Вы находитесь на странице: 1из 120

STRESS - STRAIN RELATIONSHIPS

FOR CONFINED CONCRETE : RECTANGULAR SECTIONS

A report submitted in partial fulfilment of


the requirements for the degree of Master of
Engineering at the University of Canterbury,
Christchurch~ New Zealand.

by

BRYAN D. SCOTT

February 1980
\
. '\
~

I
.
.,. I
'
I

'
J,
'l

.I
{
I
i

ABSTRACT

An experimental investigation into the behaviour of square, confined,


reinforced concrete columns was undertaken. Thirty 450 mm square, 1200 mm
high units were cast with varying amounts of longitudinal and lateral
steel. These were subjected to concentric or eccentric axial loads to
failure at slow or dynamic loading rates.

Confinement requirements of reinforced concrete columns are discussed and


the results and analyses of experimental work presented.

Results include an assessment of the significance of loading rate,


eccentricity, amount and distribution of longitudinal steel, and the
amount of confining steel.

A general stress-strain curve for rectangular concrete sections loaded at


seismic rates is proposed and compared with existing curves based on
previous static loading tests.
ii

ACKNOWLEDGEt~ENTS

The research for this report was carried out in the Department of Civil
Engineering, University of Canterbury, under the overall guidance of its
Head, Professor R. Park.

The project was supervised by Professor R. Park and Dr. M.J.N. Priestley,
both of whom have given invaluable advice, and careful guidance. My
thanks also to Professor S.M. Uzumeri (Toronto, Canada) for his introduct-
ion to this topic.

I wish to thank Mr. N.W. Prebble, Technical Officer, and all the technical
staff of the Civil Engineering Department who have been associated with
this project. Special thanks are due to Mr. A. Bell, Mr. G. Hill, and
Mr. G. Clark for their contribution towards constructing, preparing and
testing the specimens.

The financial assistance of the National Roads Board is gratefully


acknovJl edged.

Thanks are also due to Mrs. V. Grey for her tracing and to Mrs. C. Gaerty
for typing this text.

The special care and attention given by Ashby Bros. Ltd. to obtaining
uniform concrete properties is gratefully acknowledged.

Finally a special thanks to my vJife Gaile, for her support and encourage-
ment.
iii

TABLE OF CONTENTS
Page:

ABSTRACT i
ACKNOWLEDGEMENTS
TABLE OF CONTENTS Hi
LIST OF FIGURES Vi'
LIST OF TABLES ix
NOTATION X

CHAPTER ONE~ : .INTRODUCTION


1.1 General 1
1.2 Aim 2
1.3 Scope 2
1.4 Format 2
1.5 Previous Research 2
1.5.1 Chan (1955) 4
1.5.2 Roy and Sozen (1964) 4
1. 5. 3 Bertero and Fe l i ppa (1964) 7
1.5.4 Soliman and Yu (1967) 7
1.5.5 Kent and Park (1971) 9
1.5.6 Sargin (1971) 11
1.5.7 Vallenas, Bertero and Popov (1977) 12
1.5.8 Sheikh and Uzumeri (1978) 15
1.5.9 Modified Kent and Park (1979) 17
1. 5. 10 Summary 18

CHAPTER TWO : CONFINEMENT REQUIREMENTS FOR PLASTIC HINGE ZONES


IN REINFORCED CONCRETE COLUMNS
2.0 Summary 20
2.1 The Codes Considered 20
2.1.1 ACI 318-77 21
2.1.2 SEAOC (1975) 21
2.1.3 ATC (1978) 22
2.1.4 ACI Committee 343 22
2.1.5 Japanese Practice 22
2.1.6 Ministry of Works and Development, Civil Division 22
2.1.7 Draft SANZ Concrete Design Code 23
2.2 Comparison of Code Requirements for A Typical Rectangular
Column. 25
iv

2.3 Conclusions 27

CHAPTER THREE : DESIGN AND CONSTRUCTION OF TEST UNITS


3.0 Summary 28
3.1 Unit Size Criteria 28
3.2 Design of Test Units 28
3.2.1 Longitudinal Steel 28
3.2.2 Hoop Steel 29
3.3 Material Properties 32
3.3.1 Steel 32
3.3.2 Concrete 32
3.4 Construction 40

CHAPTER FOUR INSTRUMENTATION AND TESTING PROCEDURE


4.0 Summary 42
4.1 Instrumentation 42
4.1.1 Load Measurement 42
4.1.2 Longitudinal Concrete Strains 42
4.1.3 Hoop Reinforcement 43
4.2 Testing Procedure 44
4.2.1 Test Unit Preparation 44
4.2.2 Testing Procedure 45

CHAPTER FIVE : TEST RESULTS


5.0 Summary 47
5.1 General Behaviour and Visual Observations 47
5.2 Presentation of Results 56
5.3 Discussion of Results 87
5.3.1 Rate of Loading 87
5.3.2 Confinement Ratio 92
5.3.3 Distribution of Longitudinal Steel 97
5.3.4 Ultimate Compression Strain 9Z
5.3.5 Strength of Longitudinal Steel 100
5.3.6 Eccentricity of Loading 100
v

CHAPTER SIX : CONCLUSIONS AND FUTURE RESEARCH


6.0 Summary 101
6.1 Conclusions 101
6.2 Recommendations for Future Research 102

APPENDIX A : REFERENCES 103


vi

LIST OF FIGURES

Figure Title Page:


1.1 Effect of Tie Spacing on Stress~Strain Relationship
of Concrete (Roy and Sozen (1963)(10} 5
1.2 Stress-Strain Curve for Concrete Confined by (
Rectangular Ties as Proposed by Roy and Sozen (1964) l 1 ) 6
1.3 Stress-Strain Relationship of Confined Concrete in
Flexure (Soliman and Yu 19671 (151 8
1.4 Stress-Strain Curve for Concrete Confined by
Rectangular Hoops (Kent and Park 1971] (18) 10
1.5 Comparison of Analytical Curves with Experimental
Results (Vallenas et al, 19771 (24} 13
1.6 Confined Concrete with Longitudinal Reinforcement -
Analytical Curve and its Comparison with Experimental
Results (Vallenas et al, 19771 (24} 13
1.7 Stress-Strain Curve for Confined Concrete in Square
Columns (Sheikh and UzU!l)eri 19781 (2) 16
1.8 Stress-Strain Curve for Concrete Confined by
Rectangular Hoops (Park, Priestley and Gill? 1979}(29} 17
2.1 Comparison of Code Hoop Steel Requirements for a
Square Column (Park and Priestley 1979} (30) 26
3.1 Detai 1s of Test Units and Transverse Reinforcement 31
3.2 Stress.,..Strain Curve for DH24 Steel (Grade 380) 33
3.3 Stress.-Strain Curve for DH20 Steel (Grade 380) 34
3.4 Stress.-Stra in Curve for 020 Steel (Grade 275) 35
3.5 Stress-Stra i. n Curve for R10 Hoop Steel (Grade 275} 36
3.6 Stress-Strain Curve for Rl2 Hoop Steel (Grade 275) 37
3.7 Strength-Age Diagram of 200 x 100 mm Diameter Cylinders 38
3.8 Photographs of the Construction Sequence 41
4.1 Strain Gauge Locations 43
4.2 Preparation of Test Units, Testing Machine and 46
Recording Instruments
5.1 Bending of the Support Bars 49
5.2 Unit 12 Photographs Showing Test Sequence 51
5.3 Unit 15 Photographs Showing Test Sequence 52
vii

5.4 Photographs Showing Selected Features of Failure 53


for Various Test Units
5.5 Photographs Showing Selected Features of Failure 54
for Various Test Units
5.5(e) Mid-Height lateral Displacement Versus Average 55
Longitudinal Strain
5.6 Unit 1 Axial Load, Slow Speed 58
5.7 Unit 2 Axial Load, Slow Speed 59
5.8 Unit 3 Axial Load, High Speed 60
5.9 Unit 4 Eccentric Load, Slow Speed 61
5.10 Unit 4 Eccentti c Load~ Sl O'V'I Speed 62
5.11 Unit 5 Eccentric Load, High Speed 63
5.12 Unit 5 Eccentric Load, High Speed 64
5.13 Unit 6 Axial Load, Slow Speed 65
5.14 Uni. t 7 Axial Load, High Speed 66
5.15 Unit 8 Eccentric Load? Slow Speed 67
5.16 Unit 8 Eccentric Load~ Slow Speed 68
5.17 Unit 9 Eccentri. c Load, Hi'gh Speed 69
5.18 Unit 9 Eccentric Load~ Hi.gh Speed 70
5.19 Unit 11 Axi.a l Load~ High Speed 71
5.20 Unit 12 Axial Load, High Speed 72

5.21 Unit 13 Axial Load 2 High Speed 73


5.22 Unit 14 Axial Load, High Speed 74
5.23 Unit 15 Axial Load, High Speed 75
5.24 Unit 17 Axi.al Load? High Speed 76
5.25 Unit 18 Axial Load, Hi.gh Speed 77

5.26 Unit 19 Axial Load, High. Speed 78


5.27 Unit 20 Axial Load, High Speed 79
5.28 Unit 21 Axia 1 Load, Slow Speed 80
5.29 On it 22 Axi a1 Load 1 High Speed 81
5.30 Unit 23 Axial Load, Kigh Speed 82
Vi'i i

5.31 Unit 24 Axial Load, High Speed 83


5.32 Unit 25 Axial Load, High Speed 84
5.33 Unit 26 Axial Load, High Speed 85
5.34 Unit 27 Axial Load, Medium Speed 86
5.35 Plain Concrete Units Loaded at Different Rates 89
5.36 8 Bar Units Loaded at Different Rates 90
5.37 12 Bar Units Loaded at Different Rates 91
5.38 Effect of Confinement Ratio for an 8 Bar Unit 94
5.39 Effect of Confinement Ratio for a 12 Bar Unit 95
5.40 Strength Increase Versus Confinement Ratio 96
5.41 Confinement Due to Distribution of Longitudinal Steel 98
ix

LIST OF TABLES

Table Title
1.1 Summary of Tests Reported by Different Researchers.
(After Sheikh (1978) (21 3' 4
3.1 Hoop Bar Diameter and Spacing of Hoop Sets 30

3.2 Yield and Ultimate Steel Stresses 32


3.3 Test Unit Properties 39

4.1 Calculated Eccentricities 45


5.1 Summary of Results 48

5.2 The Effect of Loading Rate on Peak Stress 88


X

NOTATION

= area of concrete section confined by hoops


= gross area of section
= area of longitudinal steel
= area of rectangular hoop bar (one leg only)
= total effective area of hoop bars and supplementary cross
ties in the direction under consideration within spacing sh
b or b = core dimension
c = neutral axis depth at ultimate
cl = centre to centre distance between the longitudinal bars
d = effective depth of section
dl nominal diameter of longitudinal bar
dll = nominal diameter of transverse feinforcement
/5, = displacement
Ec = modulus of elastici'ty of concrete
E: = strain
Ec = concrete strain
=:. ultimate concrete
E:
cu strain
Esl,Es2 = minimum and maximum average 1ongitudina 1 strains
corresponding to the maximum stress in concrete (2}
E:y = steel yield strain
Eo,Eoo = concrete strain at maximum stress level
E20c = concrete strain at 20% maximum stress (18)
E50c = concrete strain at 50% maximum stress for confined
concrete (18)
= E50c - E50u (18)
E:
50u = concrete strain at 50% maximum stress for unconfined
concrete (18)
fc = stress in concrete
fl =concrete cylinder strength
c
f's = stress in transverse steel
fy = yield stress in longitudinal steel
fyh = yield stress of transverse steel
h = depth of full section
1cr = cracked section modulus
k = maximum stress ratio (24)
Ks = ratio of the strength of confined concrete to the strength
of plain concrete (2)
xi

m = cover concrete ratio for confined concrete


M = moment
Jl = displacement ductility factor
n = number of longitudinal bars in the specimen
NA = neutral axis
pe = axial load due to gravity and seismic loading
Po = 0.85 f~ (Ag - Ast) + fy Ast
pace = 0.85 f~ (Ac - As) (12}
Ps = volumetric transverse steel ratio

Pt = longitudinal steel percentage


= capacity reduction factor
1> = curvature
1>
u = section curvature at ultimate
ipy = section curvature at yield
s or sh = spacing of hoop sets
z = slope of falling branch of concrete stress strain curve (18)
1

CHAPTER ONE

INTRODUCTION

1.1 GENERAL

For moment-curvature analysis of structural members and systems it is


usually necessary to model the behaviour of the materials used. For
reinforced concrete structures under monotonic loading suitable models
exist for unconfined concrete and for steel, but limited information is
available for concrete confined by transverse reinforcement.

The stress-strain curve for unconfined concrete is well known and


generally accepted to finish at Ecu = 0.003, when crushing occurs.
However under seismic loading high ductilities are often demanded of
structures, which require ultimate concrete strains much greater than
Ecu = 0.003. These strains and ductilities can be achieved by providing
longitudinal and transverse reinforcement to effectively confine the core
concrete. The confinement is provided by allowing the concrete to arch
stirrup to stirrup vertically and bar to bar horizontally which is often
assumed to confine the core as if by an equivalent uniform lateral fluid
pressure.

A knowledge of the stress-strain curve for confined concrete is partic-


ularly important for columns with high axial load levels, when the moment
curvature characteristics of the column are largely dependent on the
concrete compressive strength and the stress strain relationship.

Early research on confined concrete was (generally) on small scale,


concentrically and monotonically loaded units, often without cover or
longitudinal reinforcement. The testing was generally carried out in load
controlled testing machines at slow loading rates.

Behaviour under these conditions has been used to predict behaviour under
seismic conditions, which are characterised by displacement control,
rapid loading rates, repeated load application, and eccentricity of
loading, Recently more realistically sized units have been used but not
under simulated seismic conditions.
2

1.2 AIM

The aim of the study of this report was to examine Experimentally the
confinement requirements of Chapter 17 of DZ 3101( 1), "Members subjected
to flexure and axial load- additional seismic requirements", in order to
further understand the behaviour of confined concrete in rectangular
reinforced concrete columns in earthquake risk areas. To minimise
interpretation problems inherant in extrapolating previous test data to
seismic conditions, the study aimed at testing near full size models at
rapid loading rates in a displacement controlled testing machine to more
closely simulate seismic conditions.

1.3 SCOPE

Thirty large scale, about half to two-thirds full size, square column
sections were designed to the revised provisions of Chapter 17 of DZ 3101,
and confined to four different axial load levels, nominally 0.1, 0.25,
0.4, 0.7 off~ Ag. These were subjected to concentric or eccentric axial
loads to failure at slow or dynamic loading rates.

1.4 FORMAT

The next section contains a brief review of previous research carried out
in the field of the stress-strain behaviour of confined concrete.
Chapter Two examines the various code provisions for confining steel.
Chapter Three outlines the design and construction of the test units along
with the properties of the materials used, while Chapter Four describes
the instrumentation and testing procedure.

Test results are presented in Chapter Five in the form of load, confined
concrete core stress/cylinder strength and hoop steel stress plotted
against longitudinal strain. The results are summarised in Table 5.1
Trends in the results are also compared and discussed in Chapter Five, A
summary and conclusions are given in Chapter Six along with suggestions
for the direction of future research.

References are listed in Appendix A.

1.5 PREVIOUS RESEARCH

Theoretical and experimental research into the behaviour of confined


concrete has been conducted by many researchers (References 2-28) at
3

various institutions throughout the world. Comprehensive literature


surveys have been collated recently by Leslie (1974)( 25 ) and Sheikh (1978)
(2) A summary of the work reviewed by Sheikh( 2 ) is given in Table 1.1.
A brief description of the salient points from the more important
researchers, based on Sheikh s survey( 2) is presented below. The work
1

summarized will concern mainly square or rectangular column units confined


by square or rectangular hoops. This report will not discuss results for
spiral column tests.

TABLE 1.1 : Summary of the Tests Reported by Different Researchers


(Sheikh(l978)( 2 )

Details of the Specimens


Researcher Size of Acore Longitudinal
Number the section Steel
mm Agross
King (1946) 164 89 X 89 0.54-0.61 4 corner bars
King (1946) 18 254 X 254 1. 34-0.66 li

Chan {1955) 9 152 X 152 0.63-0.92 li

7 152 X 152 0.92-0.96 li

7 152 dia 0.97 4 bars


Bresler and Gilbert 2 203 X 203 0.61 6 bars
(1961) 2 203 X 203 li
8 bars
Pfister (1964) 4 305 X 305 0.42-0.53 12 bars
3 203 X 457 0.36-0.49 12 bars
4 254 X 305 0.49 6 bars
Roy and Sozen (1964) 45 127 X 127 0.86x0.9 4 corner bars
Bertero and Felippa 2 76 X 76 None
(1964) 5 II
4 corner bars
2 108 X 108 None
6 II
4 corner bars
Hudson (1966) 32 102 X 102 0.46-0.47 8 bars
28 152 X 152 0.63-0.66 8 bars
Soliman and Yu (1967) 3 152 X 102 0.92-1.00 2 bars
11 II
0.44-0.92 4 corner bars
1 152 X 76 0.91 II

1 152 X 127 0.93 II

Shah and Rangan (1970) 11 51 X 51 0.83 None


Somes (1970) 42 102 X 102 0.88-0.92 None
Sargin (1971) 41 127 X 127 0.64-0.96 None
- -
4

TABLE 1.1 Continued .......


