Behavior of circular steel tube confined uhpc and uhpfrc columns under axial compression

Keywords: CFSTCs, UHPC, UHPFRC, NSC, HSC, confined concrete, steel tube, steel fibers, STCC columns... 2.2.2.1 The interaction between steel tube and concrete core 38 2.2.2.2 The diffe

CONFINED UHPC AND UHPFRC COLUMNS

UNDER AXIAL COMPRESSION

zur Erlangung des akademischen Grades

Doktor-Ingenieur (Dr.-Ing.)

an der Fakultät Bauingenieurwesen

der Universität Kassel vorgelegt von

An Le Hoang

geboren am 10 Januar 1983 in Thua Thien Hue, Vietnam

Erster Gutachter: Prof Dr.-Ing Ekkehard Fehling (Universität Kassel) Zweiter Gutachter: Prof Dr.-Ing habil Nguyen Viet Tue (TU Graz)

Tag der mündlichen Prüfung: 9 Februar 2018

Kassel, 2018

i

Vorwort der Herausgeber

Konstruktionen aus Ultrahochfestem Beton ermöglichen erhebliche Einsparungen beim Konstruktionsgewicht und damit größere Spannweiten, höhere Gebäude und filigranes Bauen mit Beton Durch Verwendung von Faserbewehrung wird ultrahochfester Beton duktil, besonders in Hinsicht auf Zugbelastung Für Druckbelastung kann jedoch auch durch Umschnürung duktiles Verhalten erzielt werden Für eine Stütze bietet sich insofern besonders die Umschnürung durch ein Stahlrohr an

Während für Verbund-Rohrstützen mit normalfestem und hochfestem Beton hierzu viele Forschungsergebnisse vorliegen und nationale und internationale Normen die praktische Anwendung erleichtern, fehlt dies für ultrahochfesten Beton weitgehend

In seiner Dissertation untersucht Herr Le Hoang An daher das Tragverhalten von Verbundstützen aus Stahlrohren mit ultrahochfestem hochfestem Beton (Concrete Filled Steel Tubes CFSTs) Er konzentriert sich dabei auf den Fall der Lasteinleitung auf den Betonquerschnitt (Steel Tube Confined Concrete CSTC) Dabei wird sowohl der Fall der Füllung mit Ulltrahochleistungsbeton ohne Fasern (Ultra High Performance Concrete UHPC) als auch mit Ultrahochleistungsbeton mit Fasern (UHPFRC) experimentell, analytisch und numerisch untersucht Die eigenen experimentellen Untersuchungen umfassen 18 kurze und mittellange Verbundstützen mit zentrischer axialer Belastung

Die Ergebnisse zeigen, dass sich insbesondere durch eine steife Umschnürung mithilfe eines ausreichend dicken Stahlrohrs die besten Ergebnisse erzielen lassen Das Last-Verformungsverhalten zeigt damit eine ausgeprägte Resttragfähigkeit im Nachbruchbereich Der Autor entwickelt einen baupraktischen Näherungsansatz, der seine Versuchsergebnisse wie auch die experimentellen Ergebnisse anderer Forscher in guter Übereinstimmung abbilden kann Die numerische Modellierung mithilfe der Finite-Elemente-Software ATENA wird dargestellt und mit den Versuchsergebnissen verglichen Dabei zeigt sich der Einfluss des Reibungsbeiwerts zwischen Beton und Stahlzylinder sehr deutlich Das numerische Modell ist in der Lage, die Traglasten sowie das Verhalten im Nachbruchbereich sehr gut abzubilden

Weiterer Forschungsbedarf wird vor allem in Hinblick auf exzentrische Belastung sowie auf längere Stützen gesehen

ii

Preface of the Editors

Structures made of ultra-high-strength concrete enable considerable savings in the design weight and thus larger spans, higher buildings as well as filigree construction with concrete By using fiber reinforcement, ultra high strength concrete becomes ductile, especially in terms of tensile load However, ductile behavior in compression can also be achieved by confinement For confinement, in particular a steel tube can be utilized, thus leading to a composite steel concrete column

While many research results are available for tubular composite columns with normal-strength and high-strength concrete and since national and international standards facilitate practical application, such support is largely absent for ultra-high-strength concrete

In his dissertation, Mr Le Hoang An examines the load-bearing behavior of steel tube composite columns with ultra high-strength concrete (Concrete Filled Steel Tubes CFSTs) He focuses on the case

of load transfer to the concrete section only (Steel Tube Confined Concrete CSTC) Both the case of filling with ultra high performance concrete (UHPC) and ultra-high performance concrete with fibers (UHPFRC) are investigated experimentally, analytically and numerically The own experimental investigations comprise 18 short and medium length composite columns with centric axial loading The results show that the best results can be achieved, in particular, by confining the concrete core

by a sufficiently thick steel tube The load-deformation behavior thus shows a pronounced residual capacity in the post-peak range The author develops an engineering approximation approach that can map his experimental results as well as the experimental results of other researchers in good agreement The numerical modeling using the finite element software ATENA is presented and compared with the test results The influence of the coefficient of friction between concrete and steel cylinder is very clear The numerical model is able to map the load capacities as well as the behavior in the post-peak range very well

Further research is needed, especially with regard to eccentric load and longer supports

Kassel, February 2018 The Editors

iii

Die Forschungsarbeiten dieser Dissertation wurden am Institut für Konstruktiven Ingenieurbau am Fachbereich Bauingenieur- und Umweltingenieurwesen der Universität Kassel durchgeführt Ich danke dem Ministerium für Bildung und Ausbildung Vietnams, dem Institut für Bautechnik der Universität Kassel (IKI, Fakultät für Bau- und Umweltingenieurwesen) und dem Deutschen Akademischen Austauschdienst (DAAD) für finanzielle Unterstützung Mein aufrichtiger Dank gilt dem Institut für Konstruktiven Ingenieurbau des Fachbereichs Bauingenieur- und Umweltingenieurwesen für die Bereitstellung der notwendigen Einrichtungen für meine Experimente

Ich möchte meinem Betreuer, Prof Dr.-Ing Ekkehard Fehling, für seine engagierte Betreuung, unschätzbare wissenschaftliche Begleitung und die großzügige Unterstützung während meiner Doktorarbeit danken

Ich danke speziell Prof Dr.-Ing habil Nguyen Viet Tue von der TU Graz für seine aufschlussreiche Anleitung und umfangreiche Unterstützung Weiterhin möchte ich mich bei Prof Dr rer nat Bernhard Middendorf und Prof Dr.-Ing Anton Matzenmiller für die Teilnahme an der Promotionskommission bedanken

Mein Dank gilt auch allen akademischen und administrativen Mitarbeitern des Instituts für Bautechnik der Universität Kassel Besonderer Dank geht an Dr.-Ing Jenny Thiemicke, MSc Paul Lorenz, Frau Ute Müller, Dr.-Ing Mohammed Ismail, MSc Yuliarti Kusumawardaningsih, MSc Attitou Abu Bakr, Dipl.-Ing Thomas Pfetzing, MSc Yahia Al-Ani für ihre kontinuierliche Unterstützung während des Testens und ihre nützlichen Ratschläge und Diskussionszeit

