall Struct1993;16:21937.

[18] Bryan ER. The design of bolted joints in cold-formed steel sections. Thin Wall Struct. Thin WallStruct 1993;16:239.

[19] Chung KF, Ip KH. Finite element modelling of cold-formed steel bolted connections. In: Proceedingsof the Second European Conference on Steel Structures, Prague, May 1999: 5036.


[20] Wheeler AT, Clarke MJ, Hancock GJ. Finite element modelling of eight bolt rectangular hollowsection bolted moment end plate connections.In: Proceedings of the Second International Conferenceon Advances in Steel Structures, Hong Kong, Dec 1999: 2374

[21] Ip KH, Chung KF. Failure modes of bolted cold-formed steel connections under static shear loading.In: Proceedings of the Second International Conference on Advances in Steel Structures, Hong Kong,Dec 1999: 2696.

[22] Chung KF, Ip KH. Finite element modelling of bolted connections between cold-formed steel stripsand hot rolled steel plates under shear. Engng Struct 2000;22:127184.

[23] Wong MF, Chung KF. Experimental investigation of cold-formed steel beam-column sub-frames:Pilot Study. In: Proceedings of the Fifteenth International Specialty Conference on Cold-FormedSteel Structures, St. Louis (Missouri, USA), Oct 2000:60718.

[24] Wong MF, Chung KF. Experimental investigation of cold-formed steel beam-column sub-frames:Comparative study. In: Proceedings of the First International Conference on Structural Stability andDynamics, Taipei, Dec 2000:58792.

[25] Chung KF, Wong MF. Experimental investigation of cold-formed steel beam-column sub-frames:enhanced performance. In: Proceedings of the First International Conference on Structural Engineer-ing, Mechanics and Computation, Cape Town (South Africa), Apr 2001; p. 151522.

Structural behaviour of bolted moment connections in cold-formed steel beam-column sub-frames

Documents

Transcript of Structural behaviour of bolted moment connections in cold-formed steel beam-column sub-frames