..-
Burdette and Hilsdorf 16 127 X 127 0. 72-1.00 None
(1971) 4 127 dia 1.00 II

Bunni ( 1975) 4 127 X 127 0.88-0.90 None


50 II
0.88-0.95 4 corner bars
PCA ( 1977) 13 254 X 406 0.68-0.72 4 corner bars
6 127 X 203 0.70 II

Bertero and Va 11 en as 3 254 X 254 0.78 8 bars


(1977) 3 229 X 229 0.96 II

3 254 X 254 0.78 None


3 229 X 228 0.96 II

Sheikh and Uzumeri (1978) 9 305 X 305 0. 77 8 bars


6 II II
12 bars
9 II II
16 bars

1.5.1 Chan (1955)( 6 )

As part of some other investigations Chan reported the testing of 9 prisms


152 x 152 x 292 mm with bent-in hoops, 7 cylinders 152 mm diameter and
305 mm high with spiral reinforcement, and 7 prisms 152 x 92 x 1321 mm
with welded hoops. These were loaded eccentrically or axially with a
transverse load at the mid point. Chans results for rectangular hoops,
when compared with unconfined concrete, showed a strength increase of more
than 50%, an increase in ultimate strain of about 500%, and that these
increases were only 50% and 70% respectively of those for equivalent
spiral reinforcement.

To determine the effect of confinement Chan ignored the hoop spacing, and
considered only the volumetric ratio of hoop steel.

1.5.2 Roy and Sozen (1963)( 10), (1964)( 11 )

Roy and Sozen developed an idealised bilineal stress-strain relationship


for confined concrete from 45 tests carried out on 127 x 127 x 635 mm
prisms with about 2% square hoop steel by volume. It was concluded that
the ductility of confined concrete was closely related to the spacing of
the hoops but the size of hoop bars and amount of longitudinal steel had
little effect on concrete properties. It was concluded from the test
results that the presence of square hoop steel increased the ductility
5

of concrete but not the concrete strength (peak stress) (See Figures 1.1
and 1.2).

I 25!

I 00
I
t:)
z
w
0::
1-
r.r:
0:: 0 75
w
0
;s
...J
>-
u
......
(f)
050
(f)
w
c::
I-
(f)

0 25

STRAIN (%)

FIGURE 1.1 Effect of Tie Spacing on Stress-Strain Relationship


of Concrete (Roy and Sozen 1963)( 10)
6

I
I
I
, _ _j _ _ _ _
0.5fc I
I
I
I
r

0.002 E5o

FIGURE 1.2 Stress-Strain Curve for Concrete Confined by


Rectangular Ties as proposed by Roy and Sozen
(1964) ( ll)

p h
E:
= 3!
'4
s
-s-
50

where h = overall depth of member


ps = ratio of volume of transverse reinforcement
to volume of concrete core
s = spacing of hoops
7

1.5.3 Bertero and Felippa (1964)( 12 )

In their discussion to Roy and Sozens paper above( 11 ), Bertero and Felippa
reported the results of tests performed on 76 x 76 x 305 mm and 114 x 114
x 305 mm p sms with square hoop steel and/or longitudinal steel under
concentric loading. Increases in concrete strength were found to be 13%
to 26%, depending on the hoop steel volume. It was concluded that
longitudinal steel alone did not enhance the concrete ductility. Hoops
alone however did, and hoops with longitudinal steel provided even greater
ductility enhancement.

1.5.4 Soliman and Yu (1967)( 15 )

Soliman and Yu conducted a study on the flexural stress-strain relation-


ship of-confined concrete. Fourteen units 152 x 102 x 1321 mm and one
each 152 x 76 x 1321 mm and 152 x 127 x 1321 mm 2343 tested under the
action of a major load and a minor load applied such that the neutral axis
was kept constant near the tension side of the unit and parallel to the
stronger axis throughout the entire range of loading.

An increase in concrete strength of up to 28% was observed by using


closely spaced rectangular hoops. It is unclear whether this was based
on gross concrete area or core concrete area. It appears also that no
consideration has been given to the spalling of the cover concrete and
that the stress-strain curve was based on the total concrete area
initially under compression.

The stress-strain curve proposed by Soliman and Yu for confined concrete


in flexure is shown in Figure 1.3 and the characteristics detailed below.
8

08fc'

FIGURE 1.3 Stress-Strain Relationship of Confined Concrete


in Flexure (Soliman and Yu 1967)( 15 )

1\
M Atie(So-S) 1.1
q" = ( 1. 4 Acc - 0. 45) ! '!'.

c At.1e S + .0028 BS 2

fcmax = 0.9 f~ (1 + .05 q") 1.2

Ece = 0.55 f~ X 10- 6 1.3

Ecs = 0.0025 (1 + q") 1.4

Ecf = 0.0045 (1 + 0.85 q") ' .. 1.5


9

where for these equations


Ac = area of concrete in compression

Acc = area of confined concrete in compression

At.1e = area of bar used for tie

B = width of bound concrete or 0.7 (depth


of bound concrete) whichever is greater
fc 1 = concrete cylinder strength

fcmax = maximum stress in confined concrete

s = tie spacing
= tie spacing at which ties are not effective
in confining the concrete, for the tests
reported in the paper S0 = 10 in.
A
In the q relation, the term Ace seems to be present in order to give the
11

stress-strain relation for thecgross concrete section which will result in


a lower value of q11 and hence lower values of fcmax' Ecs and Ecf"

1. 5.5 Kent and Park {1971)( 18 )

In 1971 Kent and Park proposed the stress-strain curve, for concrete
confined by rectangular hoops with the characteristics detailed below and
shown in Figure 1.4.
10

B
f'c
z =tan 0
r;
Confined
0.5['c r.oncrete

0.2 f'c
c /)

Unconfined concrete
I I
tc
A 0.002 c sou 50c 20<"

FIGURE 1.4 Stress-Strain Curve for Concrete Confined by


Rectangular Hoops (Kent and Park 1971)( 1S)

Region AB for E: c-< E: 0

where Eo = .002

2E
= f' (-c- (E:c)2) 1.6
fc c E: 0 Eo

Region BC for Eo ~ E:c ~ E:20c 1.7

fc = f'c {1-Z( cc- c )J


0
J.l

where z= 0.5
-E:
1.8
E5ou + E5oh co

3 + 0. 29f~
E5ou =
145f~ - 1000 1.9

1.10

where for these equations


f'c = concrete cylinder strength in MPa,
and h" = width of confined core,
Region CD c .L E20c

f c = 0.2 f'c 1.11

This relationship was proposed on the basis of an analysis of existing


experimental data. It incorporates many features of previous researchers
but no increase in the strength of confined concrete was considered,
although many researchers(l 6 , 12 ' 13 ' 15 ) observed a significant increase
in concrete strength due to confinement by rectangular hoops. Thi$
assumption was based conservatively on the test results of Roy and Sozen
(1964)( 11 ) which indicated no significant increase in strength due to
confinement by square hoops.

The falling branch was developed from the results of Roy and Sozen (19641
(ll), Bertero and Felippa (1964)(.17) and Soliman and Yu (19671(15 1,_ It
was assumed to be 1i near and fo 11 ows the same format as suggested By Roy
and Sozen (1964)( 11 ).

It was also assumed that the stress~stratn curve for the cover concrete
was either the same as the unconfined concrete~ or the confined concrete
core, or somewhere in between, at strains less than 0.004, At strai.ns
greater than 0.004 the cover was considered to have spalled and to have
zero strength.

1.5.6 Sargin (1971)( 19 ~ 20 l

Sargin tested 22 plain concrete and 41 laterally reinforced unit$:


18 plain and 31 laterally reinforced units were tested under concentric
12

loading and the remainder were tested under eccentric loading. All the
specimens were 127 x 127 x 508 mm and contained no longitudinal reinforce-
ment.

For concentrically loaded plain prisms the average concrete strength was
about 98% of the cylinder strength and the strain, over a 254 mm gauge
length, corresponding to the maximum stress was 0.0024. For eccentric
loading these values were 100% and 0.003 measured over a 127 mm gauge
length respectively.

These findings conflict with those of Sturman, Shah, and Winter (1965)( 26 )
where they found the strength of concrete in eccentrically loaded units to
be 20% higher than that of concentrically loaded units. Results reported
by Hognestad et al (1955)( 27 ) found excellent agreement between stress-
strain curves for their eccentrically loaded prisms and concentrically
loaded cylinders.

Sargin carried out a theoretical analysis of his results assuming that the
increase in strength of the 'truly confined concrete core is 4.1 times
the lateral pressure, and that the hoops yield at or before the concrete
reaches its maximum strength.

1.5.7 Vallenas, Bertero and Popov (1977)( 24 )

This study reported the results of 12 reinforced concrete columns tested


under axial load. The columns were 254 x 254 x 762 mm, with or without
8 longitudinal bars and diamond and square hoop sets. More than 20%
increase in concrete core strength was observed in some columns. An
increase in ductility of up to 300% was observed at the point of maximum
strength and up to 1000% at a point on the descending branch of the curve
where f c /f'c = 0.85. Results also showed that the addition of longitudinal
steel enhanced the strength of the confined core concrete, on average by
7%. It was shown that none of the existing curves predicted the behaviour
of their tests (Figure 1.5) and so Vallenas, Bertero and Popov suggested
an improved stress-strain relationship as follows and shown in Figure 1.6.
J.3

THEORETICAL
CD
BWME
\b
KENT AND PARK
@SARGIN
1.2
, '.,--
- -..... .. ............. EXPERIMENTAL
=,;,-:::c..:.:..:.:.:.=:...:,,:-:=
"' " .. ._ ~PLAIN CONCRETE
..., "' '-M, CONFINED

...,
1.0 WITHOUT COVER
'-..... ..... .:g; REINFORCED WITHOUT COlER
............ .................~

--
.85
.8
............. ... @)........_ ......
....... ~ ... ........__
.6 ...... ...
............ ..@...................-:-..::::::
4
\@
\ .........
' \ ... ....... .....
.2
' ......
OL-------+-----~-.------~--------~------~-------r~
0 01
' .02 .03 .04 05 06
~

FIGURE 1. 5 Comparison of Analytical Curves with Experimental


Results (Vallenas et al, 1977)( 24 )

fc/f~

L2r---:,- 8
.... , ..... ,
... ...
1.0 ... ...
...
.8
...._
..... ..... .....
r-- 9"-----i - ......
...
..... ..!
.-EXPERIMENTAL
6 ..........
4

0
A Eo,
CJl ANALYTICAL-

E.3kl
c
'
I
I
I
I
L
D

0 .01 .02 03 04 .05 06

FIGURE 1.6 Confined Concrete with Longitudinal Reinforcement -


Analytical Curve and its Comparison with Experimental
Results (Vallenas et al, 1977}( 24 )
14

Region AB
EE E E 2
j.o (_!_)-k(_!_)
fc - c Eo Eo
y- 1.12
c
1+ f.EcEo
"i("f"' - 2
L.: c
(.:1.)
so
J
Region BC

fc EL
y- k [ 1-ZE 0 ( - - 1)] 1.13
c Eo

Region CD

fc
1'- 0.3 k . . . 1.14
c

where

. 734(5) ) Ps fyh
E = . 0024 + . 006 (1 - hn 1.15
0 jf~
d"
s (p s +d p) f yh
k = 1 + .1096 (1-. 24 -h" ) - - - - - - " ' - - 1.16
If~
0.5
z = --------=---=---=-----:::~---
3 /lh" . 3 + 0.29 f~
1.17
4 Ps / s- + (145 f 1
c
- 1000 ) - 002

where for these equations


d = diameter of longitudinal steel bar
d 11 = diameter of tie bar
Ec = initial modulus of elasticity of concrete (MPa)
fyh = yield strength of tie steel (MPa)

h" - core dimension of rectangular tied colum


inside the hoop (mm)
k = ratio of maximum stress in the confined
concrete to the cylinder strength
S = hoops spacing (mm)
15

p = volumetric ratio of the longitudinal steel


= volumetric ratio of the tie steel

It is worth noting that these equations are dimensionally dependent and


the equation for E 0 appears to provide too large a strain increase. The
result is that the stress at a strain of 0.002 is significantly less than
f c The equations also suggest that strength and ductility are not
1

directly proportional to the confining force.

Gill, Park and Priestley (1979)( 28 ) have shown that when applying the model
to members under combined axial load and bending, results from moment-
curvature analysis using Vallenas, Bertero and Popovs model have under-
estimated the moment when compared with both experimental results and
other models e.g. Sheikh and Uzumeri( 2).

1.5.8 Sheikh and Uzumeri (1978)( 2 )

Sheikh and Uzumeri (1978)( 2 ) presented the results of 24 columns 305 x 305
x 1956 mm loaded in monotonic axial compression. Four different lateral
steel arrangements, involving rectangular and/or octagonal hoops were used
with 8, 12 or 16 longitudinal bars. Also examined were the volume spacing
and characteristics of the hoop steel, and volume of longitudinal steel.

It was concluded that rectangular hoop reinforcement enhances the strength


and ductility of confined concrete. The distribution of longitudinal
steel around the core perimeter increased the efficiency of the confine-
ment. A closer spacing of hoops resulted in higher concrete strength and
ductility for the same amount of hoop reinforcement and vice versa. A
higher steel strength and larger amount of hoop reinforcement resulted in
higher strength and ductility.

The model developed by Shiekh and Uzumeri is detailed below and shown in
Figure 1.7.
16

t
U1
U1
w
0::
1-
(J)

QL-----~~----------J-----L-------------------

STRAIN---+-

FIGURE 1.7 Stress-Strain Cutve for Cohfin~d Concr~t~ in Sq~are

Columns (Sheikh and Uzu~ert 1978)( 2 ]

Ks =
1
l.O + l40P
. DCC
[o - 5.5n C2b2}(1 - 0.5 ~l\2JM
b s s
... 1.18

(ff in MPa and p in KN)


s DCC

sl = 80 Ks f'c X 10- 6 (f'c in MPa) I!! ~ .. 1.19

s p f'
= 1. 0 + 2~8 [ 1 - 5(S/b) 2] s s 1.20
00 If[
(C in mm; f'c and f's in MPa}

= 0.225 ps;'b!S + s 2 1.21


S85

The units tested by Sheikh and Uzumeri, which were more realistically sized
than previous tests, showed clearly the enhancement of strength and
ductility due to confinement. The model however produces a flat plateau in
the stress-strain curve due to the use of high yield strength hoops,
(fyh = 520 or 700 MPa). This meant that when maximum concrete core
strength was reached the hoop steel was still below yield, allowing the
17

confining pressure to increase with subsequent increases of longitudinal


and lateral strain. It is felt this may not be suitable for New Zealand
conditions where lower grades of hoop steel (e.g. fyh = 275 or 380 MPa)
are commonly used which might not produce this plateau.

1.5.9 Modified Kent and Park (1979)( 29 )

In the light of research results presented more recently by Gill, Park and
Priestley (1979)( 28 ) a modified form the Kent and Park stress-strain
relationship has been proposed by Park, Priestley and Gill (1979)( 29 )
which accounts for the increase in the concrete core strength and strain
at maximum stress.

The maximum stress reached is given by Kf~, where

fyh
K=1+p F
s c

The maximum stress, Kf~, is assumed to occur at a strain of 0.002K. The


regions of the stress-stain curve are shown in Figure 1.8 and as detailed
below.

Modified Kent and


tan8m
Park , Confined Zm=--
Kf~
where K = 1+ Ps fyh
f'c
c
ill
ill
<lJ
.....
..._
1.1) Kent and Park, Confined
....<lJ
Q.J if K = 1 is assumed
.....
u
c: -----Unconfined
8

0.002K

A ~--J-J~---~--------~-------~-------~
0 0.002 0.005 0.010 0.015 0.020
Concrete Strain, Ec

FIGURE 1.8 Stress-Strain Curve for Concrete Confined by


Rectangular Hoops (Park, Priestley and Gi 11 ,1979) ( 29 )
18
:/
Region AB

1.22

Region BC (cc > 0.002K)

f c = KfTc [1 - zm (
E
c - 0. 002K) J 1.23

but not less than 0.2Kf', where


c
0.5
zm = ~-:--;~""',.,-------------
3 + 0.29f'c /["
3
145f~- 1000 + 4 Ps;sh- o.00 2K 1.24

Gill, Park and Priestley note that the increase in peak stress may not be
as much as observed in tests but the discrepancies seem to make little
difference when using the model for moment-curvature analysis as Gill,
Park and Priestley (1979)( 28 ) have shown.

It should be noted that the basis of choosing 0.002K as the strain at peak
stress is so that the second degree parabola defining the region AB
maintains the same tangent modulus at zero stress (i.e. the same initial
modulus of elasticity) regardless of the value of K. The relationship
also assumes that rectangular hoops are only one half as effective as a
circular spiral in enhancing the strength of the concrete.

1. 5. 10 Summary

There is a general agreement among researchers that the use of rectangular


confining hoops does not significantly influence the rising branch of the
stress-strain curve (i.e. the initial stiffness is unaltered) up to about
e.5 f~. An increase in ductility is also a common conclusion but the
amount is often a point of wide variance. There is disagreement as to
whether enhancement of concrete strength occurs and if so, to what extent.

Most of the tests, as shown in Table 1.1, were conducted on small units,
commonly 127 x 127 mm in section, with small core to gross area ratios and
only 4 corner bars for longitudinal steel if any at all and simple
arrangements of rectangular hoops.

The definition of the core is another variable over which there has been
much disagreement, brought about at least in part by the large variation
in calculations of volumetric hoop steel ratio and gain in concrete
strength which occur with small test models. Theoretically it is likely
that the core may be defined to the centre~line of the outer hoops but
from a practical standpoint and in keeping with current practice it will be
taken in this study to the outside of the outer hoop. The behaviour of the
cover concrete is another variable which is often ignored by researchers.
Attempts have been made to overcome this problem by casting test units
without cover, however Burdette and Hilsdorf (1971)( 2l) showed that for
columns without cover, shrinkage may result in an appreciable reduction
in the strength increase due to confinement. In this study the behaviour
of the cover has been taken as that for the large plain concrete units and
its load carrying capacity at various strain levels calculated accordingly.

King (1946)( 3) concluded from comparitive scaled up tests that the


behaviour of large size columns would be very different from small scale
models, and that a large number of parameters would require investigation
before the results of small scale models could be used to predict the
behaviour of full size units.