Ich bedanke mich bei allen Mitarbeitern des Labors für Konstruktiven Ingenieurbau und der AMPA, insbesondere Klaus Trost und Dipl.-Ing Beniamino Faion, Dr.-Ing Thomas Hahn für seine unermüdliche Hilfe bei der Probenvorbereitung und Prüfung; und allgemein den Herren Dr.-Ing Wolfgang Römer, Burkhard Deiß, Frau Anna-Katharina Reim, Herrn Timo Bauch, Herrn Dominik Hübenthal für ihre Hilfe und ihre freundliche Unterstützung beim Herstellen der Probekörper und Durchführung von Referenztests

Darüber hinaus möchte ich allen meinen Kollegen und Freunden in Vietnam und Deutschland für ihre kontinuierliche Ermutigung danken

Last but not least, möchte ich meine Liebe zu meiner Familie übermitteln, meinen Eltern Lê Văn Minh und Hoàng Thị Tuệ Thi; sowie an meine beiden jüngeren Brüder, Le Hoàng Ân, Le Hoàng Nhût; meinen Sohn, Le Hoàng Bảo Lâm; meinen Neffen, Lê Hoàng Minh Đức, in der Stadt Pleiku, Provinz Gia Lai, Vietnam, und Ihnen für ihre fortwährende Geduld und Unterstützung während meiner Zeit

im Ausland, und für ihre Anwesenheit bei mir in guten und schlechten Zeiten danken

Ich möchte meine Doktorarbeit auch meinem Großvater Hoàng Như Hàn widmen, der mehr als jeder andere mein Leben beeinflusst hat

Kassel, Februar 2018 Lê Hoàng An

iv

The research work reported in this dissertation has been carried out at the Faculty of Civil Engineering, Institute of Structural Engineering, University of Kassel, Germany I would like to express thanks to Ministry of Education and Training of Vietnam, Institute of Structural Engineering of University of Kassel (IKI, Faculty of Civil and Environmental Engineering), and German Academic Exchange Service (DAAD) for financial support My sincere appreciation is dedicated to the Institute

of Structural Engineering of University of Kassel (IKI, Faculty of Civil and Environmental Engineering) for providing necessary facilities for my experiments

I would like to express my deepest gratitude to my major supervisor, Prof Dr.-Ing Ekkehard Fehling, for his dedicated supervision, invaluable academic guidance and generous support throughout my PhD study

I would specially thank Prof Dr.-Ing habil Nguyen Viet Tue from TU Graz, for his enlightening guidance and extensive support Furthermore, I would like to thank Prof Dr rer nat Bernhard Middendorf and Prof Dr.-Ing Anton Matzenmiller for being part of the defence commission

My thanks also extend to all academic and administrative staff members of Institute of Structural Engineering - University of Kassel Special thanks go to Dr.-Ing Jenny Thiemicke, MSc Paul Lorenz, Mrs Ute Müller, Dr.-Ing Mohammed Ismail, MSc Yuliarti Kusumawardaningsih, MSc Attitou Abu Bakr, Dipl.-Ing Thomas Pfetzing, MSc Yahia Al-Ani for their continuous supports during testing and their useful advices, and discussion time as well

I gratefully acknowledge the kindly assistance from all the staff members of the Structural Engineering Laboratory and AMPA, in particular Mr Klaus Trost and Dipl.-Ing Beniamino Faion, Dr.- Ing Thomas Hahn for their tireless assistance during preparation of test specimens and testing; and generally to Dr.-Ing Wolfgang Römer, Burkhard Deiß, Mrs Anna-Katharina Reim, Mr Timo Bauch,

Mr Dominik Hübenthal for their help and their kind support in casting specimens and conducting reference tests

Still further, I would like to thank all my colleagues and friends in Vietnam and Germany as well for their continuous encouragement

Last but not least, I would like to convey my love to my family, my parents Lê Văn Minh and Hoàng Thị Tuệ Thi; my two younger brothers, Lê Hoàng Ân, Lê Hoàng Nhật; my son, Lê Hoàng Bảo Lâm; my nephew, Lê Hoàng Minh Đức, in Pleiku City, Gia Lai province, Vietnam, for their continuous patience and support when I am abroad, and for their standing by me and cheering me up through the good and bad times

I would like as well to dedicate my dissertation to my grandfather, Hoàng Như Hàn, who impacted

my life more than anybody else

Kassel, February 2018 Lê Hoàng An

v

Es ist bekannt, dass mit Beton gefüllte Stahlrohrstützen (CFSTCs) im Bereich des Bauingenieurwesens viel Aufmerksamkeit auf sich gezogen und breite Anwendungen gefunden haben Mit der rasanten Entwicklung der Betontechnologie hat sich ultrahochfester Beton (UHPC) aufgrund seiner überlegenen Leistungen, wie der extrem hohen Druckfestigkeit bis zu 200 MPa, nutzbare Zugfestigkeit und sehr hohe Haltbarkeitseigenschaften, zu einer potenziellen Alternative zu normalfestem Beton (NSC) und hochfestem Beton (HSC) entwickelt, so dass Ingenieure die Größe von Bauteilen reduzieren und die Tragfähigkeit erhöhen und neuartige Strukturelemente entwickeln können UHPC weist jedoch eine enorme Drucksprödigkeit auf, die mit der Zunahme der Betonfestigkeit einhergeht, was zu einigen Einschränkungen für seine Anwendungen in der Konstruktion führt Um diesen Nachteil zu überwinden, wurden Forschungsbemühungen auf CFSTCs mit der Verwendung von UHPC gerichtet Das Einschließen von UHPC mit kreisförmigen Stahlrohren erweist sich als eine attraktive Option aufgrund der effizienten Kombination von zwei Materialien, um ein Hochleistungselement zu bilden, das von einer signifikanten Zunahme sowohl der Festigkeit als auch der Duktilität im Vergleich zu unbeschränkten UHPC-Elementen profitiert Über umfangreiche Forschungen über das Verhalten von CFST-Stützen, die NSC oder HSC unter konzentrischer axialer Stauchung verwenden, wurde in der Literatur berichtet Daten zum Verhalten von UHPC-gefüllten Stahlrohrstützen (UHPC-FSTCs) fehlen jedoch noch Darüber hinaus sind bestehende Nachweisnormen für CFSTCs nicht auf UHPC anwendbar Vor diesem Hintergrund zielt diese Dissertation darauf ab, eine kombinierte experimentelle und theoretische Studie über das Verhalten von kreisförmigen UHPC-FSTCs-Stützen unter konzentrischer axialer Belastung nur auf dem Betonkern durchzuführen Dieses Belastungsmuster bezieht sich auf die Form von STCC-Stützen (Steel Tube Confined Concrete), die im Vergleich zu dem Fall einer Belastung des gesamten Abschnitts eine bessere Zunahme sowohl der Duktilität als auch der Festigkeit aufweisen

Diese Dissertation berichtete in der ersten Linie über eine Bewertung des axialen Stauchungsverhaltens von runden STCC-Stützen auf der Grundlage der bisherigen Versuchsergebnisse und eines entwickelten Finite-Elemente-Modells (FEM) in der ATENA-3D-Programm für diese Stützen mit unterschiedlichen Betonstärken Dies ist der Ausgangspunkt für die Hauptversuchungen, die in dieser Dissertation vorgestellt wurden Dann wurde das konstitutive Verhalten von UHPC ohne Faser und mit Stahlfasern (UHPFRC) durch Druckversuche an zylindrischen Proben und direkte Zugversuche an gekerbten Prismen untersucht, wodurch der Einfluss von Stahlfasergehalt und Aspektverhältnis auf die einaxialen Druck- und Zugbeanspruchungen bestimmt wurde