None of the previous experimental research was carried out at loading


rates simulating those likely to occur during an earthquake. Generally
tests were carried out on axially loaded units. Therefore, there must
remain doubts about the validity of the proposed curves to seismically
loaded columns subjected to combined axial load and bending moment.

It is beyond the scope of this report, but ideally a unified stress-strain


curve is required that predicts the behaviour of concrete confined by
arrangements of either circular spirals or rectangular hoops. Based on
the assumption that most researchers adopt, that the confining force from
the lateral reinforcement is equivalent to an applied lateral pressure,
there is no reason why this generalised approach should not be possible.
20

CHAPTER TWO

CONFINEMENT REQUIREMENTS FOR PLASTIC HINGE ZONES


IN REINFORCED CONCRETE COLUMNS

2.0 SUMMARY

This chapter reviews the requirements for confining steel in plastic hinge
zones in rectangular columns, as specified in the seismic codes of various
countries. In particular a comparison is made between New Zealand and
overseas recommendations. This work is summarised from a paper by Park
and Priestley (1979)( 30 ).

2.1 THE CODES CONSIDERED

The overseas code requirements for reinforced concrete design considered


in this section will be the A.C.I. Building Code( 31 ), the S.E.A.O.C. Code
( 32 ), the tentative provisions for buildings of the A.T.C.( 33 ) and the
A.C.I. Committee 343 report on the analysis and design of bridge
structures( 34 ). An indication of Japanese practice can be obtained from
publications in English for example references 35 and 36. In New Zealand
the Ministry of Works and Development have developed provisions for
confinement of bridge piers( 37 38 ) and the Standards Association of New
Zealand has prepared a draft Concrete Design Code( 1), which is now being
redrafted into its final form. The summary given below will not include
references to design of shear reinforcement, on the understanding that
transverse hoops placed for confinement will act as shear reinforcement
and that there may be cases where shear steel is the governing criteria.
Generally the transverse steel for hoops is expressed as Ash' the total
area of hoop bars and supplementary cross ties in the direction under
consideration within longitudinal spacing Sh. Sometimes ps is used to
define the total volumetric ratio of transverse steel in the section
confining the concrete core.
21

2.1.1 ACI 318-77( 31 )

Confining steel consisting of hoop reinforcement is required over the end


regions of columns adjacent to moment resisting connections over a length
from the face of the connection equal to the greater of the over a 11
thickness h (h being the larger sectional dimension for rectangular
columns), one-sixth of the clear height of the column, or 450 mm.

For rectangular hoop reinforcement, with or without supplementary cross


ties, if Pe ~~ 0.4Pb. the transverse steel should be designed as for beams
but the hoop bar diameter should not be less than 10 mm and the spacing
should not exceed d/2. If Pe > ~0.4Pb' and if a single rectangular hoop
is used, the area of one leg of the hoop bar in the direction considered
within spacing sh should be at least equal to

2.1

where ps is calculated by the greater of Equations 2.2 and 2.3.

If Pe ~ ~o.4Pb, ps is required to be at least equal to

p
s
= 0. 45 [ _g_
A - 1 _c_
Ac
f Jfyh
1
2.2

If Pe > 0.4~Pb, ps should not be less than that given by Equation 2.2, or

p = 0.12f~/fvh 2.3
s - ,J

wtth Ac taken as the area of concrete core measured to the outside of the
pertpheral hoop, The value of sh used should not exceed 100 mm.
Supplementary cross ties of the same bar diameter as the hoop may be used
to reduce the unsupported length lh, The hoop and cross tie bar diameter
should not Be less than 10 mm for longitudinal bars 32 mm diameter or
smaller~ or 12 mm for larger longitudinal bars or bundled bars.

2,1,2 SEAOC (1975}( 32 1

The potential plastic hinge zone is taken as in ACI 318-77,

For rectangular hoop reinforcement the total area of hoop bar and supple~
mentary cross ti'es (l:f any) in the direction under consideration within
spacing sh should not be less than
22

2.4

or
Ash =0.12sh"f'/f
h c yh . . . 2. 5

whichever is greater. The value of sh used should not exceed 100 mm, and
supplementary cross ties or legs of overlapping hoops should not be spaced
at more than 360 mm between centres transversely.

2.1.3 ATC (1978)( 33 )

The SEAOC provisions appear to have been followed for fully ductile frames.

2.1.4 ACI COMMITTEE 343( 34 )

For rectangular hoop steel the transverse bar diameter is specified as in


ACI 318-77. The quantity of transverse steel is not specified, but it is
stated that the tie spacing should not exceed the least dimension of the
member or 300 mm, except that when bars larger than 32 mm diameter are
bundled the tie spacing should be reduced to one-half of that value.

2.1.5 \PANESE PRACTICE( 35 36 )

No information was available to Park and Priestley (1979)( 30) regarding


current Japanese practice for bridges. However, AIJ requirements for
buildings( 36 ) specify a minimum rectangular hoop diameter of 9 mm with
spacing not to exceed the lesser of 150 mm, one-half of the smaller
column dimension, or 7.5 longitudinal bar diameters.

2.1.6 N.Z. MINISTRY OF WORKS AND DEVELOPMENT, CIVIL DIVISION( 37 38 )

Until recently the design of confining reinforcement for bridge columns


was governed bv the MWD provisions ( 37 38 ). This involves sufficient
confi111ng steel to ensure that the available structure displacement
ductility factor~ is at least 6, ~u~ed on the following steps:
(1) The ratio of the structure displacement ductility factor, ~.
to curvature ductility factor, ~u/~y' within the plastic hinge
region is calculated. The ratio depends on geometric consider-
ations, and the plastic hinge length which is taken to be that
given by Baker's formulaC 39 }.
23

(2) Calculate the yield curvature ~y' and hence the required ultimate
curvature ~u
(3) Calculate the "ultimate" compression strain e:cu corresponding to
~u' based on a conservative idealisation for the stress-strain

curve of confined concrete.


(4) Calculate the required volumetric ratio of confining reinforce-
ment.

For rectangular hoops and supplementary cross ties ps is obtained using

e:cu = 0.0021 { 1 + 150ps + (0.7 - 10 s) %} 2.6

Equation 2.6 is based on the work of Baker and Amarakone( 39 ) and Chan{ 6 ),
but was made significantly less conservative than the results of that work
in the light of the findings of more recent tests (see the December 1977
amendment ( 37).

It is understood that the MWD will be adopting the requirements of the new
SANZ Concrete Design Code (at present in draft form( 1)) when that SANZ
code is available in its final form.

2.1.7 DRAFT SANZ CONCRETE DESIGN CODE( 1)

The confinement provisions of the draft SANZ Concrete Design Code, DZ3101,
are based on the ACI/SEAOC requirements modified to take account of the
effect of axial load level.
(a) First Draft
In the first draft of DZ3101, issued for comment in 1978, the
potential plastic hinge regions were specified as in ACI 318-77.
The total area of rectangular hoop reinforcement, including
supplementary cross ties if any, in the direction under consider-
ation in potential plastic hinge regions, when Pe ~ 0.6f~Ag, was
required to be not less than

2.7

nor
f1 p
= o.12shh" T-yh [o.33 + 1.67 f~c g J 2.8
24

where Pe was not to be taken as less than O.lf'A


c g
. The diameter
of hoop or tie bar was to be at least 10 mm, and the maximum
centre specing of hoop sets was not to exceed the smaller of one-
fifth of the smaller member section dimension, 150 mm, or six
times the diameter of the longitudinal bars. The yield force of
the hoop bar or supplementary cross tie was to be at least one-
sixteenth of the yield force of the longitudinal bar or bars it
was to restrain. Other rules were also given to ensure adequate
lateral support of the longitudinal bars. These equations were
based on theoretical moment-curvature analyses( 25 ) conducted
using the stress-strain model proposed by Kent and Park (1971)( 18 ).
(b) Revised Draft
The potential plastic hinge region for Pe ~ ~0.3f~Ag is now
recommended as not less than the longer column section dimension
in the case of a rectangular section, or where the moment exceeds
0.8 of the maximum moment at that end of the member. When
Pe > ~0.3f'A
c g
the potential plastic hinge region is increased to
1.5 times the above value.
In potential plastic hinge regions when rectangular hoops, with
or without supplementary cross ties, are used and either
Pe ~ ~0.7f~Ag or Pe ~ ~0.7P 0 , the total area of transverse steel
within spacing sh should not be less than

A f' p
~- JT-yh e
Ash = 0.\hh" [ 1 (0.5 + 1.25 ~f'A
2.9
c g

nor
f' p
e
Ash = 0.12shh 11
/ (0.5 + 1.25 ~f'A
2.10
yh c g

When the load Pe has been obtained using a capacity design


procedure, the value of the strength reduction factor ~ in all
the above equations can be taken as unity.
These modifications of Equations 2.7 and 2.8 were made on the
basis of the test results on near full scale columns obtained by
Gill, Park and Priestley (1979)( 28 ).
25

2.2 COMPARISON OF CODE REQUIREMENTS FOR A TYPICAL RECTANGULAR COLUMN

The difference between the confinement ratios required by the above codes
is illustrated in Figure 2.1 for a typical 700 mm square column confined
by an arrangement of square and octagonal hoops such as would be used in
a building.

Figure 2.1 illustrates that the Japanese requirements apparently result in


very low quantities of hoop steel.

The step change in the ACI 318 requirements has again been taken to occur
at Pe = 0.1f~Ag in Figure 2.1, which is a reasonable approximation for
0.4~Pb. The SEAOC and ATC recommendations do not permit a reduction in
the hoop content at low axial load levels. The SANZ (revised draft)
quantity shows a linear increase in hoop content with axial load from 50%
of the SEAOC moment at zero axial load to 1.38 times the SEAOC amount at
Pe = 0.7f cA.
1
g
Using the arrangement of one square hoop plus one octagonal
hoop per set, shown in the column section in Figure 2.1, the total effect-
ive area of hoop bars per set, Ash' is taken as twice the area of the
square hoop leg plus 1.414 (i.e. twice 1/12) times the area of the
octagonal hoop leg. Hence if both hoops are of the same size bar with leg
area Asb' the value of Ash would be 3.414Asb The SEAOC code requirements
for the column shown in Figure 2.1 could be met using 16 mm diameter
square and octagonal hoop bar with the hoop sets placed at 88 mm centres.
The longitudinal steel used in this example consisted of twelve 32 mm
diameter bars.

As illustrated in Figure 5, the MWD approach requires substantially more


hoop steel than does the other approaches at high axial load levels. At
low axial load levels only very small quantities of confining steel are
again required because of the high emphasis of the d/c ratio on Ecu in
Equation 2.6. For example, Equation 8 indicates that Ecu = 0.010 will be
available if c/d < 0.19 even if ps = 0. Also there are large differences
between the hoop requirements for various ~pier demands in Figure 2.1.
26

P1 = 0.0197
fy = 375MPa

fyh = 300!-1Pa

E f~ = 30MPa
40 E
a Cover to hoops
a
" = 25mm

700mm

' SECTION

30

NZMWD
NZMWD
~-for llpier =6

20
NZMWD
for ll pier =4

-CI -C
~V) lf)
SANZ SANZ

10

0.1 0.2 0.3 0.1. 0.5 0.6 0.7


Axial Load Ratio Pe
f~ Ag

FIGURE 2.1 Comparison of Code Hoop Steel Reguirements for


a Square Column (Park and Priestley (1979)( 30)
2.3 CONCLUSIONS

The various overseas and New Zealand code recommendations for transverse
confining steel in potential plastic hinge regions of columns and piers in
seismic design show vast differences in the required quantities of
transverse steel and it is evident that this is still a matter of some
controversy.

Recent tests on near full scale reinforced concrete columns containing


spiral steel or rectangular hoop steel, under simulated seismic loading,
at the University of Canterbury, have shown that the quantities of
confining steel recommended in the draft SANZ Concrete Design Code, with
slight modifications mainly to take into account the effect of the
possible increase of actual concrete strength over a specified f~ and to
avoid the spread of plastic hinging into less heavily confined regions,
will result in available displacement ductility factors for columns of at
least 8.

The provisions for confining steel which have been used by the Ministry
of Works and Development for ductile bridge piers appear to be very
conservative when axial load levels are high and are in need of revision
to avoid the use of excessive quantities of confining steel. This
observation is made not only from comparison with the quantity of
confining steel required by the draft SANZ Concrete Design Code but also
in the light of the results of the column tests recently conducted at the
University of Canterbury. At low axial load levels the MWD confinement
provisions may be unconservative but the requirements of transverse steel
for shear will govern in that case and result in a greater quantity of
transverse steel being placed.
28

CHAPTER THREE

DESIGN AND CONSTRUCTION OF TEST UNITS

3.0 SUMMARY

This chapter describes the design of the test units, the properties of
the materials used and the construction method employed.

3.1 UNIT SIZE CRITERIA

In 1978 the University of Canterbury installed a 10 MN Dartec Universal


Testing Machine in the Civil Engineering Department, which now permits
the testing of large scale building components at rapid loading rates.

Considerable effort was made to test as large a unit as possible, in


order to avoid the complications caused by scale effects that have been
noted by other researchers (e.g. King (1946)( 3) and Sheikh and Uzumeri
(1978)( 2)). The 10 MN capacity of the Dartec testing machine was the
dominant factor in determining unit size. Other factors which influenced
final column dimensions were the expected enhancement of concrete
strength due to confinement and rapid loading rates; a realistic longitud-
inal steel ratio; the expected yield and ultimate strengths of the
longitudinal steel; the ease of formwork construction and a suitable
height to core dimension ratio. Based on these considerations a unit
size of 450 x 450 x 1200 was chosen.

3.2 DESIGN OF TEST UNITS

3.2.1 Longitudinal Steel

The use of high strength steel (Grade 380) in reinforced concrete building
columns is becoming increasingly common. The early strain hardening
of the steel improves the performance of the column at high strains by
compensating for the loss of capacity due to spalling cover concrete.
This approach applies principally where column moments are induced from
overstrength beam drive and there is a desire to prevent column hinging.
However if columns are the fundamental earthquake resisting elements. as
29

in a bridge where plastic hinging must occur in the piers, the column
overstrength may be very high with Grade 380 steel, resulting in excessive
shear forces. To model current building practice 18 of the 30 units were
constructed with Grade 380 longitudinal steel.

It has been shown that distribution of longitudinal steel has a marked


effect on ductility( 40) and confinement( 24 ), and so two longitudinal steel
distributions were chosen to represent current practice. The longitudinal
steel was also selected to conform with NZS DZ3101( 1 ) and the volume of
steel to be similar between units. The final choices for longitudinal
reinforcement were 8 DH 24 bars with Pt = 0.0186, or 12 DH 20 or 12 D 20
bars with Pt = 0.0179 as shown in Figure 3.1, where D signifies Grade 275
deformed bar (nominal f y = 275 MPa) and DH signifies Grade 380 deformed bar
(nominal f y = 380 MPa).

3.2.2 Hoop Steel

The hoop steel was designed to conform with a proposed modification of


NZS DZ 3101( 1 ) for Special Transverse Hoop Reinforcement in Potential
Plastic Hinge Zones, as detailed in Chapter Two.

The equations used in their modified form, restated here from Chapter Two,
were:

A f~. 1.25P
= o.35hh''Cf- 1) f (0.5 + f'A e 3.1
c yh c g

or f' 1. 25P
Ash = 0.12Shh" / (0.5 + ...,.,f';-;:-A_e ) 3.2
yh c g
whichever is greater, where Pe is not greater than 0.7f~Ag.

Four levels of confinement were chosen to cover the range of axial loads,
Pe. These were Pe equal to 0.1, 0.25, 0.4, and 0.7 of f~Ag. Confinement
to an axial load level of 0.25f~Ag was chosen to be typical of current
practice and to be representative when considering the effect of the
other variables.

Equation 3.2 governs the hoop steel quantity because of the test unit
dimensions chosen and from this equation the required spacing of the hoops
was determined for a given transverse bar area. All hoop steel was plain
bar Grade 275. The hoop bar diameter and spacings are given in Table 3.1.
30

TABLE 3.1 Hoop Bar Diameter and Spacing of Hoop Sets

Pe Hoop Bar Diameter Hoop Spacing


f'A
c 'g mm mm
0.1 11"\
lU 98
0.25 10 72
0.4 12 88
0.7 12 64

The hoop spacing was reduced by half for 200 mm at each end of the test
unit to ensure that failure occurred in the 800 mm long central region.
The configuration of hoop sets is shown in Figure 3.1
.[ 11 Yl
0
0

"'
31

!6mm
111-L
16 mm. :-------llM=====$$======!F1 Suppo ~r-----
Support -----JI ~~
bar bar HI-
DH 24 DH20 7

I 0 0
0 0
"!"

R10
H ~

Hoopsets -~~~ Jl r sets


1!!11
1-
5tJ = 98 mm nn==f=~~=======U/. Bmm F=
0
F- = 0

"'
E

~r
~-r

~~ ~~~~~;~~~:~
0
sh varies 0
between units -Cl__ "'

BDH24

RTO

' 183

450

FIGURE 3,1 Details of Test Units and Transverse Retnforcement


32

3.3 MATERIAL PROPERTIES

3.3.1 Steel

The reinforcement for each bar size was chosen from a single supply batch
to ensure uniform properties. Three randomly selected samples of each
diameter bar were subjected to full extensometer tensile testing in order
to establish the stress-strain relationships for the steel.

This assumes that the static stress-strain curve for steel in tension can
be applied to steel in dynamic compression. For these test units the
longitudinal steel carries a maximum load of about 1.5 MN so that an error
of 10% due to any discrepancy between tension and compression yield, or
from dynamic loading rates represents only 2% error in concrete core
stress.

Table 3.2 summarises the yield and ultimate stresses while Figures 3.2 -
3.6 show the measured stress-strain curves.

TABLE 3.2 : Yield and Ultimate Steel Stresses

Bar Yield Stress Ultimate Stress


(MPa) (MPa)
DH 24 394 646
DH 20 434 708
D 20 272 416
R 12 296 424
R 10 309 436

3.3.2 Concrete

The concrete used in all specimens was delivered ready-mixed with a target
strength at 28 days of 25 MPa, a slump of 75 mm and maximum aggregate
size of 20 mm.

Cylinders 200 x 100 mm diameter were cast with each pour and tested at 7,
14, 28, and 42 days, and also during testing, Cylinders were cured in
the fog room at 2oc in 100% relative humidity while the test specimens
were covered with hessian and polythene with the top surface kept moist
for 7 days before stripping the formwork. They were then allowed to stand
for another 5 weeks (6 in total) before testing.
600

d: fy = 394 MPo
~
I
(/")
fu = 647 MPo
t] E5 = 205 GPa
~ Cy = 0.00187
ll)

0 I I ~
0 0.01 0.02 0.03 0.01, 0.05 0.06
STRAIN

FIGURE 3.2 : STRESS-STRAIN CURVE FOR DH24 STEEL (GRADE 380) w


w
!

800

600
-
-
A
0
Q
~
I
ll)

I
ll)
lu fy = 1.31. fv!Pa
((
1--
ll)
fu = 708 MPa
Es = 185 GPo
[ y -- 000213

200
II

0"--------J'-------L ~------~-----~--------~'~
0 0.01 0;02 0.03 0.01. 0.05 0.06

STRAIN

FIGURE 3.3 : STRESS~STRAIN CURVE FOR DH20 STEEL (GRADE 380) w


..j:::.
1.00

. 16.1. [y ...
11-- 1
300 r.LI_ _ _ _ _ _ __

0 fy = 272MPa
Q
~ fu = 1.16 MPa
V)
E5 = 211 GPa
V)
IJ.J
200 [y = 0.00129
~
V)

I'
100

ol I I I I I I~
0 0. 01 0.02 0. 03 0.01. 0. 05 0.06
STRAIN
w
FIGURE 3.4 : STRESS-STRAIN CURVE FOR 020 STEEL (GRADE 275) (.)1
J
400

14 6.49Cy 1
300
fy = 309MPa
fu = 443MPa
0 E5 = 185 GPo
Q
~
I y = 0.00167
~ 200
~
lr)

100

0~--------~--~----~--------~--------~~--------~--------~
0 0.01 0.02 0.03 0.04 0.05
STRAIN
w
FIGURE 3.5 : STRESS-STRAIN CURVE FOR RlO HOOP STEEL (GRADE 275) 0'1
A
400

' I..... 6.09 [y .. I


300 I -
fy = 296MPa
0 fu = 424 MPa
a..
~
I
E5 = 192 GPo
ll) [y = 0.00141
V)
Lu 200
e:
t.t)

100

0 ~--------~--------~--------~--------~--------~--------~~~
0 0.01 0.02 0.03 0.04 0.05 0.06
STRAIN

FIGURE 3.6 : STRESS,STRAIN CURVE FOR R12 HOOP STEEL (GRADE 275) w
'--l
38

Figure 3.7 shows the strength-age properties of the three batches of


concrete as measured by standard 200 x 100 mm diameter cylinder tests.
The average cylinder strength measured at 6 weeks is given in Table 3.3,
which also summarises reinforcement details for all units. It will be
noted that strength at 6 weeks was very close to the target value of 25 MPa
for all three batches.

-.
~
ll)
"<
,,
....__

~ 1.0
(::J

~
~
(/)
.....
ltJ
(.!J
Q::
~
)(
Batch 1
i?: o Batch ;;
~ 0.5 X
ltJ 0 x Batch 3
~
(/)

Q::
l!J
~
~
<..J

~
0
00
<..J 10 ;;o 30 40
AGE (DAYS)

FIGURE- 3.7 STRENGTH-AGE DIAGRAM OF 200 x 100MM DIAMETER CYLINDERS


39

TABLE 3.3 Test Unit Properties


.........
"' .s::::
0... Number and Hoop
::E: +-'
.__; m Diameter Diameter Volume
c:
-o QJ
c: $....
Specimen of Yield and Yield of Hoop
c:C +-'
Vl
0
~

s.... Number Longitudinal Strength Spacing Strength Hoop Type


ZQJ
-o Bars Steel
.s:::: c:
Ui-
+-" ,.,--
ra>, (mm) (MPa) (mm) (MPa)
rou

1 0 0 Q

2 12 'V 20 434 10 'V 72 309 ,0182

(V)

lJ")
N
-
3
4
5
6 8
II

II

II

'V 24
II

!I

II

394
~I

1,_1

fl
~

~
,,
1.1
II

i).

,,
,0174
D
~
2 "''
. \.~
~
II II I{
t-1 7
!I II I~ ~~ ll
8
9
,, H II. ~~

"'
10 II ,
..
'' I~
"'
11 Q 0 Q

12 12 'V 20 434 10 'V 98 309 \0140

co.
<:j-
N

2
13
14
15
16
"
II

II

II
~

I),

II

IJ
10
12
12
'V

'V

'V

10. "' 72
72
88
64
'*
296
l1

309
.0182
,0224
,0309
,0182
D
17 8 'V 24 394 10 "' 98 lil
,0134

~
N
u
18
II II
10 rv72 ,0174
19 " H 12 'V 88 296 ,0213
12 "' 64
ll
20 " ~I
,0293

21
22
23
12
0
'V

!I

II
20 272
I}

II
10
0
'V

10 "' 72
98 309
I)

I)
,01400
Q

,0182
,0224
N 24 12 "' 88
<:j-
N 25 ll II,
12 'V 64 ~~.
,0309
2 26 0 0 0
!I l,l "!
(V) 27
II 1.1 I~
28
"
~~
29 I"'

II II II
30
40

3.4 CONSTRUCTION

The height of the test units was chosen to be 1200 mm for two reasons.
Firstly this gave a suitable aspect ratio of 3 to 1 height to core dimens-
ion, and secondly the 18 mm construction plywood was produced in a 2400 x
1200 mm module. This meant that test units could be cast in batches of
ten in two moulds, each with five test units. The formwork was supported
by 38 x 38 x 4.7 mm angle at all exterior joints and at the third points of
height on each side. The test units were tied at these levels with 10 mm
bolts into 16 mm diameter rods that were to be used as lifting points and
also during testing as supports for the linear potentiometers used for
longitudinal strain measurement.

The construction sequence consisted of tying the cages, fixing the strain
gauges to the transverse reinforcement and placing the cages in the
painted and oiled moulds. The cages were fixed in position by the bolted
tie rods.

The ready mixed concrete was placed in 3 lifts and well vibrated.
Eighteen 200 x 100 mm test cylinders were cast and vibrated on a vibration
table with each pour. Figure 3.8 shows some aspects of the construction
sequence.
41

FIGURE 3.8 PhOtographs of the Cons truction Sequence


42

CHAPTER FOUR

INSTRUMENTATION AND TESTING PROCEDURE

4.0 SUMMARY

To obtain the strain and load measurements for analysis various instruments
were required to record the data. This chapter details the instruments
used and describes the testing procedure.

4.1 INSTRUMENTATION

Two methods of recording data were used, depending upon the loading rate.
For dynamic loading rates all measurements were recorded electronically
and synchronised with a 2HZ timing signal generator. For 'slow' loading
rates all measurements were recorded manually while loading continued at
the preset rate.

4.1.1 Load Measurement

Two methods of load measurement were employed, The first (Pressure Load),
as part of the Dartec testing machine equipment, measured load across
the pressure transducer valves of the machine hydraulics to a precision of
1 kN. The second (Strain Load) measured load from strain gauges installed
on the columns of the Dartec machine. Pressure load was recorded on
three Hewlett-Packard X-Y (Type 70358) Plotters, plotted against Dartec
displacement, sum of East and West, and sum of North and South potentio-
w.eter strains respectively, Strain load was recorded along with Dartec
displacement against time on a two channel Hewlett-Packard chart recorder,
(Type 7402A).

4.1.2 Longitudinal Concrete Strains

These were also recorded by two methods, The first was the Dartec machine
stroke displacement to give an overall unit strain. The second method
recorded linear potentiometer displacement over a central 400 mm gauge
length from four 50 mm Sakae 20 FLP 50A 10 Kohm linear potentiometers. By
allowing for the flexibility of the Dartec machine quite good agreement
43

was found between the two records. As well as being recorded in pair
summation plotted against load, as described above, individual potentio-
meter readings were recorded on a 20 channel Bryans Southern Ultra-violet
Chart Recorder.

4.1.3 Hoop Reinforcement

To measure strainson hoop reinforcement, electrical resistance strain


gauges were fixed to the underside of the transverse reinforcement. In
general six gauges were located on each unit at three different levels
within the test region, as shown in Figure 4.1. For tests carried out at
slow loading rates additional strain gauges were used to obtain further
information.

Gauges at:-
k'l F
~J i'1 "

r
Jl -- 2,10 ~

H H H
-- 4,11 H
H H H H
H ~I - - 7,12 H H~ H H
rl H H
~
f
- n H
~. f
"

8 3 8

7 4
7

STRAIN GAUGE POSITIONS

FIGURE 4.1 Strain Gauge Locations


44

To affix the strain gauges, gauge sites were smoothed with emery paper and
cleaned thoroughly with Methyl Ethyl Keytone. 5 mm KFC-5-C1-11 strain
gauges were used throughout, applied with Kyowa CC-15A cement and joined
to Showa 5FC-5T self-adhesive terminals. The strain gauges were water-
proofed with at least four layers of Shinkoh SN/4 compound and mechanical
protection was provided by two layers of Scotch 3M vinyl mastic. All
strain gauge leads were threaded through electrical Spaghetti which was
11 11

double tied to the stirrups close to the strain gauge site and which
provided mechanical protection to the strain gauge leads. The strain
readings were recorded individually on the Bryans-Southern Recorder for
tests carried out at dynamic loading rates. For tests at slow loading
rates, load, Dartec displacement and strains, both longitudinal and
transverse, were recorded manually. Potentiometers were connected to a
Hewlett-Packard Digital Volt Meter and strain gauges to calibrated Budd
P350 strain indicator, respectively.

4.2 TESTING PROCEDURE

4.2.1 Test Unit Preparation

In preparation for testing the units were checked for vertical plumb, and
if necessary, vertically aligned by placing washers under the appropriate
corners. The units were then set in plaster, firstly on a 20 mm steel
base plate and then at the top against the cross he~d of the machine~to en;..
sure even lo~d distribution to the units, Installation and testing of the
units was carried out wtth the spherical seating~ normally located at the
top platten~ removed, It was felt that a more uniform compression field
with lower tendency for buckling of the units at high strains would
result from the rigid contact at top and bottom.

For eccentrically loaded units a roller bearing was installed at top and
bottom of the test unit, to load the units ~t a predetermined eccentricity~
49 mm for 12 bar units and 33 mm for 8 bar units. These eccentricities
were calculated from a laminar analysis using a stress~stratn curve at
various strain gradients, The intention of the eccentric loading was to
produce a strain gradient varying from zero at the outside of the confined
core on one side to a maximum at the other. However 1 as the requl'red
eccentricity, based on the stress-strain curve obtained from the approp':"
riate concentrically loaded unit (unit 2 or 6}, was a functiDn of the
maximum compression strain (see Table 4.1} this could only be obtained
45

approximately. The chosen eccentricities satisfy the required strain


gradient at a peak compression strain of 0.01. Note that no allowance was
made for additional eccentricity resulting from flexural deformation of
the units (P- 6 effect). Results presented in Chapter Five show this to
be a significant factor.

Potentiometers were mounted on the brackets and strain gauges wired to the
appropriate recorder. The Dartec machine and recorders were even
calibrated for load, strain and displacement. A preload of 2 MN was used
for load calibration.

4.2.2 Testing Procedure

For the bulk of the tests which were performed at the rapid loading rate
of 20 mm /s, equal to an average strain rate of 0.0167 /s, a countdown
had to be given to approximately synchronise starting the time clock,
camera (recording at 4 frames per second), the video tape (for 3 tests),
the Hewlett-Packard two channel recorder, the Bryans-Southern recorder,
and the Dartec machine operation. The stroke controls of the test machine
were set to ramp longitudinal displacement at 20 mrrVs to a maximum
displacement of 50 mm, corresponding to an average longitudinal strain of
0.0417 and a test duration of 2.5 seconds. This was sufficient to ensure
the failure criterion of first hoop fracture was met in all but one
eccentrically loaded unit. Eccentric units had a stroke limit of only
30 mm to protect the potentiometers.

Test preparation and instrumentation can be seen in Figure 4.2,

TABLE 4.1 Calculated eccentricities

MAX. STRAIN 8 BAR UNIT 12 BAR UNIT


e~ e mm
.001 90 95
.004 64 73
before spalling)
.004 56 63
after spalling)
.01 33 49
.02 24 40
46

FIGURE 4. 2 Pteparation of Test Units. Testing Machin~ and


Recording Instruments.
47

CHAPTER FIVE

TEST RESULTS

5. 0 SU~1~1ARY

This chapter presents the results from the 25 tests performed, describes
the visual observations made and discusses trends in behaviour. Graphic
results are given for each unit and all are summarised in Table 5.1.

5.1 GENERAL BEHAVIOUR AND VISUA.L OBSERVATIONS

The appearance of vertical cover cracks was always the first sign of any
distress in the test units. These spread rapidly as cover crushing
occurred causing the cover to spall in quite large slabs. As expected,
this was particularly evident for units with closely spaced hoops, With
the cover gone load still continued to increase a small amount and then
began to decrease as the core concrete started to arch between the hoops
and longitudinal bars as its outer edges fell away. The load decreased
until buckling of the longitudinal bars occurred, This was invariably
associated with fracture of the hoops at or near the buckle. As the hoops
snapped the core concrete in the vicinity was reduced to fine rubble and
flowed, or was ejected from the core, The test ceased at the end of the
50 mm stroke travel,

In spite of the precautions taken to ensure axiality of the applied load


some strain gradient was evident in both the strain readings and the
inclined failure plane.

It was also found after testing that the horizontal bars supporting the
linear potentiometers had often deformed near one end, about 50 mm from
the longitudinal bars inside the core, indicating that initially plane
sectioRs were not remaining plane for the duration of the test, This
occurred despite the fact that the end sections remained plane and
parallel. To compound the problem further this phenomenon did not occur
in all tests, It was noticed that the horizontal bars bent only if they
were near a buckle in the longitudinal bars as shown in Figure 5,1,
TABLE 5.1 : SUMMARY OF RESULTS

BATCH UtHT NUMBER HOOP VOLUME HOOP TYPE RATE PEAK AVERAGE MAXIMUM STRAIN AVERAGE PEAK
NUMBER NUMBER AND DIAMETER OF SET OF OF LOAD STRA!N CORE AT STRAIN COMP'N
AND DIAHETER AND HOOP SHAPES LOAD LOADING AT STRESS MAXIf~UM AT STRAIN
CYLINDER OF SPACING STEEL Strain/ 5 PEAK CORE FIRST AT
STRENGTH LONGITUDINAL LOAD STRESS HOOP FIRST
mm/ 5 fcc
BA~~ r;- FRACTURE HOOP
(MPa) (mm (mm) m (MN) FRACTURE !

1 0 0 0 Concentric ,000033 4.38 ,00113 0,86 ,0018


"' I
2 .12 "' 20 10"' 72 .0182 " .
" 7.07 .00315 1.24 .0052 ,0052

,..,
"'
N
3
4
5
6
"
II

II

8 "'24
II

"
II

II
II

II

"
.0174
D "
Eccentric
"
Concentric
.0167
.000033
,0167
8.41
5.49
6.40
6. 72
.0030
.0027
.0033
.0041
1. 54

1.22
.0040

.0044
.0215
.0274
.0188
.0325
.0743
.0609
1

~
I 7 II II II II
.0167 7.85 .0032 1.47 .0038 .0271
... 8 II
" " Eccentric 5,54 .0044 .0206 .0649
9 II II II II
.0167 6,65 .0026 - -
10 II

" " " Not Tested


"' - - - - - -
11 0 0 0 Concentric .0167 5,75 .0012 1.17 .0012 - - !

12 12 "' 20 10 "' 98 .0140 II


" 8.50 .0025 1. 55 .003 .0167

..,a:>
N

I
13
14
15
16
17
II

II

II

II
10
12
12
10
"'
"'
"'
'"
72
88
64
72
.0182
.0224
.0309
.0182
.0134
D II

II

II

Not Tested
II

"
II

"'
8.65
8.80
9.40
-
.0035
.0033
.0052
-
1.65
1.67
1.86
-
,004
.0045
.0055
-
.0203
.0289
.0304
- -

~
8 "' 24 10 "' 98 Concentric .0167 7.90 .0027 1.48 .004 .0214
N
18 " 10 "' 72 .0174 II II
8.50 .0025 1.60 .0025 .0287
19 II
.0213 ,, 8.40 .0035 .0359
12 "' 88 " .003;~ 1.62
20 II
12 "' 64 .0293 " II
8.80 .0039 1. 75 .004 .0382
21 0 0 0 II ,,_. ';!' 4.78 .0018 0.97 .0018 -
22 12 "' 20 10"' 98 .0140
,, 7.30 .0017 1.41 .0028 .0238
"

..,
N

I
23
24
25
26
27
"
"
II

0
10 "'
12 "'
12 "'
0
II
72
88
64
.0182
.0224
.0309
0
D II

II

II

II
II

"
II
7.45
7.80
8.50
.QOOG-3J... 6.20
5.40
.0021
.0023
.0030
.0010
.0006
1.49
1.57
1. 79
1.27
1.10
.0030
.0035
.0040
.0010
.0006
.0287
.0284
.0323
-
-
~

-
" II II
.00167
,..,
28 II II
" Not Tested
.,.
- - - - -
29 ,, II II II
-- - - .,.
-- .
30 II
" II II
.,.
- - - ..j::::o
co
49

Buckling of
~==~====~I /24mm
longitudinal bar ~bar
bending support

bar with< ~~~~~~~*15~ 16mm


~ ~ ~ ::5 support
bar

FIGURE 5,1: Bending of the Support Bars

It appears that very high localised strai:ns associated with longitudinal


bar buckling caused local section distortion, However? as this phenomenon
occurred towards the end of testing, errors in longitudinal strain are not
considered to be great,

The problem of defining an 'ultimate~ concrete strain has recently


received some attention (e.g. Park~ Priestley? and Gill C29 lL Baker and
Amerazone(4l1 and Corley(4ll oath 1ist expressions for ultimate compress~
ive strain, but experimental studi~s by Potangaroa, Park and PtiestleyC42 l
and Gill, Park and Priestley(2B1 have i.ndicated these expressions to be
very conservative, In this series of tests marked strength degrada,tion of
the core was always initiated by fracture of an interna,l hoop~ Generally
a considerable number of hoops fractured before testing was complete~ and
50

the concrete was still able to carry a significant load after fracture of
(say) 3 or 4 hoops. It thus seems to be a reasonably conservative
decision to define the ultimate compressive strain as the strain at which
fracture of a hoop first occurs.