Experimentelle Versuchungen an 18 kreisförmigen Stahlrohr-eingeschlossenen UHPC (CSTC-UHPC) und UHPFRC (CSTC-UHPFRC) Stumpfstützen und Mittelstützenwurden durchgeführt Alle Proben hatten einen Durchmesser von 152.4 mm Versuchparemeter

vi

Mittelstützen Die Versagensarten aller Proben waren hauptsächlich mit dem Querkraftversagen des Betonkerns verbunden Zusätzlich zeigten die Versuchsergebnisse, dass eine Verbesserung der Festigkeit und Duktilität erreicht werden kann, indem nur der Betonkern belastet wird Die inhärente Sprödigkeit von UHPC und UHPFRC bei der Stauchung war durch die zusammengesetzte Wirkung von zwei Materialien deutlich eingeschränkt Bei der Eingliederung von Stahlfasern zeigte sich jedoch sogar bei Verwendung von 2 Vol.-% Stahlfaser keine merkliche Steigerung der Festigkeitsverbesserung Darüber hinaus kann die Verwendung von Stahlfasern die Duktilität der Mittelstützen leicht erhöhen, während die Duktilität der kurzen Stützen nachteilig beeinflusst wird Die Festigkeits- und Duktilitätsverbesserung ist bei dickerer Stahldicke signifikant erhöht Unter den in dieser Studie untersuchten variablen Parametern hat die Stahlrohrdicke den größten Einfluss auf das Verhalten von CSTC-UHPC- und CSTC-UHPFRC-Stützen Daher wäre es sinnvoller, UHPC ohne Stahlfasern in Kombination mit dickeren Stahldicken für diese Art von Stützen zu verwenden

Basiert auf der Analyse der Versuchsergebnisse wurden die Formeln zur Vorhersage der begrenzten Spitzenspannung und ihrer entsprechenden Dehnung und eine vereinfachte Spannungs-Dehnungs-Kurve für CSTC-UHPC- und CSTC-UHPFRC-Stützen vorgeschlagen Die Anwendbarkeit der aktuellen Entwurfsnormen wie EC4 (2004), AISC (2010), AIJ (2001), ACI 318R, CISC (2007) und einige verfügbare analytische Modelle für Beton, die durch Stahlrohre eingeschlossen sind, wurde auch durch den Vergleich der Grenzlasten zwischen Vorhersagen und Versuchsergebnisse validiert Schließlich wurde ein FEM in ATENA-3D Programm entwickelt, um 18 getestete Proben zu simulieren und den Einfluss des Reibungskoeffizienten auf das Druckverhalten modellierter Stützen zu untersuchen Die Vorhersagen von FEM zeigten sehr gute Übereinstimmung mit den Versuchsergebnissen

Schlüsselwörter:

CFSTCs, UHPC, UHPFRC, NSC, HSC, eingeschlossener Beton, Stahlrohr, STCC Stützen

vii

It is well known that concrete filled steel tube columns (CFSTCs) have drawn much research attentions and widespread applications in the field of civil engineering In addition, with the rapid development of concrete technology, ultra high performance concrete (UHPC) has recently become a potential alternative to normal strength concrete (NSC) and high strength concrete (HSC) because of its superior performances such as extremely high compressive strength up to 200 MPa, usable tensile strength and very high durability properties, thus allowing engineers to reduce the size of structural members and to increase the load bearing capacity, and to develop novel structural elements However, UHPC exhibits enormous compressive brittleness accompanying with the increase of concrete strength, leading to some limitations for its applications in construction To overcome this drawback, research effort has been directed towards CFSTCs employing UHPC Confining UHPC with circular steel tubes

is found to be an attractive option due to the efficient combination of two materials to form a high-performance member that benefits from a significant increase in both strength and ductility as compared to unconfined UHPC members An extensive amount of research has been reported in previous literature on the behavior of CFST columns employing NSC or HSC under concentric axial compression However, data on the behavior of UHPC filled steel tube columns (UHPC-FSTCs) is still lacking Moreover, existing design codes for CFSTCs are not applicable to UHPC Set against this background, this dissertation aims at performing

a combined experimental and theoretical study on the behavior of circular UHPC-FSTCs columns under concentric axial loading on only the concrete core This loading pattern refers

to the form of steel tube confined concrete (STCC) columns, which is found to exhibit a better increase in both ductility and strength as compared to the case of loading on the entire section This dissertation reported primarily an assessment of the axially compressive behavior of circular STCC stub columns based on the previous test results and a developed finite element model (FEM) in ATENA-3D software for these columns with various concrete strengths This provides the starting point for the main experimental investigations presented in this dissertation Then the constitutive behavior of UHPC without fiber and with steel fibers (UHPFRC) was investigated by the compression tests on cylindrical specimens and direct tension tests on notched prisms, thereby determining the influence of steel fiber content and aspect ratio on the uniaxial compressive and tensile responses

Experimental tests on 18 circular steel tubes confined UHPC (CSTC-UHPC) and UHPFRC (CSTC-UHPFRC) stub and intermediate columns were conducted All the specimens were 152.4 mm in outer diameter Test variables included: steel tube thicknesses of 5.0 mm, 6.3

mm and 8.8 mm; steel fiber volumes of 0%, 1% and 2%; column lengths of about 600 mm for stub columns and about 1000 mm for intermediate columns The failure modes of all specimens were mainly associated with the shear plane failure of concrete core In addition, the test results indicated that an improvement in the strength and ductility can be obtained by loading on only the concrete core The inherent brittleness of UHPC and UHPFRC in

viii

even with the use of 2% steel fibers by volume Furthermore, the use of steel fibers may slightly increase the ductility of the intermediate columns, while there was an adverse influence on the ductility of the short columns The strength and ductility enhancement were significantly increased with thicker steel thickness It is found that, among the variable parameters investigated in this study, the steel tube thickness had the most tremendous impact

on the behavior of CSTC-UHPC and CSTC-UHPFRC columns Hence, it would have more sense to use UHPC without steel fibers in combination with thicker steel thickness for this type of columns

Based on the analysis of test results, the formulae for predicting the confined peak stress and its corresponding strain and a simplified stress-strain curve for CSTC-UHPC and CSTC-UHPFRC columns were proposed The applicability of current design codes such as EC4 (2004), AISC (2010), AIJ (2001), ACI 318R, CISC (2007) and some available analytical models for concrete confined by steel tube was also validated by comparison of ultimate loads between predictions and test results Finally, a FEM in ATENA-3D was developed to simulate 18 tested specimens and to investigate the effect of friction coefficient on the compressive behavior of modelled columns The predictions of FEM showed very good agreement with the test results