It should be noted that fracture of the outer (rectangular) hoops occurred


later, if at all, than fracture of the inner hoops. This was due to bond
loss on the outer stirrups, from loss of cover, resulting in an averaging
of strain over the core thickness.

Fracture of an inner hoop and subsequent local degradation of the core


also caused a loss of anchorage bond to the outer hoops, resulting in a
tendency to unwind rather than fracture.

The series of photographs included in this section show typical behaviour


during testing. Figures 5.2 and 5.3 show time sequences of photographs
for two units (12 and 15 respectively) taken at high speed with a Nikon
SLR camera with a motorised drive unit at approximately 0.35 second
intervals. Figures 5.4 and 5.5 show selected features of the tests
performed. The horizontal metal bands seen in these figures at the centre
of the units were made loose fitting (25 mm clearance all round) to provide
protection to the potentiometers but in no way provide any confinement.

Figure 5.3 shows vertical cracking of the cover has occurred by t = 0,39 s
just before a peak load of 8.5 MN was reached and by the next frame at
t = .79 s load has fallen to 6.4 MN. The next photograph at t = 1.13
occurs just before first hoop fracture and the explosive frame at t = 1.48
is the third hoop to fracture. By the last sequential photograph at
t = 2.22 s six hoops have fractured and two more will do so before the
end of the test. The final photograph shows the condition of the core at
the end of testing.

Figure 5.3 shows a similar sequence of photos to Figure 5.2. In this


series for unit 15 the effect of a higher confinement steel content can be
seen. Vertical cover cracks do not appear until t = 0.79 s, which, in
this case, concides with a peak load of 9.4 MN. The higher load demand on
the Dartec machine has slowed the loading rate from 20 mrn/s to 15 mm/s, so
first hoop fracture does not occur until t = 2,88 s while still carrying
a load of 7.5 MN. Falling slabs of cover can be seen at t = 2.53, 2.83,
3.10 s. As they fall from the top of the unit. Slabs of cover from the
central test region can be seen held in place by the protective bands.
,.......
G')
c
::0
rn

.
(.]1

N
..
c
:::s
- '.
M-
.......
N

-a
:Y
0
M-
0
0
~
CJ .
0
:::::l"'
Vl

(/)
t =0 t ;::: 0,39 t = 0.79 t = 1.13
:::::l"'
0
:E:
.......
:::s

c
ro
:::s
()
ro

(.]1
.......
t = 1.48 t = 1. 85 t = 2.22 End of Test
,........
GJ
c
;:o
fT1

.w
Ul

..
c
~
_.,
M-
.....
Ul

-o
:::J"'
0
M-
0
,
.0
QJ
0
;:s'
VJ
t = 0.19 t;:: 0.79 t = 1. 39 t = 1. 92
tn
;:s'
0
.:<.:
-:!
~

c
(!)
~
()
J\)

t = 2.53 t = 2.83 t = 3.10 End of Test


53

(a) (b)

(c) (d)
FIGURE 5.4 Photographs Showing Selected Features of Failure For Various
Units
54

(a)_ Co 1

~1 ~1
FIGURE 5.5 Photographs Showing Selected Features of Failure 'for
Various Test Units
55

Unit 5
e = 49mm

,/
~-~unit8
Unit 9 ~ e =33mm
=
e 33mm /"'
1/:/-
~~
o~~----------~------------~------------~------------~~-
0 0.005 0.010 0.015 0.020

AVERAGE LONGITUDINAL STRAIN

MID- HEIGHT LATERAL DISPLACEMENT VERSUS AVERAGE LONGITUDINAL STRAIN

FIGURE 5.5 (e) :

Figure 5.4(a) snows unit 17 after testing and is typical of all units
tested axially. Failure has occurred in the central region. longitudinal
bars have buckled and in this case for lightly confined units the core
has deteriorated rather badly,

Figure 5.4(b) shows a close-up of the fracture of three hoops from unit 12
(see also Figure 5.2). Note also the loss of anchorage to the outer hoops
and the bending of the horizontal potentiometer oars.

Figure 5,4(c) depicts the double bu-ckling of a 20 mm dia Grade 380


longitudinal oar and Figure 5,4(d} shows a close.,.up view of the excellent
condition of Unit 15 (see also Figure 5.31 after testing,

Unit 11, the plain unit tested at high speed can be seen in Figure 5,5(al
and (B). The cone failure at each end joined by a large vertical crack
56

was typical of all plain units tested. The failure was always brittle and
sudden with rapid loss of load capacity.

The eccentrically loaded units 4 and 5 can be seen in Figures 5.5(c) and
(d) respectively. Note the buckling of the longitudinal bars on the
compression face and the large evenly spaced tension cracks on the opposite
face. These cracks are slightly inclined due to the shear induced by the
moment gradient resulting from high, but variable P - ~ moments. As P - ~
moments are a maximum at mid-height and zero at top and bottom, a moment
gradient with height, and thus a shear force distribution, is involved.

Figure 5.5(e) shows the significance and magnitude of the central deflect-
ion as it relates to the applied eccentricity and longitudinal strain.

5.2 PRESENTATION OF RESULTS

Results are presented graphically for all units tested, in Figures 5.6 to
5.34 which show three graphs per unit. These show (a) Load versus
longitudinal strain curves, (b) Hoop steel stress versus longitudinal
strain, and (c) core concrete stress versus longitudinal strain. On all
of these is indicated the longitudinal strain at which the first hoop
snapped. The load-strain curves show the sequential fracture of ties as
steps of decreasing load. The cover concrete and steel components of the
load are shown individually and summed so that the remaining load carried
by the core is defined. The load carried by the cover was calculated
from the stress strain relationships of unit one, tested at the slow rate.
The results from unit eleven, tested at the fast rate, were not available
until much later in the series, For this reason unit one is compared
with all units rather than unit eleven 1 If there was a difference of
0.15 f~ in stress at a given strain between units one and eleven, the
difference in load carried by the cover would be 0.13 MN. The load and
stress carried by the core is then in error by less than 2%, It wa,s found
that if the cut-off strain for cover contribution was taken as 0,003 or
0.004, the stress strain curve peaked at that strain, It therefore
seemed reasonable to use the stress-strain curve for unit one to a, strain
of ,01 to model the behaviour of tne cover, From this graph the core
stress-stra,in curve was calculated by subtracting the components described
above from the total load. Also shown on the stress-strain curve is the
appropriate Modified Kent and Park stress-strain relationship detailed in
Chapter One. The Hoop steel stress-strain curve shows an average of the
hoop steel stresses and indicates when the hoops have yielded relative to
57

the longitudinal strain.

For the eccentrically loaded units (4, 5, 8 and 9) the first graph for
each (Figures 5.9, 5.11, 5.15 and 5.17) shows the load strain curves for
tension and compression strains at the outside of the core and for an
average of these. Note that initially both sides were in compression and
as the longitudinal strain increased, lateral displacement increased,
resulting in an effective increase in eccentricity of load (P- ~effect).
This effected a change in strain gradient resulting in the crack patterns
seen in Figure 5.5 (c) and (d). The second graph for each of the
eccentrically loaded units (Figures 5.10, 5.12, 5.16, and 5.18) shows a
comparison of theoretical and experimental load and movement plotted
against average strain. The theoretical curve was found from a laminar
analysis using nine 50 mm strips. Assuming an average strain over the
strip, the stress at that strain was found from the corresponding unit
with the same longitudinal and transverse steel and loaded axially at the
same rate. The cover contribution was treated in the same manner as for
axially loaded units.

Five units (10, 16, 28, 29, 30) shown in Table 5.1 as 1 not tested' were
left for future comparison of age effects,
Pt = 0
Ps--0
f~ =25.3MPa
A
10
Unit 1
(Plain Concrete)

-... Kent & Pork curve


~ 6
-
C)
~(J 1.0
for unconfined concrete
Static load test
Controlled f~ =25.3MPo
( )(
..,_J
lr) 10 Stroke controlled cylinder
4 t8 0.8
/!A\ test in DARTEC
e: I
V)
0.6 ,r! \ Unit ((Plain Concrete)
'/ I
~ I
Lu II'I
2 ct 0.4 I \\

~
u 0.2
I '\\
\\\\\
....... ~_
........ -..-.._ ___ _
0 -- --~
0 0. 01 0.02 0.03. 0 0.01 0.02 0.03
STRAIN STRAIN
(a) LOAD-STRAIN CURVE (b) STRESS -STRAIN CURVES
(J1
co
FIGURE 5.6 : UNIT 1 AXIAL LOAD, SLOW SPEED
C[
~ 400

~
(:l:
,/

Pt = 72DH20 = 0.0786 V)
fy = 434 MPa -...J
As= 3770mm; ~ 200
~
Ps = RJ0-72 =.0.018.2
! yh=
fc
309MPa
V)

~ 1st hoop
10 ~ 2S.3MPa
~ I fracture
Unit 2 i-/
00 0.01 0.02 0.03 STRAIN
(b) HOOP STF?ESS-LONGITUDINAL STRAIN
8
':,._lJ
)(

ti 7.4
llJ Modified Kent- Park

-
<: 6
g:
V) 1.2 Unit 2

-
~
~
Unit 1
(Plain Concrete)

4~
~

a"'( I ~ 1.0
I
a-...J
u~ 0.8
ClJ
..... I
~ I A
lJ I II
c: I I I I
8 I I \
I I
I
Cover+ Steel
I
I \
I

2 f+ f'......_ /
I
I 1st hoop 0.4
I \
\
I 1st hoop
I \ 1 fracture

Qij
ClJ
..._
V)
L/
fracture

0.2
\
\
\
', ...........
v
I
I
0 f =-=-=--= I t I I I ?:J;l:r- 0 :;:p..-
0 0.01 0.02 0.03 STRJJN 0 0.01 0.02 0.03 STRAIN
(a) LOAD-STRAiN CURVES (c) CORE STRESS-STRAIN CURVE

(J1
1.0

FIGURE 5.7 : UNIT 2 AXIAL LOAD, SLOW SPEED


-r 400
-
~
t l)
tl)
Lu
.....
Pt = 72DH20 =0.0786 ~
tl)
fy = 43t.MPa
As= 3770mm 2 cd 200
Ps = R10-72 = 0.0182 ._Lu 1st hoop
tl)
fyh= 309 /vfPa fracture
70
Unit 3
f~ =25.3 MPa ~
a
::t
v
I
0
0.01 0.02 0.03 STRAIN
1.6 (b) HOOP STRESS-LONGITUDINAL STRAIN
8
Modified Kent-Park
1.4
~l.> Unit 3

-- 6
)(
1.2
Unit 1

r'---."~
tl)
(Plain Concrete)
~ ~
I

J ~ 1.0
Q
"G: I tl)
a....! .....
Qj I

I I~1
<l.J I Lu I I
1....
I ~ 0.8
II
l.>
c: I
0
(.j I ~
Cover+ Steel
I 1st hoop ~ 0.6
(.j
'! \ i lsi hoop
2fl- /'-....._/
v
I

I
fracture
0.4 .,IJ I\\ \ .

\
I:..-----
I
I
/fracture

~I I 0.2
\ ~~
~~~~--~TRt
I I -
0 I _ _ _ ' -, -I - - - - - 1.
0 ~ ' I
JL_____
0 0.01 0.02 0.03 STRAIN 0 0.01 0.02 0.03 STRAiN
(a) LOAD-STRAIN CURVES (c) CORE STRESS -STRAIN CURVE
(J)
0

FIGURE 5.8 : UNIT 3 AXIAL LOAD, HIGH SPEED


Pt = 72DH20 = 0.0786
fy = 434MPa
A= 3770mm 2
P5 = R10-72 = 0.0182
yh=
-~ 309MPa
f~ =25.3MPa

-d
Q Unit 4
=
e = 49mm

-.1

Tension strain
Compression str01i1

4 I
I
--
r -1st hoop 1st hoop I
1 fracture fracture I
I
I 2
I
I I
I I
I I

-0.03 -0.02 -0.01 0 0.01 a~ aro a~ a~


.:..- ~

STRAIN 0.05 0.06 0.07 0.08


LOAD-STRAIN CURVES

(J)
f-1

FIGURE 5.9 : UNIT 4 ECCENTRIC LOAD, SLOW SPEED


Pt = 12DH20= 0.0186
fy =434MPa
As= 3770mm 2
Ps = Rl0-72 = 0.0782
fyh= 309MPa
10 f~ = 25.3MPa 400
e = 49mm
Unit 4

+
8
- Measured from 300 +
Measured from
test of Unit 4 test of Unit 4
---Calculated from
+
--- Calculated from
-6
stress-strain curve ~ stress strain curve
of Unit 2. ~
-
~
C)
(
1-..
dj200
I" ..... ...........
'/
of Unit 2.

.....
..... ,
-.J' ~ I ',,
4 ~
I ',
I ',
100 '
I \
2

0 ~0 ~
0 0.01 0.02 0.03 0 0.01 0.02 0.03
AVERAGE STRAIN AVERAGE STRAIN
(a) LOAD-AVERAGE STRAIN CURVE (b) MOMENT-AVERAGE STRAIN CURVE
THEORY- EXPERIMENT COMPARISON

0)
N
FIGURE 5.10 : UNIT 4 ECCENTRIC LOAD, SLOW SPEED
Pt = 720H20 = 0.0786
fy = f.3f.MPa
As= 3770mm 2
Ps = RJ0-72= 0.0182
fyh= 309MPa
-~ f(; = 25.3MPa.

-
C)
<
Unit 5
e = f.9mm

-.J

Compression strain

c_____
--
l
I f,
I I
I 1st hoop
L/acture 1st hoop I1 1st hoop I
fractu~ fracture I

I 2 ~
I I I
I I I
1 - I - -- I ------ I

-0.03 -0.02 - 0..01 0 0.01 0.02


-0. 03 0.04 0.05
;::.:::--.
STRAIN 0.05 0.06 0.07
LOAD-STRAIN CURVES

(j)
w
FIGURE 5.11 : UNIT 5 ECCENTRIC LOAD, HIGH SPEED
Pt =72DH 20 =0.0186
t'y =394MPa
As =3770mm 2
P5 =R10-:72 =0.0182
yh =309MPa
10 tc = 25.3MPa 400
e =49mm
Unit 5

8 - - Measured from - - - Measured from


test of Unit 5 300 test of Unit 5

--
- - - Calculated from
stress-strain -
e \
'X- .!L Calculated from
stress-strain
~ 6
Q
~
\
cuNe of Unit 3 <:
-
~

'-
<: 200
''" curve of Unit 3

C
-.J
\
' .... ~l ~
~ ""
4 '.......
.... ,
''-1
I
I
11
I
I' X
"" X

I 100 I

2 I I
I
I
I
o~------~------~------~--~~ 0 ~
0 0.07 0.02 0.03 0 0.01 0.02 0.03
AVERAGE STRAIN AVERAGE STRAIN
(a) LOAD-AVERAGE STRAIN CURVE (b) MOMENT- AVERAGE STRAIN CURV~
THEORY- EXPERIMENT COMPAR[SON

(J)
-t:>
FIGURE 5.12 : UNIT 5 ECCENTRIC LOAD, HIGH SPEED
-&
~
'-
, ,.
~
Pt = 8DH24 = 0.0179 2:
lr)
fv = 394MPa -.J

As= 3679mm
2
[jj 200
I-:
Ps = RJ0-72 =0. 0774 lr)
1st hoop
A fyh= 309MPa ~ fracture 1
70 f/: = 25.3 MPa ~ "--1
Unit 6 I
0
0 0.01 0.02 0.03 STRAIN

8 .........(.) (b) HOOP STRESS-LONGITUDINAL STRAIN


)(

~ --- Modified Kent-Park


Lu
2:lr) A
-
~
6 1.2
Unit 6
Unit 1
. :::--.....
-------- (Plain Concrete)

-------
'-
a 1.0 .........:::
<
-.J
4 ~/ ~ 0.8 I
~ I
ll
c:
a !
Cover+ Ste I 0.6
~~ e
l..)
I
I
211 -1 0.4 1st hoop I
fracture
1st hoop I 1

~
.._
Cll
....
Cll
fracture I 0.2
lr) '---1
I ;;z-0 0~-....._--'-'-_l_--=::.:::i:=------l..--,L_-.J._..J__~~
0. I03 STRAIN
00 0.07 0.02 0.07 0.02 0.03 STRAIN
(a) LOAD-SrRAIN CURVE_$ /c) COB~. STRESS -.STRAIN.CURV~.
CJ)
(J1

FIGURE 5.13 : UNIT 6 AXIAL LOAD, SLOW SPEED


-
~ 400
~
lr) . ,..-.
tj
Pt = 8 DH 21. = 0. 0179 ~
ly = 394 MPa lr)

As= 3679 mm 2 Lj 200


Ps = RJ0-72 = 0.0171. ~
lr) 1st hoop
ylf 309MPa fracture
~
v
1
10 f = 25.3MPa
~
Unit 7 I
0 -~-- ---::z:;-.
0.07 0.02
1.6 (b) HOOP STRESS-LONGITUDINAL STRAIN
8

, 7.4 ~ ~ -~ --- Modified Kent-Park


..._u Unit 7
- 6 ~ 1.2 I["-.. ---- u
mt 1
~
-....;.
~ , ...........__.............._ (Plain Concrete}
Lu
a ~ 1.0
d lr)

~0.8 ~I
...,J
.....Q)
Q) I
4
..... I 1\
ll 1st hoop 1 Lu I I
c:
fracture a:: I \
Q 1
~ 0.6 1 1 I 1st hoop
~
Cover + Steel <...>
.
8 / \ I fracture
2 I 0.1. \ l--/
._
Q) 0.2 ~
'\ I1
~
lr) ' ' ..... ...._ I
0~----==------~~-----~--~ ~ 0 ----- ;:J:;:a-
0 0.01 0.02 0.03 STRAiN 0 0.01 0.02 0.03 STRAIN
,

(a} LOAD-STRAIN CURVES (c) CORE STRESS-STRAIN CUF?VE


Q)
O'l

FIGURE 5.14 : UNIT 7 AXIAL LOAD. HIGH SPEED


Pt =8DH24 =O. 0179
fy =394MPa
10
As= 3679mm2
Ps = R10-72 =0.0174
--
~ 8
Unit 8
yh= 309MPa
f =25.3MPa
Cl e = 33mm
d-...J
Compression strain

c;:;.: _ _ _ _ _ _ __

4
I I I
1 1st hoop 1st hoop 1
I fraCture I
1 . fracture 1st hoop I
~ fractu~ ~
I 2 I
I I
I I
l I
I I I
I I I

-0.03 -0.02 ,0.01 0 0.01 0.02 0.03 0.04 o.os


STRAIN -- I I 1:=:--
0.05 0.06 0.07
LOAD STRAIN CURVES
en
'-.1

FIGURE 5.15 : UNIT 8 ECCENTRIC LOAD, SLOW SPEED


Pt = 8 DH24 = 0.0779
y = 394 MPa
As= 3679mm 2

10
A
~ 1~1 Ps =R70-72 =0.0174
fy = 309MPa 400
f~ = 25. 3MPa
Unit8

8
I
I

- - Measured from 300 Measured from

--
test of Unit 8
- - - Calculated from -
<:
E: -x-lt..
test of Unit 8
Calculated from

~
6 stress-strain
curve of Unit 6 -
~

J......
..,.,--x-x--
stress -strain
curve of Unit 6
a
d ....... ~ 200 /
...,J 0 1/
~
4 I
I
I
100
I X

2 k

0 ------l.~--:::a- 0 I I l=-=
0 0.01 0.02 0.03 0 0.07 o. 02 0.03
AVERAGE STRAIN AVERAGE STRAIN
(a) LOAD-AVERAGE SffiAIN CURVE (b) MOMENT-AVERAGE STRAIN CURVE
THEORY- EXPERIMENT COMPARISON

01
co
FIGURE 5.16 : UNIT 8 ECCENTRIC LOAD, SLOW SPEED
pt = 8 DH21. =0.018
fy = 391./vfPa
As = 3679mm
Ps = Rl0-72 = 0
fyh= 309/vfPa

-
.~
._
Unit 9
tc = 25.3 MPa
e = 33mm
Q
~
a--.1
Tension Compression strain
strain

c _______
Average strain

I.

-0.03 -0.02 -0.07 0 0.01 0.02 0.03 0.01. 0.05


STRAIN --- ~
0.05 0.06 0.07
LOAD-STRAIN CURVES
(j)
\.D

FIGURE 5.17 : UNIT 9 ECCENTRIC LOAD, HIGH SPEED


Pt = 8DH24 = 0.018
f) 1 = 394/V!Pa
As =3679mm 2'
Ps = R!0-72 =0.01
A fyh = 309/V!Pa
10 f~ = 25.3/V!Pa
e = 33mm
Unit 9

8
Measured from 300 - - Measured 'from
test of Unit 9
---Calculated from
-
E:
<:
test of Unit 9
- _ Calculated from
-6
~
-
a
I -~
I
I

\
stress-stra1i1
curve Unit 7 -
~

1-.
<:
Lu 200
X
stress-strain
curve Unit 7

\ ~
C I \ a
-..J ~.
f.
X

II
100
II
2

0 ~0.._ _ _ ___.___ _ _ _..__ _ __.___ _---:=--


0 0.01 0.02 0.03 0 0.01 02o. 0.03
AVERAGE STRAIN AVERAGE STRAIN
(a) LOAD-AVERAGE STRAIN CURVE (b) MOMENT-AVERAGE STRAIN CURVE
THEORY-EXPERIMENT COMPARISON

-....J
0

FIGURE 5.18 : UNIT 9 ECCENTRIC LOAD, HIGH SPEED


Pt =0
Ps = 0
f = 24.8MPa

10
Unit 71

1.2
- 6
~
~

~ ':.._u 1.0 Unit 71


C )(
.----- Unit 1 {Plain Concrete)
-.1 V)
A
4 t] 0.8 II
I I
g: I I
V) 0.6 I \
I I
~ I I
l!.J 1 I
2 ~ 0.4 \
I
\
25<....> \
\
---
' ' .... .... ..... __

0 I ~ 0 - ;;:J:::to-
0 0.01
- 0.02 0.03 STRAIN 0 0.01 0.02 0.03 STRAIN
(a) LOAD- STRAIN ci.Jml (b) STRESS -STRAIN CURVES

"-..)
I->
FIGURE 5.19 : UNIT 11 AXIAL LOAD, HIGH SPEED
-
~
...._
1.00
(./)

~
~
-------
Pt =72DH20 =0.0186 (./)
fy =1.31.MPa.
Gj 200
As= 3770mm 2
~
Ps = RTO- 98 =O.Oi40 (./)
1st hoop

~
10 yh= 309/v!Pa
1 fracture
Unit 12
~ =21..8/v!Pa ~

0.01 0.02 0.03 STRAIN


(b) HOOP STRESS-LONGITUDINAL STRAIN
8
II \
\(.) 1.4 --- Modified Kent-Park
(./) Unit 12
~ 1.2 ----- Unit 1

""
;? 6
~ ~ (Plain Concrete)
(./)
a 1.0
C ~ I
-...J ClJ
....... Lu
4 ClJ
..... ~ 0.8 1,\I ::1
~
(J
c: I I
0
<...> <...> I I I
Cover 0.6 I
+Steel I I : 1st hoop
2
II
" .
-
~
/
--.
I \
\
V fracture
1st hoop \ I
I \
1 fracture \ I

O~' ~I
0.2 \ I
~ ' ', ............... ______ _I
I I I ~0
0 0.07 0.02 0.03 STRAIN 0 0.01 0.02 0.03 STRAIN
(a) LOAD-STRAIN CURVES (c) CORE. STRESS-STRAIN CURVES

"-..1
N

FIGURE 5.20 : UNIT 12 AXIAL LOAD, HIGH SPEED


-a
Q. 1.00
~
V)
V)
Lu
Pt = 72DH20 = 0.0786 ~
V)
fy = 1.3~MPa
Gj 200
As= 3770mm 2
~
Ps = Rl0-72=0.0182 V) 1st hoop
fracture
1- fyh= 309MPa ~ 1
10
f~ = 21..8MPa
a ~
Unit 73 ::c: I
:l>-
0.01 0.02 0.03 STRAIN
1.6~ / " ' (b) HOOP STRESS-LONGITUDINAL STRAIN
11
-
8

"' \._U
)(
1. 4

1.2
r----~
-- Modified Kent-Park
Unit 13
-----. Unit 1
<: 6

h' l
V) (Plain Concrete}
~
..._ V)
Lu I . ___
Q
~ 1.0
C .....ClJ V)
-...J
I.
ClJ
1...
u I Lu I
c:: I C3 0.8 I
Cover+ Steel
8 I 0::
(.j
I
I
<: 0.6 I
8 I 1st hoop
2~f'....._/ t -l 1st hoop 0.4 \
I fracture
1-c- ____.,/
1 fracture \
\

~ \

0~ ~~
0.2
' ,. _
...... ......

0 0.01
I
I
II
0.02
I ~
0.03 STRAIN
0
0 0.01
------ 0.02
--->--
0.03 STRAIN
(a) LOAD-STRAIN CURVES (c) CORE STRESS-STRAIN CURVE
"'-J
w
FIGURE 5.21 : UNIT 13 AXIAL LOAD, HIGH SPEED
-~ 400
-......:.

~
lJ.J
Q::
Pt = 72DH20 =0.0186 h.
ll)
fy = 431. MPa Gj 200
As= 3770mm 2
~
J.~ = R72-88 =0~0224 V)
1st hoop
10.
~1h= 296MPa ~ fracture I
fc =24.BMPa ~ "--J'
Unit 11.
'l..l 0'--------l.._ _ _ ____L._ _ __j__.J,__ _ _ ~

~ ! 0.01 0.02 0.03 STRAIN


8 ~ 1.6 ~ / ' \ . (b) HOOP STRESS- LONGITUDINAL STRAIN
~
1-:
ll) 1.4 Modified Kent-Park
Unit 11.
1::
-~ 6 ~ 1.2
Unit 1

-
CJ
~(.)
(Plain Concrete)

J
( I 1.0
I
..._
<lJ
I
It~
-..J <lJ
.... I
l..l
c: 1st hoop I
Cover + Steel 0
I II I I
o.6 ~1
(.)
fracture
\ I
~ t \ 1st hoop I
~
II / [ I fracture 1
0.4 \
~I
2
\
\
--
ClJ
0.2
\
\
~
ll) ............ ........ _ I
0 l --.. I I I I I ;)>- 0 ----- ;'E:3-
0 0.01 0.02 0.03 STRAIN 0 . 0.07 0.02 0.03 STRAiN
(a) LOAD-STRAIN CURVES (c) CORE.STRESS-STRAIN CURVE

"'
+:>
FIGURE 5.22 : UNIT 14 AXIAL LOAD, HIGH SPEED
Pt = 72DH20 =0.0186
fy = 431. MPa
As= 3770mm 2
~u
,. , =R12-64= 0.0309
'5 )( Modified Kent-Park
yh= 296MPa ~ Unit 15
70 f~ =21..8MPa ~
1-::
Unit 1
U) (Plain Concrete)
~ 1.8

8 ~~ 1.6
II l ~ (..)

1.4
I
.r------ .........__
--
~
6
I
I
I
1.2 I
,
------.-.JI
a
I 1. 0 ~I, I
I
d
....,j
I
4 1st hoop I 0.8 ft-1 1A1 1st hoop j
Cover + Steel I fracture
frcrcture II 1 1

\j 0.6 ~II \I ~I
I I
II . / L I
I \ I
2 0.4
'. I
\ I
0.2 \ I
' , ...... _ I
0, - I I f !I >-- 0 I I --- ;;z::..-
0 0.01 0.02 0.03 STRAIN 0 0.01 0.02 0.03 STRAIN

(a) WAD-STRAIN CURVEs_ (b) CORE STRESS~STRAIN CURVE

"-!
CJl
FIGURE 5.23 : UNIT 15 AXIAL LOAD, HIGH SPEED
-;p ~.ao A
-~
~

---
Pt = 8DH21.=0.0179 e:
lr)
fy = 391./V/Pa ...,J

As= 3679mm 2
tti 200
h..:
Ps = Rl0-98 = 0.0131. lr) 1st hoop

10
A
Unit 17
yh= 309/V/Pa
f(: = 21..8MPa
~
~
0
0 0.07 0.02
v fracture

0.03 STRAIN
~

(b) HOOP STRESS-LONGITUDINAL STRAIN


8 ~(.J
)(

~ 1.1. --- Modified Kent-Park


IJ.J
-~ 6
g:
lr) 1.2
Unit 17
----- Unit 1 (Plain
-
a
~
~ 1.0
Concrete)

( <...,)
...,J
<:
I. ,--, a<...,) 0.8
A
f\
~
<l! I I
1... I I
u 1 1 I 1st hoop
Cover+ Steel c:
8 fracture
I
/ \ V I

II 1st hoop
0.1. \ I

v
2 fracture
II I~ '--L - \

--
<l!
(lJ
......
0.2 '\
',, ......_
I
II
lr)
I
'U ~
I I I >- 0 ------ I I I ~
0
0 0.01 0; 02 0. 03 STRAIN 0 0.01 0.02 0.03 STRAIN
(a) LOAD-STRAIN CURVES ( c ) CORE STRESS -STRAIN CURVE
........
0"1

FIGURE 5.24 : UNIT 17 AXIAL LOAD, HIGH SPEED


-. Q
400
-
Q.
~
l f)
V)
lJ.J
,.. ... -
Pt = BDH21. =0.0179 ~
lf)
fy =391. MPa
2 Ljj 200
As= 3679mm

10
A
Ps = R!0-72 = 0.0171.
yh= 309MPa
f = 21..8MPa
~
lf)

~
a
velure 1st hoop

::r:
Unit78
0 ~
0.01 0.02 0.03 STRAIN
'u A
~ 1.6 ~ .-.... (b) HOOP STRESS LONGITUDINAL STRAIN
8 lf)
lf)
lJ.J Modified Kent-Park
P: 1.1.
h: Unit 78
lf)
----- Unit 1 (Plain
~ 1.2 Concrete)
~
(.)
<: 1.0
a(.)
I
I 0.8 fH l\
1 1st III\
I I

v
hoop
I \ I
: fracture I \ : 1st hoop
I \
\ I fracture
0.1.

0.2
\
\
\
\
L/
I
I I ::::-- 0 I
' ......_ __ _
I
, -- I
I ! .....
'
0 I -
I I I >--
0.03 STRAIN 0.01 0.02 0.03 STRAIN
(c) CORE STRESS-STRAIN CURVE

'-...!
'-...!

FIGURE 5.25 : UNIT 18 AXIAL LOAD, HIGH SPEED


-
Q
~
~
CJ
400
.
ll)

~
Pt = 8 DH 24 =0.0179 r:::
ll)

fy =39f.MPo -..J

As= 3619mm 2
[jj 200
h..
ll)
P5 =R72-88 =0.0213 1st hoop
A
10
fyh=
f~'
296MPo
=2f..8MPa
8::t fracture
~
1

c I
Unit 79
~(..)
0
)(
0.01 0.02 0.03 STRAIN
ll) (b) HOOP STRESS-LONGITUDINAL STRAIN
8 ~ 7.6
g: Modified Kent-Pork
ll) 7.4
Unit 19
~
-
~
6 ~ 1.2
.Ir----. --.. --....:::' ..
----- Unit 1 (Plain Concrete)

- <....>
<:
a
a
<
-..J
f.
II
.....Q)
~
u
1st hoop
<....> ,
I
{,~I \
---------, I
c: fracture ., I
Cover + Steel ~ It
\ i
8 0.6 !lf I
~
I
2
I 0.1.
' I\
1st hoop
fracfu~
\
--
Q) 0.2
\

' ', l

or .___ j I
Q)
......
ll)
r I ,
...... ...... ........ -... ___ _ I
0 ~ 0.02 0.03 STRAIN
0.01 0.02 0.03 STRAiN 0 0.07
(a) LOAD-STRAIN CURVE_S (c) CORE STRESS-STRAIN CURVE N
co
FIGURE 5.26 : UNIT 19 AXIAL LOAD, HIGH SPEED
--..
d:~
-~
l r)
400

~
Pt = 8DH21. = 0.017 lr)

y = 391.MPa Ltj
~ 200
As= 3679mm 2
lr)
Ps = R72-61.=0.0293 1st hoop
~
10
fyh= 296MPa
'f: = 21..8MPa ~
fracture
.......___ ..,
1

I
Unit 20
0 ~.----------------------------------~~
Lu ~\.J 0.01 0.02 0.03 STRAIN
f...: )(
Lu!.0 (b) HOOP STRESS-LONGITUDINAL STRAIN
8
f5~ Modified Kent-Park
~~ Unit 20

r ----
-- --..-----
<...)lr)
Unit I (Plain Concrete)

--.. 6 jZ2 i --.--:_~


-
~
C)
C <lJ
......
<lJ
I ,
17.0 !
I. I
~

I
I
-...1 ~-
I.
(..)
1st hoop 1.8 j' ,\ I
I
8 fracture 11 1st hoop l

~
'/ I fracture l
Cover + Steel .6 I \ '-._ !
1 I ~
2 r;-J, I ~.4I I \
\
lI
--<l.J
<lJ 1.2r
I
\ ', I
I
o! '-L.
0.07 0.02
......
lr)
I
0.03 STRAIN
J_
0
''------
0.07 O.D2 0.03
i _.
STRA 'J'I/
(a) LOAD STRAIN CURVES (cf CORE.STRESS STRAIN CURVE
.........
l.O

FIGURE 5.27 : UNIT 20 AXIAL LOAD, HIGH SPEED


Pt =0
Ps =0
f =24.2MPa

10
Unit 27

- - U n i t 21
- - - - Unit 1
- 6 (Plain concrete)
~
-....;;.
.........lJ 1. 0
a
~
)(

......(
4 ~ 0.8
~
t-.::
ll) 0.6
~
ll.J
2 ~ 0.4
~
<..> 0.2

0 L I~ 0 -- ~
0 0.01 0.02 0.03 STRAIN 0 0.01 0.02 0.03 STRAIN
(a) LOAD-STRAIN CURVE (b) CORE STRESS-STRAIN CURVE

co
0

FIGURE 5.28 : UNIT 21 AXIAL LOAD, SLOW SPEED


(/)
(/)
Lu
Pt = 72D20 = 0.0786 ~
(/)
fy = 272MPa -.J
As= 3770mm 2 }jJ 200
h:
Ps = R10-98 = 0.07.40 (/)
1st hoop
A
v
yh= 309MPa Q. I fracture
10 f = 24.2MPa 8
::t:
Unit 22
0 ~
0 0.01 0.02 0.03 STRAIN

8 (b) HOOP STRESS -LONGITUDINAL STRAIN

7.4
Modified Kent- Park
Unit 22
--.. 6 ':.....u1.2 Unit 1 (Plain Concrete}
~ ~"'"- . ~
)(

t3 1.0 I
-......;. (/)

a ~.
C ~ I
-.J 4 (/) 0.8 .............:
/A ""i
~ ,I I
~
QJ
.....
u
I Lu
~ 0.6
!" \
I I
I
I 1st hoop
fracture

v
I 1
Cover + Steel c:
0 1st hoop ~
2 lJ lJ 0. 4
fracfute \
\ I
\
...._ 0.2 fl-
\
\
I
QJ
'' _ I
or
. 0
.___ ,
0.01
QJ

ti)y,
0.02
1
,

~o
0.03 STPAJN 0
...... .....

0.01
----- .'J.02
I

0.03 STRAit-'
~

(a) LOAD-STRAIN CURVES (c) CORE. STRESS -STRAIN CURVE


CXl
t-'

FIGURE 5.29 : UNIT 22 AXIAL LOAD, HIGH SPEED


a4ooA
a...
~ /
~
Pt = 12D20 =0.0186 Lu
fy = 272 MPa e:
V)
As= 3770mm 2 [jj 200
Ps = RT0-72 =0.0182 ~ 1st hoop
fyh= 309MPa (/)

10
Unit 23
f = 24.2MPa ~
~
0 ~------------~--------~----~~----~~
v
I fracture

0. 01 0.02 0.03 STRAI


(b) HOOP STRESS-LONGITUDINAL STRAIN
8 1.6

1.4 Modified Kent-Park


Unit 23
~(..)

--
~
6
V)
~ ~2 r--~
-
Unit 1 (Plain Concrete)

CJ ~
(/)
1.0 1 I ----
~- --.._
d fL1 ~ ------~
~ o.8
I
-.J
4 ..
..._
(J) I
Lu
f,\
,i \
I
CJ
<...