Keywords: CFSTCs, UHPC, UHPFRC, NSC, HSC, confined concrete, steel tube, steel

fibers, STCC columns

2.2.2.1 The interaction between steel tube and concrete core 38

2.2.2.2 The different Poisson’s ratio of steel and concrete 39

2.2.2.3 The failure mechanism in steel tube and concrete core 42

2.2.2.5 Confinement effect in circular CFST columns 47

2.2.2.6 Classification of axial load versus vertical deformation of CFST

columns under concentric compression

49

2.2.3 An overview of the experimental investigation on circular CFST columns under

2.2.3.3 Tests on steel-fibers reinforced concrete filled steel tube columns 61

2.2.3.4 Tests on circular CFST columns employing UHPC or UHSC 64

2.2.4 Overview of some existing design guidelines for CFST columns 67

x

3.2.1 Comparison between circular STCC columns using UHPC and circular

UHPC-FSTCs loaded on entire section

77

3.2.6 Summary of axial stress-strain models for concrete confined by steel tube 84

3.2.7 Recalibration of confined peak stress and strain in four models 87

3.2.7.1 Comparison of confined peak stress and strain in four models with

previous test results

87

3.2.7.2 Modification of equations in the model of Hatzigeorgiou (2008) 89

3.2.7.3 Modification of equations in the model of Johansson (2002) 89

3.2.7.4 Modification of equations in the model of Sakino et al (2004) 90

3.2.7.5 Modification of equations in the model of Han et al (2005) 90

3.3 Proposed model for axial stress-strain curve of confined UHPC by steel tube 91

3.4 Comparison of modified models and proposed model with previous test results 93

4.2.2 Re-evaluation of analytical strength models for confined concrete 103

4.3.1.5 Modelling of interfaces between steel tube and concrete core 114

xi

4.3.2.2 Verification of the axial loads-strain curves of the steel tube 117

4.3.2.3 Verification of the axial load-strain curves and the ultimate load 118

5.3.3 Evaluation of the strain at the peak stress and the elastic modulus 141

5.3.4 Effects of steel fiber on the stress-crack opening relationship in tension 144

5.3.4.1 Observation of the stress-crack opening relationship from the test

results

144

5.3.4.3 The relationship between the fiber efficiency cf0 and the fiber factor K 150

6.4.2 Reference tests for determining the concrete properties 160

6.4.2.4 Direct tensile strength of unnotched prisms 163

7.3.1 Load versus axial strain curves at the mid-height of the steel tube 185

7.3.2 Load versus stress curves at the mid-height of the steel tube 188

7.3.3 Lateral-to-axial strain ratio of the steel tube at the mid-height section 192

7.3.4 Load contribution of the steel tube and the concrete core 194

7.4.3 Vertical displacement between the concrete and the steel tube at LVDTs-V4,

8.2.1.2 Effect of steel fibers on the performance indices 214

8.2.2.2 Effect of steel thickness on the performance indices 221

8.2.3.2 Effect of L c/D on the performance indices 228

8.3 Theoretical calculation of ultimate axial load and strain of circular STCC columns 229

8.4 Proposed formulae for the confined peak stress (f cc) and strain (cc) in circular STCC columns with UHPC and UHPFRC infilled

233

8.6 Simplified stress-strain model for UHPC and UHPFRC confined by steel tube 238

8.7 Verification of the simplified stress-strain model for UHPC and UHPFRC confined

by steel tube columns

241

8.8 Comparison of ultimate axial loads in tests with the predictions from previous analytical models

243

8.9 Comparison of ultimate loads in the tests with the predictions from the design codes 249

8.10 Influence of three different concrete strength ranges including UHPC, HSC, and NSC on the strength enhancement

xiii

9.4.4 Interaction between the concrete core and the steel tube 267

xiv

Capital and Small Latin Letters

A c Cross sectional area of concrete core

A s Cross sectional area of steel tube

D Outer diameter of steel tube

t Thickness of steel tube

L Length of steel tube

L c Length of concrete core

SR Strength ratio

DI Ductility index

SI Strength enhancement index

SE Strain enhancement index

N res or N r Residual load

N u Ultimate load

N y Yield load

N s Load carried by steel tube

N c Load carried by concrete core

N u,pre Predicted ultimate load

HI Hardening index

D/t Outer diameter of steel tube-to-steel thickness ratio

f c Compressive strength of unconfined concrete

f cc Compressive strength of confined concrete

f cu Cubic compressive strength of unconfined concrete

f cc,pre Confined peak stress obtained from prediction

f c,cyl Compressive strength of concrete obtained from 150 mm x 300 mm cylinder test

f spl Splitting tensile strength of concrete

f co Compressive strength of concrete at the onset of plastic flow

w cr Crack width at cf,cr

p yield Confining pressure in yield condition

f bc Equibiaxial compressive strength

f p The proportional limit in the equivalent stress- strain curve of steel tube

c Shear cohesion

t

s

E Tangent modulus in the elastic-plastic range

E c Elastic modulus of concrete

E s (or E y) Elastic modulus of steel

V f Steel fiber volume

l f Length of steel fiber

d f Diameter of steel fiber

K tt Initial elastic shear stiffness

K nn Initial elastic normal stiffness

L t Crack band size in tension (ATENA-3D)

L c Crack band size in compression (ATENA-3D)

N u,test Ultimate load obtained from test

w ct or w ctu Crack opening at the complete release of stress

E 1 Strain hardening modulus

L c /D Length of concrete core-to-outer diameter of steel tube ratio

L/D Length-to-outer diameter of steel tube ratio

xv

p Applied confining pressure

t 500 Flow time of fresh concrete

f t or f ct Matrix tensile strength of concrete

w 0 Crack width at cf0

f rp , f l Lateral confining pressure

e The eccentricity to describe the roundness of the failure surface

s Finite element size of concrete core (in ATENA-3D)

f u Ultimate stress of steel in tension

The units of strength (i.e, f c , f cc) and strain (i.e, c ,cc ) are MPa and ‰, respectively The unit of loads (i.e., N u,

N res) is kN The length and diameter in this study are mm

Greek Letters

ξ Confinement index (factor)

c Strain of concrete at the peak stress f c in compression

pl Plastic strain of concrete

d Limit compressive strain at the zero stress of concrete

ct Strain of concrete at the peak stress f ct in tension

ctu Strain corresponding to zero stress in the descending branch of tension

u Relative opening of contact surface

cc,pre Strain of concrete at the peak confined stress f cc obtained from prediction

μ Friction coefficient between concrete and steel

σ cc Stress of confined concrete at the peak load

σ v Longitudinal stress of steel tube

ν c Poisson’s ratio of concrete

v Longitudinal strain of steel tube

σ r orl or σ lat Lateral confining pressure

χ Reduction factor for the buckling curve

η c Coefficient of confinement for concrete core

K or

r

k Value of confining factor determined empirically from the test data

ν Lateral-to-axial strain ratio of steel tube

85% Axial strain when the load decreased to 85% of the ultimate load

 Equivalent stress of steel

v sp The Poisson’s ratio of the steel in the elastic-plastic range

α Parameter controlling the shape of the descending branch in the stress-strain curve

cf,cr Cracking stress in direct tension test on notched prism

cf0 Fiber efficiency

 Fiber orientation factor

c Strain of concrete at f co

Oliveira Correction factor proposed by De Oliveira et al (2010)

c Strain of concrete at the peak confined stress f cc

σ c Stress of concrete at f c

cc,pre Strain of concrete at the peak confined stress f cc obtained from test

σ cp Compressive strength of confined concrete

σ h Hoop stress of steel tube

ν s Poisson’s ratio of steel

h Hoop strain of steel tube

δ Steel contribution ratio

η a Coefficient of confinement for steel tube

u Strain at the ultimate load N u

y Strain at the yield load N y

xvi

 Equivalent strain of steel

σ hp and σ vp The hoop and longitudinal stresses at the end of the elastic range

β Parameter controlling the slope of the descending branch in the stress-strain curve

Abbreviations

UHPC Ultra high performance concrete

UHPFRC Ultra high performance fiber reinforced concrete

NSC Normal strength concrete

HSC High strength concrete

HPC High performance concrete

UHSC Ultra high strength concrete

Eq (Eqs.) Equation(s)

FE Finite element

RPC Reactive powder concrete

L-S Load versus strain

LVDTs Linear Variable Differential Transducers

SG Strain gauges

SD Standard deviation

CFST Concrete filled steel tube

CFSTCs Concrete filled steel tube columns

STCC Steel tube confined concrete

CSTC-UHPCs Circular steel tube confined UHPC columns

CSTC-UHPFRCs Circular steel tube confined UHPFRC columns

UHSC-FSTCs Ultra high strength concrete filled steel tube columns

UHPC-FSTCs Ultra high performance concrete filled steel tube columns

HSC-FSTCs High strength concrete filled steel tube columns

NSC-FSTCs Normal strength concrete filled steel tube columns

FEM Finite element method

UHSS Ultra high strength steel

COV Coefficient of variation

FEM Finite element model

FE Finite element

SFRC Steel fibers reinforced concrete

1

CHAPTER 1 INTRODUCTION

1.1 Background

Concrete filled steel tube columns (CFSTCs) possess numerous advantages over the conventional reinforced concrete and steel columns in both terms of structural performance and construction sequence, such as high compressive strength and fire resistance, large stiffness and ductility (Johansson et al 2002) Moreover, the use of steel tube as a permanent formwork and thus reduces the construction cost and the amount of labor (Han et al 2014).Consequently, in recent decades, the pace of CFSTCs has been increasing rapidly and their applications are more popular in multi-storey buildings, bridge piers and other supporting

structures (Liew and Xiong 2012)

In practice, depending on type and function of CFSTCs in construction, load can be imposed

on the concrete core or on the entire section It is well known that when concentric load is applied only to the concrete core, the confinement induced by the steel tube is more efficient

in comparison with the case where the steel tube and the concrete core are loaded simultaneously According to Han et al (2005) and Yu et al (2010), for CFSTCs, it is recommended that only the concrete core is to be loaded to obtain better strength and ductility, which are amongst the most important design factors This loading pattern refers to the form of the steel tube confined concrete (STCC) columns as shown in figure 1.1

Figure 1.1: The concept of STCC columns (following Sun 2008)

Within the last two decades, the use of ultra high performance concrete (UHPC) has been gaining increasing popularity in the civil engineering community owing to the superior performance and continued advancement of material technology (Schmidt and Fehling 2005) The use of UHPC in the construction industry has received significant recent research attention because of the excellent performance offered by UHPC over NSC and HSC However, it can be argued that the inherently brittle nature of UHPC accompanying with very

- Addition of the high strength steel fibers

- UHPC core confinement reinforcement by stirrups

- Using steel jacket pipe to provide confinement for UHPC core

- Using high-strength longitudinal reinforcement

Likewise, Popa and Kiss (2013) reviewed some different methods of preventing UHPC column at brittle fracture and pointed out advantages as well as disadvantages of each method The solutions in this study were shown as follows:

- Column wrapped with steel sheet

- Composite column by combining UHPC core and NSC with reinforcement

- Using steel stirrups to produce the confinement for UHPC

- UHPC core wrapped by Fiber Reinforced Polymer (FRP)

Pu et al (2004) stated that, for ultra high strength concrete (UHSC), if brittleness can not be overcome by itself, UHSC becomes a low performance concrete, thus leading to some limitations for its application in the construction From this point of view, these authors suggested that the effective way to improve the fragility performance of UHSC columns is using steel tubes in which UHSC is filled The combination of steel tube and UHSC results in the tremendous bearing capacity and excellent deformation

UHPC with compressive strength up to 200 MPa exhibits enormous compressive brittleness and steel fiber plays little act in improving compressive ductility, so it is necessary to utilize steel tubes to alleviate the brittleness of UHPC (Yan and Feng 2008, Fehling et al 2014) Among the solutions mentioned above, the CFSTCs have emerged as one of the dominant options for impeding the brittleness of UHPC or UHSC under compression In addition, UHPC filled steel tube columns (UHPC-FSTCs) can exploit the best attributes of both steel tube and concrete core, thus allowing engineers to reduce the cross section, to economize on concrete and to achieve small dead load owing to very high compressive strength while still obtaining increased stiffness, strength, energy absorption and ductility Consequently, UHPC

is an attractive alternative to NSC and HSC for CFSTCs (Liew and Xiong 2012) It has been found that although there have been many studies on NSC filled in steel tube columns (NSC-FSTCs) or HSC filled steel tube columns (HSC-FSTCs), there is relatively little research on UHPC-FSTCs On the other hand, current design codes for composite columns may be only applicable for NSC and HSC For instance, Eurocode 4 (EC4 2004) allows the use of concrete cylinder strength up to 60 MPa, while American Institute of Steel Construction (AISC 2010)

is applicable to CFSTCs with normal weight concrete cylinder strength from 21 MPa to 70 MPa Additionally, the Chinese standard (DBJ 2003) and the Japanese code (AIJ 2001) limit the maximum compressive strength of concrete to 80 MPa and to 90 MPa, respectively With

3

some recent advances, Australia standard (AS 2014) for composite bridges and building allows the concrete cylinder strengths up to 100 MPa (Aslani et al 2015) Hence, so as to accelerate the application of UHPC in CFSTCs and in order to extend the current design guidelines, further researches on UHPC-FSTCs are necessary

It is well established that due to the smaller lateral deformation when using concrete with higher compressive strength, the confinement in CFSTCs employing HSC, UHSC and UHPC

is less effective compared to that in NSC-FSTCs (Tue et al 2004a, Yan and Feng 2008, Xiong 2012) More recently, several experimental studies have been conducted to examine the performance of circular UHPC-FSTCs or UHSC-FSTCs, such as Tue et al (2004a, 2004b), Schneider (2006), Yan and Feng (2008), Liew and Xiong (2010, 2012), Xiong (2012), Liew et al (2014), Guler et al (2013), Chu (2014), but more attentions have been paid

to these columns subjected to the concentric loading on the entire section rather than on the concrete core Based on the experimental findings, it is concluded that although UHPC-FSTCs can achieve very high load bearing capacities compared to NSC-FSTCs and HSC-FSTCs, the post-peak behavior is still brittle with the sudden failure after peak load induced

by the natural brittleness of UHPC (Liew and Xiong 2010) Furthermore, it is also recommended that the confinement effect should be neglected in short circular UHPC-FSTCs under loading on the entire section (Yan and Feng 2008, Guler et al 2013, Liew et al 2014) However, the columns loaded only on the concrete core possess sufficient performance on ductility and higher ultimate strength compared to those loaded on the entire section (Tue et

al 2004b, Liew and Xiong 2012, Xiong 2012) This may be attributed to the fact that research attention should be given to STCC columns using UHPC

The aforementioned research gap is a main motivation for the study in this dissertation

1.2 Aims and Objectives

The principal aim of this dissertation is to investigate the compressive behavior of STCC columns using UHPC under axial loading through experimental study, previous test results and finite element model (FEM) In order to fulfil this aim, the following objectives were established:

 To review on the general mechanical characteristic of UHPC as a new innovative material and CFSTCs under compression, thereby an assessment of compressive behavior

of STCC columns using UHPC was carried out

 To study the compressive behavior of circular STCC columns employing UHPC with and without steel fibers under axial loading taking into account the confinement effect

 To investigate the effect of steel fiber content, steel tube thickness and column length on strength and ductility of STCC columns employing UHPC

 To evaluate the suitability of design codes and previous analytical models for STCC columns with UHPC infilled

1.3 Methodology

To achieve the above-mentioned objects, the following methodologies were applied:

 Analyzing the compressive behaviour of STCC columns using UHPC based on the literature review, wherein a comparison of previous test results with design codes and

analytical models is conducted

 Developing suitable material models and finite element model (FEM) in ATENA-3D software to simulate the compressive behavior of circular STCC stub columns for various concrete strength infilled and validating by comparing with previous test results Then a

parametric study for these columns using established FEM is also conducted

 Conducting an experimental test program on 18 circular STCC columns under uniaxial compression These composite columns are classified into three groups depending on steel tube thickness Each group includes 3 short columns of 600 mm length and 3 slender columns of 1000 mm length corresponding to steel tube thicknesses of 8.8, 6.3 and 5.0

mm The outer diameter of circular steel tube for all tested specimens is 152.4 mm In addition, in each group, UHPC with steel fibers at the contents of 0, 1%, and 2% is cast for

1 short and 1 slender columns

 Investigating the axial load versus axial strain of columns, the axial load versus longitudinal strain and lateral strain of steel tube which are taken from the tests

 Studying on the effect of steel fiber content, steel tube thickness and column length on strength and ductility of circular STCC columns using UHPC based on the test results

 Evaluating the suitability of existing design codes and previous analytical models through comparison with the test results

 Developing a FEM using ATENA-3D package to predict the compressive behavior of circular STCC columns using UHPC with and without fibers and verifying the FEM by comparing with the test results

 Proposing a stress-strain curve for UHPC confined by steel tube and a formula for estimating the load bearing capacity of circular STCC columns using UHPC with and without steel fibers

5

1.4 Layout of dissertation

This dissertation reports the outcomes of the study on the compressive behaviour of circular STCC columns with UHPC infilled To obtain the outlined objective, this dissertation is being structured in 10 chapters as follows:

Chapter 1 introduces the statement of studying problem related to UHPC and CFSTCs using

UHPC infilled, discusses the research motivation, and proposes the study objectives and methodology

Chapter 2 presents a significant and thorough review on the mechanical properties of UHPC

and CFSTCs through previous research results

Chapter 3 comprises an assessment of compressive behaviour of circular STCC columns

with UHPC infilled based on the previous test results An investigation of the effect of steel tube thickness, confinement index on the strength and ductility is made, then the appropriateness of design codes and previous analytical models is also evaluated In addition,

a simplified stress-strain model for confined UHPC by circular steel tube is proposed and used for an extensive parametric study

Chapter 4 presents a developed Finite Element Model (FEM) using ATENA-3D software for

simulating the compressive behaviour of circular STCC stub columns under axial loading After validating FEM by comparing with previous test results, a parametric study on the effect

of steel tube thickness, concrete strength, steel yield strength on strength and ductility for circular STCC stub columns is conducted

Chapter 5 investigates the influence of steel fiber content and aspect ratio on the uniaxial

tensile and compressive behavior of UHPC, thereby contributing to get better insight into the

basic properties of UHPC

Chapter 6 provides information about the main experimental program on 18 circular STCC

columns This section explains in detail the procedure of preparation of specimens, material, instrumentation and test setup

Chapter 7 presents the test results and discussion on them, encompassing test observations,

the axial load versus axial strain of columns, the axial load versus longitudinal strain and lateral strain of steel tube, failure modes

Chapter 8 discusses the effect of steel fiber content, steel tube thickness and column length

as well on the strength and ductility In addition, this section also conducts a comparison and

an evaluation of the experimental results with predicted results obtained from existing design codes and previous analytical models Lastly, a stress-strain model for UHPC confined by steel tube and a formula for estimating the load bearing capacity for circular STCC columns using UHPC infilled are proposed A comparative study on the effect of NSC, HSC, UHPC

on the strength enhancement of circular STCC columns are also carried out

6

Chapter 9 describes the FEM performed in ATENA-3D software for simulating the tested

specimens The accuracy of FEM is validated by comparison with test results

Chapter 10 highlights significant findings from the study and recommendations for future

research

7

CHAPTER 2 LITERATURE REVIEW

2.1 Ultra High Performance Concrete (UHPC)

2.1.1 Definition of UHPC

Ultra High Performance Concrete (UHPC) is well-known as one of the latest advance in concrete technology Moreover, UHPC is a new family of concretes which exhibits superior mechanical and durability properties over traditional normal strength concrete (NSC) and high strength concrete (HSC) To date, the definition of UHPC has varied among countries but there are general commonalities associated with it In the recommendation of AFGC/SETRA (Scientific and Technical Committee in France), UHPC is classified as a new generation concrete with a very high compressive strength exceeding 150 MPa, possibly attaining 250 MPa and the use of internal fiber reinforcement ensures the ductile behavior of UHPC under tension However, it was emphasized that depending on the composition and the temperature

of heat – treatment, the compressive strength of UHPC ranges between 200 MPa and up to

800 MPa (Richard and Cheyrezy 1995, Dugat et al.1996), while The US Army Engineering Research and Development Center (ERDC) classifies UHPC as cementitious materials with unconfined compressive strengths ranging from 138 to 276 MPa (Roth et al 2008) According

to The Federal Highway Administration (FHWA) in the United States, UHPC is defined as follows:

‘UHPC is a cementitious composite material composed of an optimized gradation of granular constituents, a water-to-cementitious ratio less than 0.25, and a high percentage of discontinuous internal fiber reinforcement The mechanical properties of UHPC include compressive strength greater than 21.7 ksi (150 MPa) and sustained post- cracking tensile strength greater than 0.72 ksi (5 MPa) UHPC has a discontinuous pore structure that reduces liquid ingress, significantly enhancing durability as compared to conventional and high-performance concrete.’ (Graybeal 2011)

UHPC is also characterized by a very dense matrix with a compressive strength from 150 MPa to 250 MPa (Schmidt and Fehling 2005, Schmidt 2012) In addition to the higher packing density of fines and compressive strength, which is defined for Ultra High Strength Concrete (UHSC), the term ‘Ultra High Performance’ refers to the outstanding durability and low ratio water/cement (Schmidt 2012)

2.1.2 Development of UHPC

In 1970s, new advances in admixture and material technologies allowed to produce a type of concrete with a better dispersion of binder particles in the water and improved compact matrix On the same time, the use of high strength concrete started and gained a great deal of attention from researchers due to plenty of dominant characteristics such as higher compressive strength, flowability, elastic modulus, flexural strength, lower permeability and

8

better durability over normal strength concrete The origin of ultra-high strength cementitious paste with low porosity was investigated by Roy et al (1972) and Yudenfreund et al (1972) The significant enhancement in the compressive strength of the cement paste was achieved through heat-curing (250oC) under pressure (50 MPa) or applying low water-to-cement ratio (0.2) In the beginning of 1980s, Bache (1981) developed a type of concrete named as DSPs

(Densified Systems containing ultra-fine Particles) with a very high compressive strength of

120-270 MPa, ascribed to denser spaces between the cement particles using ultra-fine particles and an extremely low water content with a large quantity of high-range water reducer The development of DSPs was also directed towards a less brittle composite material known as CRC (Compact Reinforced Concrete) in 1986 The product of CRC was based on a high fiber content reinforced with a high amount of conventional steel reinforcement The general principles of DSP material were then taken over by many researchers to set background for UHPC in years later The 1990s emerged a commencement of the development for UHPC with the first concept introduced by De Larrard and Sedran (1994) Somewhat later, a forerunner of UHPC named as RPC (Reactive Powder Concrete) offering a