(..) I 5 06 lf \ I 1st hoop
I ~ \ I
II
Cover + Ste~::/
/ 18 I (.) o.4 / \ ~
fracture

2
I 1st hoop l
' ---.. TQ;
(J)
~ure 0.2 tI '\,. .,
'
I
II
0 II' ~ I f ti I I '):::>- 0 ,_____ I I ~
0 0.01 0.02 0.03 STRAIN 0 0.01 0.02 0.03 STRAIN
(a) LOAD-STRAIN CURVES (c) CORE STRESS..;.STRAIN CURVE
co
N

FIGURE 5.30 : UNIT 23 AXIAL LOAD, HIGH SPEED


-D
1.00

-lD
Q.
~
V) /

Pt = 12020 =0.0186 e::ll)


fy = 272MPa
A5 = 3770mm 2 Gj 200
Ps = R72-88 = 0.0221.
~
ll)
1st hoop
A fyh= 296MPa
10
Unit 21.
f~ = 21..2 MPa v
I fracture

0. ~
0.01 0.02 0.03 STRAIN
(b) HOOP STRESS-LONGITUDINAL STRAIN
8 1. 6
Modified Kent-Pork
...-...;:u 1. I. Unit 21.
)(
Unit 1 (Plain Concrete)
-
~
6 ~
l!j
1.2 r-------
I ~

..._ e:
l/) 7,0
I
a I

( I L ~ I I
...J I. Qj I ~ 0.8 !',\ I
..._ I ~ 1I I 1st hoop

Cover+ Steel 8
~
u
! 8 0.6 f\ ~ure
2
/ I
1 1st hoop
j fracture
0.4 I . \

\
\
I
I
I
_ I _/ 0.2 ', I
~ Qj
Qjl I ' ,, 1
0 I I V) _j_ I ~ 0 ------ I I I >-
0 0.01 0.02 0.03 STRAIN 0 0.01 0.02 0.03 STRAIN
( o) LOAD-STRAIN CURVES (c) CORE STRESS-STRAIN CURVE
co
w
FIGURE 5.31 : UNIT 24 AXIAL LOAD, HIGH SPEED
- .t.OO (b) HOOP STRESS-LONGITUDINAL STRAIN
d:
~ .........
U)
U)
Pt = 12020 =0.0186 ~
fy =272/v!Pa I-:
U)
2
As= 3770mm --.J 200
Ps =R72-64 =0.0309 lB
A fyh= 296/v!Pa
I-:
U)
1st hoop
fracture~
10 ~ = 2L..2MPa ~
a
Unit 25
::t:
0
1.8

8 1.6

1.4
r---
~I --------~
1 ------

- 6 1.2
~
-
CJ ~(.) 7.0
)(
I

I
I

II
6 .!!
Ql U) ' } -. - Modified Kent-Park
--.J 4 I...
(.) ~ 0.8
c: /\ - Unit 25 I
a ~ f\ ----- Unit 1 (Plain Concrete) 1

U) 0.6 1 \
1

1st hoop
Cover + Steel ~ f \ 1 fracture
~ ! ./
1

I \
2 0.4 .\ ~
~ \ I
l) Q2 \ I

0
~
d I
--1
JJ
U) I
I
I ::r:.--0
I ','",,_ I----- I I
I
I ---.:::c:-
0 0.01 0.02 0.03 STRAiN 0 0.07 0.02 0.03 STR,J.IN
(a) LOAD-STRAIN CURVES (c) CORE STRESS-STRAIN CURVE
00
..j:::>

FIGURE 5.32 : UNIT 25 AXIAL LOAD, HIGH SPEED


Pt = 0
P5 =0
f = 2G.2 MPa

A
10
Unit 26:

8
_;__Unit 26
7.4
---- Unit 7

--
~ 6
~(J
7.2
(Plain concrete)

C) )( 1. 0
6
....,J ~
4 e:
Lu
0.8 ~
II
V)
I \
~ 0.6 I \
I \
~ f I
2 ~ 0.4 I
a \
<...>
0.2
' .....
.....
.........
I I I ;::=- O I 'j-- I I .:;:.s-
00 0.01 0.02 0,03 STRAIN 0 0.01 0.02 0.03 STRAIN
(a) LOAD-STf?AIN CURVE (b) CORE STRESS-STRAIN CURVE
co
(.)"(

FIGURE 5.33 : UNIT 26 AXIAL LOAD, HIGH SPEED


Pt = 0
P5 =0
f~ ~~ 24.2MPa

10
Unit 27

- U n i t 27
----Unit 1
(Plain concrete)
- 6 1.2

-~
Cl
~u 1.0
)(

6
--.J ~
4 ~
I-:
0.8
V)

~ 0.6
~
l_)
2
~
l_)
0.4 "
I
0.2 1..I
' ....... ................ ____ _
0 0 I ~
0 0.01 u.02 0.03 STRAIN 0 0.01 0.02 0.03 STR..J,IN
(a) LOAD-STRAIN CURVE (b) CORE STRESS-STRAIN CURVE
co
en
FIGURE 5.34 : UNIT 27 AXIAL LOAD, MEDIUM SPEED
87

5.3 DISCUSSION OF RESULTS

5.3.1 Rate of Loading

Figures 5.35, 5.36 and 5.37 illustrate the influence of loading rate on
the core compressive stress-strain curve for plain concrete (units 21, 26,
27), 8 bar (units 6, 7, 17), and 12 bar (units 2, 3, 13) units respect-
ively. Table 5.2 shows that strength increase due to rate of loading is
typically 25% at peak stress but by a strain of 1~ ~ 2% this may have
fallen to (say) 10%. The falling branch in this region (peak stress to
(say) 2% strain) is much steeper than the slow test and at larger strains
tends towards the curve of the corresponding slow test.

The slow rate (0.000003 = .004 mm/s) was representative of that rate at
which cylinders were loaded in a load controlled, standard cylinder test;
and of previous research tests which took (say) 30 minutes to complete.
The fast rate (0.0167/s = 20 mm/s), as indicated, was representative of
the loading rate expected during a seismic attack and the medium rate
(0.00167/s = 2 mm/s) to indicate the sensitivity of strength increase to
changes of loading rate.

Note in this series of tests the fast loading rate was only indicative of
that expected during an earthquake and may be exceeded in reality. However,
the influence of small changes in loading rate (i.e. factors of 2 or 3)
are likely to be small in comparison with the variation from slow to fast
within these tests (i.e. a factor of 5000) as shown in the figures
mentioned above.
88

TABLE 5.2 The Effect of Loading Rate on Peak Stress

UNIT RATE RELATIVE fcclf~ RATIO ARITHMETIC


t:./S AVERAGE FOR
mm/s FAST TESTS
1200
PLAIN UNITS
1 .0000033 0.0002 .86 .89
11 .0167 1.00 1.17 1.21 1.260
21 .0000033 0.0002 .97 l.OO(base)
26 .0167 1.00 1.27 1. 31
27 .00167 0.10 1.10 1.13
8 BAR UNITS
6 .0000033 0.0002 1.22 l.OO(base)
7 .0167 1.00 1.47 1.20 1.250
18 .0167 1.00 1.60 1. 30
12 BAR UNITS
2 .0000033 0,0002 1.24 l.OO(base}
3 .0167 1. 00 1. 54 1.24 1.245
13 .0167 1.00 1. 55 1.25

avera 11 (Fast) Average 1.252


89

---- Unit 21
t = 0.00003/s
Unit 26
1.2
t = 0.0167 /s
--- Unit 27
. t = 0.00167/s
1. 0
'..._u
)(

ll) 0.8
ll)
Lu
e:
ll)
0.6
~
Lu
0::
lJ
::<: 0.4
a
lJ

0.2

0.005 0.010 STRAIN

CORE STRESS-STRAIN CURVES

FIGURE 5.35 Plain Concrete Units Loaded at Different Rates


90

Experimental Results
Units 6, 7 & 18
p5 .:: 0.0174
1. 8
- - - Modified Kent & Park
---Proposed Curve
1.6

Unit 7
1.4

)( 1. 2
ll)

lD
~
ll) 1.0
Lu
1--
l.u
5<:: 0.8
0
l.>
0.6

0.4
1st hoop
fracture ~~

~2 I
0 1------L------'-------'------'---'---1 -~
0 0.01 0.02
...._____.._!

0.03 STRAIN

CORE STRESS-STRAIN CURVES

FIGURE 5.36 8 Bar Units Loaded at Different Rates


91

1. 6
Experimental Results
Units 2, 3 & 13
1. 4 P5 -= 0.0182
- - - Modified Kent & Pork
- - - - Proposed Curve

0.8

0.6

1st hoop
0.4 fracture

0.2

0 ~------------~------------~~~--------~L-----------~~
0 0.01 0.02 0.03 STRAIN
CORE STRESS-STRAIN CURVES

FIGURE 5.37 12 Bar Units Loaded at Different Rates


92

A proposed alteration to the Modified Kent and Park curve for confined
concrete is shown in Figures 5.36 and 5.37 to account for the influence
of dynamic loading rates. This is achieved by applying a factor of 1.25
to the peak stress, the strain at peak stress and the slope of the falling
branch of the modified Kent and Park curve. The stress-strain curve for
dynamic axial loading then becomes (refer figure 1.8)
Region AB . Ec ~ 0.002K

5.1

fyh
where K = 1.25 ( 1 + pS ~) 5.2
c

Region BC

K f~ [ 1 - zm (Ec - 0,002K)] 5.3

but not less than 0.2Kf~, where

z = 0.625 5.4
m --~~~~~------------------
3 + 0.29f~ 3 JhO
145f' - 1000 + 4 ps;~ - 0.002K
c h

The proposed curve is also shown in Figures 5.38 and 5.39 with the
different levels of confinement for the 8 and 12 bar units respectively.

5.3.2 Confinement Ratio

The effects of confinement ratio on the behaviour of the units tested at


the fast loading rate is shown in Figures 5.38 and 5.39 for 8 and 12 bar
units respectively. The proposed curves based on Equations 5.1 to 5.4
are included for comparison.

Note from these figures the large increases in strength obtained from the
rapid loading rates (as discussed above} and confinement of the core
concrete.

For a quite modest confinement ratio (ps about 0.018) an average increase
in core strength of 23% was obtained for the slow tests while a strength
93

increase of about 55% was obtained for corresponding units under rapid
loading. For high confinement ratios (ps about 0.030) strength increases
of about 80% were obtained.

The confinement ratios are 0.0134, 0.0174, 0.0213 and 0.0293 for the 8 bar
units, and 0.0140, 0.0182, 0.0224, and 0.0309 for the 12 bar units for hoop
set spacings of 98, 72, 88 and 64 mm respectively. It will be noted from
the similarity between the two centremost curves in each figure that the
strength increase from confinement has been largely offset by a reduction
resulting from increased spacing of the larger diameter hoop steel used
for the units with the higher confinement ratio.

It will be seen that the proposed curve represents the dynamic tests as
accurately as the Modified Kent and Park curve represents the static tests.
Time in this study does not permit a full regression analysis of the vari-
ables but the proposed curve is considered to give a good representation
of behaviour at this stage. The proposed curve, like the modified Kent
and Park curve, peaks at a lower strain than generally obtained during
testing, which, as previously mentioned, was noted by its authors, and is
also accepted here on the basis that the influence on moment-curvature
calculations is insignificant.

The effectiveness of the transverse reinforcement is seen in both the


increase in strength and by the maintainence of a large proportion of the
load at high longitudinal strains. Figure 5,40 plots the peak stress
against confinement ratio. It will be seen that a factor of 1.25 applied
to the Modified Kent and Park equation appears to be a good conservative
estimate of the influence of loading rate. Shown also is the strength
increase Ks predicted by Sheikh and Uzumeri (Equation 1.18). This was
calculated for a 12 bar unit with a hoop spacing of 72 mm and hoop steel
stress of 309 MPa, at various confinement ratios, It will be seen in
Figure 5,40 that K5 gives a good estimate of strength increase for rapidly
loaded units at medium levels of confinement but appears to err at high
and low levels of confinement. However, no allowance has been made for
dynamic loading rates, and would therefore overestimate the strength
increase obtained from slow loading rates.
94

Experimental
Results
1. 8 Proposed
Curves

1. 4

1.2

~u
1. 0 I
)( Unit 19
=0.0213
P5
I
1.1)
1.1)
Lu
0.8
Unit 18l .........
I
~
1.1) Unit 17
P5 =0.0174 ~ I
Lu
......
0.6 P5 = 0.0134 I I
Lu
5<:
I I
a(..)
0.4 1st hoop
fracture
I
I
0.2
I
0
0 0,01 0.02 0.03 STRAIN
L
CORE STRESS- STRAIN CURVES

FIGURE 5.38 Effect of Confinement Ratio for an 8 Bar Unit


95_

Experimental
Results

1. 8
---- Proposed
Curves

1. 6
-.... ...........

--
...........
1. 4 -.... ......;,_
-....
'lJ
......
)(

ll)

~
1. 2 -
I
I
::: 1. 0
I
ll)
I
Lu
..._
l.t.J
I
Q:
l,)
0.8 Unit 12 I
<::: p_'i=0.0140
a
l,) I I
0.6 I I

0.4
1st hoop
fracture
-H
I I
0.2
I I
I I
I i
0 ~
0 0.01 0.02 0.03 STRAIN

CORE STRESS- STRAIN CURVES

FIGURE 5.39 Effect of Confinement Ratio for a 12 Bar Unit


96

- - Proposed curve
- - - - t 5Y. Limit of proposed curve
- - - Modified Kent & Park
-----Sheikh & Uzumeri
0 12 Bar unit 15
+ 8 Bar unit _
0 ....
1.8 e Plain unit ---025
-- +2o
~--
+--
13
0 -19+
- --
1. 6 18 --
!; -- 024 ---
,..,..,..,. >""- ~ 3 ....... .,..,.--- -

- 17+ --
--
9+023_-
__.
-----
1. !,

1. 2

1. 0 21

0.8

0.6

0 0.01 0.02 0.03 P5

STRENGTH INCREASE _VERSUS HOOP STEEL VOLUME

Note that Units 1? 2, 6 and 21 were tested at slow speed while the
remainder of experimental results refer to high speed tests.

FIGURE 5,40 : Strength Increase Versus Confinement Ratio


97

5.3.3 Distribution of Longitudinal Steel

Figure 5.41 shows comparitive curves for 8 and 12 bar units with similar
amounts of longitudinal and confining steel. In each case the curve for
a 12 bar unit lies above the curve for the comparable 8 bar unit,
indicating that better confinement is obtained from wider distribution of
the required longitudinal steel.

5.3.4 Ultimate Compression Strain

Examination of Figure 5.41 indicates that the ultimate compressive strain


(at first hoop fracture) is greater in each case in the 8 bar unit than
the 12 bar units. This appears to be because the length over which yield
penetration can spread is greater in the internal diamond of an 8 bar unit
than the internal octagon of a 12 bar unit. This would also be the reason
for first fracture of an inner hoop rather than an outer hoop.

A comparison of the theoretical and experimental ultimate compressive


strains can be seen in Table 5.3. The theoretical ultimate concrete strain
was calculated from Baker (Equation 5.4) and Corley (Equation 5.5):
Baker

Ecu = 0.0015 [ 1 + 150ps + (0.7 - lOps) % J~ 0,01 (5.4)

Corley
b psf h 2
= 0.003 + 0.02 ~ + ( y } (5.5)
L 137.8

where ps = volumetric hoop steel ratio


d = effective depth of section
c = neutral axis depth (at the ultimate moment)
b = width of beam
Z = distance from the critical section to the point
of contraflexure.

The ratios of% and ~were considered to be 1.0 and zero respectively for
the eccentric units, and of course zero for the axially loaded units.
These equations were developed from test results on simply supported beams
(Corley) and have resulted in different criteria to govern what should be
designated as ultimate concrete strain.
98

12 Bar Units
---- 8 Bar Units
1.8 0 Unit Number

1.6

1. 4
- -- -- --
~(J
1.2
~-l f
I . . . . . "'
)(

lr)

tti
~
lr)
1.0 I 'I
Lu
1-- I I
I
~ 0.8
I I
{...)
<:
I I
a I
{...)
0.6 I
0.4

0.2

0.01 0.02 0.03 . STRAIN


CORE. STRESS-STRAIN CURVES

FIGURE 5.41 Confinement Due to Distribution of Longitudinal Steel


99

TABLE 5.3 Comparison of Ultimate Concrete Strains

UNIT pS
Baker Corley Exp. Exp. Exp.
fyh Eq. 5.4 Eq. 5.5 Baker Corley
2 .0182 309 .0056 .0047 .0223 .3.98 4.74
3 II II II
.0047 .0215 3.84 4.57
II
4 II
.0064 .0047 .0743* 11.61 15.81
5 II II II
.0047 .0609* 9.52 12.96
6 .0174 II
.0054 .0045 .0325 6.02 7.22
7 \1 II II
.0045 .0271 5,02 6.02
8 II II
.0062 ,0045 .0649* 10.47 14.42
12 .Q140 309 .0047 .0040 .0167 3.55 4.18
13 .0182 II
.0056 ,0047 .0203 3.63 4.32
14 .0224 296 .0065 .0053 .0289 4.45 5.45
15 .. 0309 II
.0085 ,0074 .0304 3.58 4.11
17 .Q134 309 .0045 .0039 .0214 4.76 5.49
18 .0174 II
,0054 .0045 .0287 5,31 6.38
19 .0213 296 .0063 .0051 ,0359 5.70 7.04
20 .. 0293. II
... 0081 ,0070 .0382 . 4. 72 5,46

22 .0140 309 .0047 ,0040 .0238 5,06 5,95


23 .0182 !I
.0056 .0047 .0287 5.13 6.11
24 .0224 296 .0065 .0053 .0284 4.37 5.36
25 .0309 II
,0085 ,0074 .0323 3.80 4.36
* Peak strain at lst hoop fracture for eccentric tests

The ratios of experiment/theory in Table 5.3 are $imilar to those found by


Gill, Park, and Priestley( 28 ), and Potangaroa, Park and Priestley( 42 ), as
are the strains at first hoop fracture for the eccentric units. These
generally lie between 2 and 3% for the axially loaded units and 6 to 7.5%
for the eccentrically loaded units. In each case there is also an increase
in strain at first hoop fracture with increasing confinement ratio.

It was impossible to establish a spalling strain for the dynamic tests,


however for the slow axial units it was about 0.004 and for the slow
eccentric units it was about 0.005.