230 MPa compressive strength was also developed by Richard and Cheyrezy (1995) A maximum grain size of 0.6 mm to reduce the micro-crack size and optimize the particle

packing density and the addition of straight steel fiber volume of 1.5% - 3% (l f /d f = 13mm/0.15mm) to obtain higher ductility were the key technique for this type of RPC Furthermore, a compressive strength up to 800 MPa could be attained for the RPC using steel aggregate under curing process with a temperature of 400oC and a pressure of 50 MPa Following the research by Richard and Cheyrezy (1995), a new patent of RPC was later commercialized under the trade name ‘DUCTAL’ produced by Bouygues-Lafarge-Rhodia in France In 2000s, there were an increasing number of companies and researchers throughout the world developing UHPC One of the most remarkable developments of UHPC is the priority program SPP1182 granted by the German Research Foundation in 2005 and coordinated by the University of Kassel The developments of concrete compressive strengths, in general; and UHPC, in particular, can be seen in figure 2.1

Figure 2.1 The development of: a) concrete compressive strength (taken from presentation by Tue) and

b) concrete technology over 100 years and UHPC (Schmidt et al 2006)

9

2.1.3 Constituent materials of UHPC

2.1.3.1 Principle

To achieve excellent mechanical properties, some basic principles which have been identified

by many previous studies must be adopted to produce UHPC, can be summarized as follows:

- The homogeneity of concrete is enhanced due to the exclusion of coarse aggregate It was suggested that the maximal aggregate size should be less than 600m (Richard and Cheyrezy 1995) Moreover, UHPC can be produced using either fine aggregates concrete with maximum size of 0.5 mm or coarse aggregates concrete with maximum size of 16

mm (Schmidt 2012, Schmidt et al 2015)

- A low water/cement ratio, normally between around 0.2 and 0.3, which leads to very dense and strong structure of hydration products and a minimization of capillary pores (Schmidt and Fehling 2005)

- High packing density of the mixture by the optimization of the fine particles grading (Ma and Schneider 2002)

- Improvement of the properties of the cement matrix by the addition of pozzolanic admixture, such as silica fume (Richard and Cheyrezy 1995; Papadkis 1999) The silicafume content in UHPC is normally 25-30% of cement (Ma and Schneider 2002) Besides, the addition of high quality superplasticizer and large quantities of superfine silica fume and quartz causes a low water/binder ratio to reduce porosity to adjust the workability and improve compressive strength (Wang et al 2012)

- Enhancement of the microstructure can be provided by the post-set heat-treatment (Richard and Cheyrezy 1995) It has been proven that heating during hardening at an optimum temperature 90°C or 200°C can enhance the pozzolanic reaction of micro-silica

and results in a significant reduction of porosity (Cheyrezy et al 1995) Moreover, UHPC

can be cured under pressure condition, which increases the density by reducing entrapped air, removes excess water and accelerates chemical shrinkage as well

- The use of steel fibers or other types of fibers is to control the crack widths and to increase tensile, flexural and shear strength under different types of structural actions and

to provide sufficient ductility for concrete (Schmidt 2012, Fehling et al 2014)

2.1.3.2 UHPC compositions

It has been documented that a typical UHPC usually comprises cement, silicafume, fine or coarse aggregates (e.g., quartz sand, basalt gravel, basalt chippings, and bauxite), crushed quartz, superplasticizer, fibers and water as well The raw material required in producing UHPC must ensure mechanical homogeneity, maximum particle packing density and minimum size of flaws (Schmidt et al 2005) Figure 2.2 shows typical components for producing an UHPC mixture

10

Figure 2.2: The typical components of UHPC (taken from presentation by Tue 2008)

In comparison with NSC and HSC, UHPC is generally distinguished as a fine-grained concrete, with the special selection of aggregates, higher binder content and lower water-cement ratio The typical compositions of concrete materials at different levels of performance is illustrated in figure 2.3 As can be seen in figure 2.3, UHPC exhibits a significant enhancement in compressive strength and a minimum porosity compared to other types of concrete

Figure 2.3: Comparison of strength and porosity between UHPC and NSC, HPC (Lee et al 2013)

Cement: The proportion of cement used in UHPC is relatively higher than that used in NSC

and HSC Ordinary Portland Cement type I (CEM I 42.5R/52.5R) or Portland Cement with high sulphate resistance (CEM I 42.5R HS/52.5R HS) is usually used to produce UHPC Because of high cement content in the mixture, UHPC is more susceptible to high shrinkage, while the needs of water, ettringite formation and heat of hydration are reduced

Sand: Various types of sand are usually used to provide a good paste-aggregate interfacing

bonding due to their high hardness Besides, the fine aggregate like quartz sand plays an important role in reducing the maximum paste thickness The mean particle size of sand is often smaller than 1mm, whereas Wille et al (2011) found that an optimum sand-to-cement ratio was about 1.4 for a quartz sand particle size of 0.8mm

Crushed Quartz: Cement can be partially replaced by crushed quartz powders so as to

achieve the optimum packing of the combination of particles It was pointed out by Ma and Schneider (2002) that up to 30% of cement can be replaced by crushed quartz powders without reduction in compressive strength Moreover, the addition of crushed quartz powders can lead to an improvement in flowability of fresh UHPC

Silica fume: According to DIN EN 13263-1, silica fume is added into UHPC mixture as a

microfiller to fill the voids between coarser particles and to reduce the friction between angular particles Furthermore, the addition of silica fume generates secondary hydrates by

11

puzzolanic reaction from cement hydration Therefore, the presence of silica fume not only contributes to the increase in strength but also to compact concrete density Ma et al (2003) noted that the optimal silica fume content could be up to 25% to obtain the densest mixture and the compressive strength was found to be highest with 30% silica fume used in some tests

Superplasticizer: This admixture is based on polycarboxylate ethers (PCE) The use of

superplasticizers is found to be particularly suitable for UHPC mixture using high quantity of silica fume Besides, superplasticizers provide not only a high early strength but also an improved flowability for UHPC due to the dispersion of fine particles Experience shows that the dosage of superplasticizers ranges between 2-4% of the volume fraction

Water: It is well known that water plays a key role in the cement hydration In addition, the

water/binder (w/b) ratio has a significant influence on the porosity and compressive strength

An optimal w/b ratio of 0.14 for UHPC was noted by Richard and Cheyrezy (1995), while several other researchers indicated that w/b ratio is commonly ranged from 0.18 to 0.25 (Schmidt and Fehling 2005)

Fibers: It was argued that fibers are added into UHPC mixture to increase the ductility and

the post-peak behavior, thereby avoiding brittle failures and reducing the autogenous shrinkage However, the workability is obviously reciprocal with fiber content Richard and Cheyrezy (1995) recommended that 2% volume of steel fibers is the most economic content

to ensure the workability for UHPC mixture

2.1.4 Applications of UHPC

2.1.4.1 General advantages

It has been proven that there are a number of advantages offered by UHPC when compared to conventional concrete and even steel structures It is also argued that the extreme high compressive strength of UHPC allows engineers to reduce the cross section of structure, resulting in the use of less material and a decrease in dead load Figure 2.4 illustrates the comparison of material section weights of beams made by UHPC (produced by Ductal) and other types of material such as steel, conventional reinforced concrete and pre-stressed concrete as well Interestingly, the depth of UHPC beam was found to be a half of reinforced

or pre-stressed concrete, leading to a reduction in the weight of beams by about 70% Consequently, UHPC can be adopted for longer span structures Moreover, the addition of steel fibers in UHPC also allows engineers to eliminate steel reinforcement in some cases, while the flexural and shear capacity of member can be ensured Besides, UHPC is also advantageous in terms of service life and reduced maintenance cost for structures owing to its superior durability and low permeability The risk of deterioration induced by environment in long term can be alleviated by using UHPC According to Racky (2004), for constructing a column, UHPC requires a lower energy and raw materials consumption, indicating a more sustainable building material in comparison with NSC and HSC In brief, the UHPC technology utilized for structural applications is expected to have an extended service life,