100

5.3.5 Strength of Longitudinal Steel

It would seem from Figure 5.40 that the strength of longitudinal steel has
only a minor effect on confinement. Because the longitudinal bars buckle
earlier with Grade 275 steel the confinement of the core is reduced and a
lower peak stress is reached. It was noted that double buckling of these
bars occurred more frequently than with the high yield (Grade 380) steel.
The individual graphs (Figures 5.29 ~ 5.32} show excellent behaviour of the
falling branch and may in some respects be considered better than the high
yield steel. Strain at first hoop fracture is also increased slightly
with the Grade 275 steel.

5.3.6 Eccentricity of Loading

For the eccentric tests no stress-strain curve is possible but the


comparison between theory and experiment in Figures 5.10, 5.12, 5.16 and
5.18 raises some doubts about using results obtained from axially loaded
units for analysis under eccentric loading, especially if large flexural
deformations are involved. The graphs show the theory to be conservative
beyond peak load or moment indicating that a flatter falling branch is
appropriate for members with strain gradients.

Figure 5.5(e),with central lateral displacement plotted against longitud-


inal strain, shows that only small displacements occur at low longitudinal
strains. Displacements become quite significant beyond about 0.004 and
rise quite rapidly, and almost linearly in this region.

The ultimate compressive strain (at first hoop fracture) at 0.060 to 0.075
was much higher than obtained from the axial load tests. However, the
longitudinal strain adopted for the axially loaded units was the a~erage
value. As noted earlier, strain gradients existed in the axially loaded
units (typically up to 20% of average at failure} with hoop fracture
always occurring on the side with maximum compression. The ultimate
compressive strains noted in Tables 5,1 and 5.3 for axially loaded units
are thus conservative.

Although Baker and Corley both make allowances for the presence of strain
gradient when predicting ultimate concrete strain the ratios shown in
Table 5.3 for experiment/theory are still of the same order of magnitude
for eccentrically loaded units as for axially loaded units.
101

CHAPTER SIX

CONCLUSIONS AND FUTURE RESEARCH

6. 0 SUM~1ARY

This chapter draws together the conclusions for the research c~rrted out
and makes recommendations for future research_.

6,1 CONCLUSIONS

Cal Large strength increases were obtP.ined from the effective confine . . .
ment of core concrete and rapid loading rates, Concrete core strengths up
to 186% of cylinder strength were obtained from these tests

(b} The confinement requirements of DZ 3101 (1 J Chapter 17 provided


excellent confinement of the core concrete for the v~rtous axi_a,l load
levels considered.

Cc1 Loading rate i.nfluenced both peak stress and slope of th.e falling
branch of the core stress-strain curye. For the fast loading rate tn
these tests (0.0167/s} peak stress and slope of the falli.ng branch. l,t.lere
increased by about 25%.

(d} An i.ncrease in confinement ratio increased the peak stress attai.nedl


the strain at first hoop fracture? and decreased the slope of the falling
branch. An increase in hoop set spaci.ng tended to reduce the efficiency of
confinement.

(e} The presence of a strain gradient increased the strain at first


hoop fracture quite significantly (2 to 3 times the strain in axial load
tests}. It also resulted in a slower decrease in load and moment with
increasing strain which may be considered a better behaviour than predicted
by analysis using stress-strain curves from axi,ally loaded tests.

(f) Theoretical predictions of ultimate concrete strain based on


accepted equations by Baker and Corley are unrealistic and unduly conserv~
at;ve, Ultimate concrete strain of axially loaded units in these tests
ranged from 1.67% to 3,82% and increased with increasing confinement ratio.
102

A definition of ultimate concrete strain of 11 the strain at first hoop


fracture 11 is felt to be appropriate for columns.

(g) Early buckling of Grade 275 longitudinal steel reduces the


effective confinement of the core concrete, when compared with Grade 380
steel, resulting in a lower peak stress (about 10%), and an increase in
strain at first hoop fracture (also about 10%).

(h) An increase in the number of longitudinal steel bars resulted in


a better confinement of the core concrete for a given total steel area.

6.2 RECOMMENDATIONS FOR FUTURE RESEARCH

(a} There still remains a need for experimental work to be carri.ed out
on full size or near full size test units.

(b) The presence 6f a strain gradient i~ not fully understood and its
influence, while it appears to be beneficial, needs to be examined.

(c} The increase in cylinder strength of confined core concrete has


been established experimentally but a rational theoretical analysis r:1eeds
to be developed, for both circular and rectangular sections if possible,
Such an analysis would need to include the effects evident in these tests
such as loading rate, confinement ratio, presence of a strain gradient
and distribution of longitudinal steel.

(d} All the units tested had very similar concrete cylinder strengths.
Further investigation is necessary to estab l i.sh th.e inf1 uence of concrete
strength on the stress strain relationship for confined concrete.

(e) Five units in this series of tests have been left to be tested at
some later date in order to establish the influence of the age of concrete
at the time of testing. Particular attention should be paid to any
change of sti.ffness or a more brittle behavi~our with age.
103

APPENDIX A

REFERENCES

1. Draft Code of Practice for the Design of Concrete Structures


11

(DZ 3101) Standards Association of New Zealand, 1978.


11 ,

2. SHEIKH, S.A., Effectiveness of Rectangular Ties as Confinement


11

Steel in Reinforced Concrete Columns Ph.D. Thesis, University 11 ,

of Toronto, Canada, 1978.


3. KING, J.W.H., The Effect of Lateral Reinforcement in Reinforced
11

Concrete Columns The Structural Engineer, Vol. 24, No. 7,


11 ,

July 1946, pp.355-388.


4. KING, J. W. H. , Further Notes on Reinforced Concrete Co 1umns
11 11 ,

The Structural Engineer, Vol. 24, Nov. 1946, pp.609-616.


5. KING, J.W.H., Some Investigations of Effect of Core Size and
11

Steel and Concrete Quality in Short Reinforced Concrete Columns 11 ,

Magazine of Concrete Research, Vol. 2, Jan. 1949.


6. CHAN, W.W.L., The Ultimate Strength and Deformation of Plastic
11

Hinges in Reinforced Concrete Frameworks Magazine of Concrete 11 ,

Research, Vol. 7, No. 21, Nov. 1955, pp. 121-132.


7. BRESLER, B., and GILBERT, P.H., Tie Requirements for Reinforced
11

Concrete Columns ACI Journal, Proceedings Vol. 58, No. 5,


11 ,

November, 1961, pp.555-569.


8. PFISTER, J.F., Influence of Ties on the Behaviour of Reinforced
11

Concrete Columns ACI Journal, Proceedings Vol. 61, No. 5,


11 ,

May 1964, pp.521-536.


9. SZULCYNSKI, T., and SOZEN, M.A., Load-Deformation Characteristics
11

of Reinforced Concrete Prisms with Rectilinear Transverse


Rei nforcement Structural Research Series No. 224, Civil
11 ,

Engineering Studies, University of Illinois, Urbana, Sept. 1961,


54 pp.
10. ROY, H. E. H., and SOZEN, M.A., A Model to Simulate the Responses
11

of Concrete to Multi-Axial Loading Structural Research Series


11 ,

No.268, Civil Engineering Studies, University of Illinois, Urbana,


June 1963.
11. ROY, H.E.H., and SOZEN, M.A., Ductility of Concrete Proceedings
11 11 ,

of the International Symposium on the Flexural Mechanics of


Reinforced Concrete, ASCE- ACI, Miami, November 1964, pp. 2130224.
104

12. BERTERO, V.V., and FELIPPA, C., Discussion of 'Ductility of


11

Concrete', by H.E.H. Roy and M.A. Sozen Proceedings of the 11 ,

International Symposium on the Flexural Mechanics of Reinforced


Concrete, ASCE- ACI, Miami, November 1964, pp. 227-234.
13. STOCKL, S., Discussion of 'Ductility of Concrete~, by H.E.H. Roy
11

and M.A. Sozen Proceedings of the International Symposium on


11 ,

the Flexural Mechanics of Reinforced Concrete, ASCE- ACI, Miami,


November 1964, pp. 225-227.
14. HUDSON, Fred M., Reinforced Concrete Columns: Effects of Lateral
11

Tie Spacing on Ultimate Strength Paper No. 10, Symposium on


11 ,

Reinforced Concrete Columns, ACI Special Publication SP.13, 1966,


pp. 235-244.
15. SOLIMAN, M.T.M., YU, C.W., The Flexural Stress-Strain Relationship
11

of Concrete Confined by Rectangular Transverse Reinforcement 11 ,

Magazine of Concrete Research, Vol. 19, No. 61, Dec. 1967,


pp. 223-238.
16. SHAH, S., and RANGAN, B.V., Effects of Reinforcement on
11

Ductility of Concrete", Journal of the Structural Division,


ASCE, ST6, June 1970, pp.1167-1184.
17. SOMES, Normal F., Compression Tests on Hoop-Reinforced Concrete
11 11 ,

Journal of the Structural Division, Proceedings of the American


Society of Civil Engineers, ST7, July 1970, pp.1495-1509.
18. KENT, D.C., PARK, R., Flexural Members with Confined Concrete
11 11 ,

Journal of the Structural Division, ASCE, Vol. 97, ST7, July


1971, pp.1969-1990.
19. SARGIN, M., Stress-Strain Relationships for Concrete and the
11

Analysis of Structural Concrete Sections Study No. 4, Solid 11 ,

Mechanics Division, University of Waterloo, Waterloo, Canada,


1971, 167 pp.
20. SARGIN, M., GHOSH, S.K., and HANDA, U.K . ., Effects of Lateral 11

Reinforcement Upon the Strength and Deformation Properties of


Concrete Magazine of Concrete Research, Vol. 28, No.75-76,
11 ,

June-Sept., 1971, pp. 99-110.


21. BURDETTE, Edwin G., and HILSDORF, Hubert K., Behaviour of 11

Laterally Reinforced Concrete Columns Journal of the Structural 11 ,

Division, Proceedings of the American Society of Civil Engineers,


ST2, February 1971, pp.587-602.
22. BUNNI, N.G., Rectangular Ties in Reinforced Concrete Columns
11 11 ,

ACI Special Publication SP50-8, 1975, pp.193-210.


23. KAAR, P.H., FIORATO, A.E., CARPENTER, J.E., and CORLEY, W.G.,
Limiting Strains of Concrete Confined by Rectangular Hoops",
11

Tentative Report, Research and Development, Construction


Technology Laboratories, Portland Cement Association, PCA R/D
Ser. 1557, January 1977.
105

24. VALLENAS, J., BERTERO, V.V., and POPOV, E.P., "Concrete Confined
by Rectangular Hoops and Subjected to Axial Loads", Earthquake
Engineering Research Centre, College of Engineering, University
of California, Berkeley, California, Report No. UCB/EERC-77/13,
August 1977, 114 pp.
25. LESLIE, P.O., uouctility of Reinforced Concrete Bridge Piers",
Master of Engineering Report, University of Canterbury, New Zealand,
1974.
26. STURMAN, G.M., SHAH, S.P., and WINTER, G., "Effect of Flexural
Strain Gradients on Microcracking and Stress-Strain Behaviour of
Concrete", ACI Journal, Proceedings Vol. 62, No. 7, July 1965,
pp. 805-822.
27. HOGNESTAD, E., HANSON, N.W., and McHENRY, D., "Concrete Stress
Distribution in Ultimate Strength Design", ACI Journal, Proceedings
Vol. 52, No. 12, Dec. 1955, pp. 455-579.
28. GILL, W.O., PARK, R., and PRIESTLEY, M.J.N., "Ductility of
Rectangular Reinforced Concrete Columns With Axial Load Research
11 ,

Report 79-1, Department of Civil Engineering, University of


Canterbury, February 1979.
29. PARK, R., PRIESTLEY, M.J.N., and GILL, W.O., "Ductility of Square
Confined Reinforced Concrete Columns", Paper submitted to ASCE.
30. PARK, R., and PRIESTLEY, M.J.N., "Code Provisions for Confining
Steel in Potential Plastic Hinge Regions of Columns in Seismic
Design''.

31. Building Code Requirements for Reinforced Concrete (ACI 318-77)"


11

American Concrete Institute, Detroit, 1977, 102 pp.


32. Recommended Latera 1 Force Requirements and Commentary
11 11 ,

Seismology Committee, Structural Engineers Association of


California, 1975, 21 pp plus commentary and appendices.
33. "Tentative Provisions for the Development of Seismic Regulations
for Buildings Applied Technology Council, US Government Printing
11 ,

Office, Washington, 1978, 505 pp.


34. Analysis and Design of Reinforced Concrete Bridge Structuresu,
11

ACI Committee 343, American Concrete Institute, Detroit, 1977,


116 pp.
35. OKUBO, T., and IWASAKI, T., "Summary of Experimenta 1 and
Analytical Seismic Research Recently Performed on Highway Bridges",
Proceedings of a Workshop on Seismic Problems Related to Bridges,
Applied Technology Council, Palo Alto, California, January 1979.
36. "Design Essentials in Earthquake Resistant Buildings",
Architectural Institute of Japan, Tokyo, 1970, 295 pp.
106

37. "Ductility of Bridges with Reinforced Concrete Piers", CDP 810/A,


Ministry of Works and Development, April 1975 (plus December 1977
amendment), 109 pp.

38. ''Highway Design Brief", CDP 701/D, Ministry of Works and


Development, September 1978 (plus November 1978 addendum), 52 pp.

39. BAKER, A.L.L., and AMARKONE, A.M.N., "Inelastic Hyperstatic Frames


Analysis", Proceedings of International Symposium on the Flexural
Mechanics 9f Reinforced Concrete", ASCE- ACI, Miami, Nov. 1964,
pp. 85-142.

40. PRIESTLEY, M.J.N., PARK, R., and NG, K.H., "Seismic Behaviour of
Reinforced Concrete Bridge Piers", Master of Engineering Report,
University of Canterbury, New Zealand, 1976.

41. PARK, R., and PAULAY, T., "Reinforced Concrete Structures",


John Wiley and Sons, New York, 1975, 769 pp.

42. POTANGAROA, R., PARK, R., and PRIESTLEY, M.J.N., ''Ductility of


Spirally Reinforced Concrete Columns Under Seismic Loading'',
Master of Engineering Report, University of Canterbury, New Zea 1and,
1979.
Classn:
THE STRESS STRAIN RELATIONSHIP FOR CONFINED CONCRETE
RECTANGULAR SECTIONS.
Bryan D. Scott
ABSTRACT:
An experimental investigation of square confined concrete
columns subjected to concentric or eccentric axial loads to
failure at slow or dynamic loading rates. Results are
presented for 25 units tested and conclusions made about
loading rate, eccentricity, distribution of longitudinal
steel and confinement ratio.
Department of Civil Engineering, University of Canterbury,
Master of. Engineering Report~ 1980.