12

leading to less environmental impact over time It is also important to state that UHPC can lead to a reduced effort for scaffolding and formwork, transport, preparation of building site

Figure 2.4: Comparison of material section weights among UHPC (Ductal), steel, prestressed

and reinforced concrete (Perry 2006)

To date, UHPC has become a versatile material for a large number of structure in constructions such as in artwork, acoustical panel, precast elements, pedestrian bridges, and some highway bridges

2.1.4.2 General disadvantages

It has been recognized that, in practice, the biggest disadvantages for the applications of UHPC is the initial cost due to its very high cement content and steel fibers addition Currently, UHPC is much more expensive than NSC or HSC, as depicted in figure 2.5 Moreover, because UHPC is relatively new to the industry and there is a lack of established standards for its applications, UHPC has become to be limited in some common structures compared to NSC or HSC Another difficulty of UHPC in practice is that it is more suitable for precast facility rather than onsite utilization

However, most researchers maintained that the total costs of the construction using UHPC can

be lower than that using conventional concrete This is due to the fact that UHPC exploits a significant reduction in member sizes, quantity of material and life-cycle cost as well

Figure 2.5: The comparison of material costs among UHPC, HSC and NSC (taken from presentation

by Tue 2008)

2.1.4.3 Worldwide examples of UHPC applications

UHPC has been considered as major breakthroughs in concrete technology, and it has attracted much attention from academics and engineers throughout the world over past decades Due to some different characteristics compared to conventional concrete such as

13

NSC and HSC, UHPC has been subjected to a significant amount of rigorous material and structural testing before implementing it in many applications To date, there have been a large number of pilot projects using UHPC with the view to industrialize this material as an alternative to NSC and HSC for sustainable construction and to bring a revitalization for structures in building

The first structure made of UHPC was a footbridge named as Sherbrooke bridge, which was constructed in Quebec, Canada, in 1996 It is a prestressed hybrid pedestrian bridge with a total length of 60 m and using six 10-m prefabricated match-cast segments for assembling the main span, as shown in figure 2.6(a) A 30 mm thin UHPFRC slab with a width of 3.3 m was produced and prestressed transversely by sheathed monostrands The remarkable technique involved UHPFRC truss webs confined by stainless steel tubes and prestressed longitudinally

by both internal and external prestressing strands

Likewise, another early use of UHPC was the Sakata-Mirai footbridge in Japan built in 2003 This bridge is a post-tensioned structure with a span length of 50.2 m and assembled by six precast segments, as shown in figure 2.6(b) It is interestingly noted that there was no traditional steel reinforcing bars used for the precast box-girders, while the maximal bearing capacity is reached by only steel fibers in combination with external prestressing The bridge

is extremely light with a weight of about 56 tons, which is approximately one-fifth the weight

of a conventional reinforced concrete structure

The first practical use of UHPC in construction in Germany was also the footbridge over River Nieste near Kassel, built in 2004 An inverted U-Shape deck was produced and posttensioned in the factory with a total length of 12 m, a thickness of 0.1 m and a width of 3

m, as depicted in figure 2.6(c) There were no bar reinforcements for UHPC deck except for the anchoring regions of the tendons and for the integrated transverse girder at the supports

By using precast UHPC, this bridge had very short construction time and reduced construction cost owing to its low weight of about 12 tons Following the success in short span footbridge, a longer span footbridge named as the Gärtnerplatz bridge over River Fulda was constructed in 2007 by using a hybrid steel-UHPC structure This bridge included 6 spans with a total length of 134 m and a longest span of 36 m, as seen in figure 2.6(d) A three-chord truss with varying depth was adopted for the longitudinal beam UHPC and steel tubes were used for top chords and bottom of the truss, respectively To make up the bridge deck, precast UHPC slabs with a width of 5m were produced and pretensioned transversely High strength steel fibers (diameter 0.15 mm, length 17 mm) with 0.9 percent by volume were added in the selected UHPC mix

Other examples of footbridges employing UHPC as major structures constructed in elsewhere can be found in other countries such as the Footbridge of Peace – Seonyu in Korea built in

2002 (arch-bridge with a span of 120 m including six precast-prestressed segments using double-T cross section and a deck with transverse ribs), the 81.2 metre span Mikaneike Footbridge and the 64.5 metre span Hikita Footbridge in Japan completed in 2007, the cable-stayed Footbridge within the scope of the Super Bridge 200 project of the Korean Institute of

14

Construction Technology built in 2009, the bridge for cyclists/pedestrians using UHPC deck panel with reinforcement in Purmerend-Netherlands built in 2008

(a) Sherbrooke Footbridge –Quebec, Canada, 1996 (Blais and Coutoure 1999)

(b) Sakata-Mirai Footbridge, Japan, 2003 (taken from Resplendino and Petitjean 2003)

(c) First UHPC in Germany: bridge over River Nieste near Kassel (Fehling et al 2004b)

(d) Gärtnerplatz Bridge in Kassel, Germany (Fehling et al 2008b)

(e) Pi-girder in the Jakway Park Bridge (Keierleber 2008, 2011)

Figure 2.6: Applications of UHPC in bridges around the world

15

(f) Batu Bridge of 100 m designed by Dura Company, Malaysia (on website http://www.dura.com)

(g) Mars Hill Bridge Wappello County Iowa, U.S.A (Endicott 2007; Bierwagen and Abu-Hawash 2005)

(h) UHPC waffle deck panel with UHPC filled cast joints in Wappello County Iowa (Aaleti et al 2014)

Figure 2.6: Applications of UHPC in bridges around the world (continued)

UHPC has been found to be a greatest potential material for bridge technology, especially for precast girder sections in highway bridges UHPC with superior properties provides a great opportunity for longer girders with shallower depths than conventional precast concrete New bridge girder sections made of UHPC have been developed mainly in North America, Europe and Australia in recent years In addition, the Pi-girder (with a cross-section that resembles the Greek symbol) has emerged as one of the dominant precast girder sections with highly optimized section, thus opening new opportunities for new integral bridge concept with rapid installation The first bridge using UHPC Pi-girders was the Jakway Park Bridge in Iowa, U.S.A completed in 2008 This bridge was assembled using three Pi-girders in the middle-span at a length of 15.5 m and a depth of 0.84 m, as seen in figure 2.6(e) Beside Pi-girders, UHPC has been applied in precast deck panels, box-girders (see figure 2.6f), I-beams and T-beams, inverted or double -T-beams The first French UHPC bridge built in 2001 was Road bridge at Bourg-les-Valence using double-T-beams with a depth of 900 mm, a top chord width of 2200 mm and a thickness of 150 mm Some other highway bridges in France using UHPC can be listed such as Pont de la Chabotte road bridge in 2004 (having a span of 47.4 m, the beam made up 22 UHPC precast segmented box-girders), Pont Pinel road bridge in 2007 (using 17 prestressed UHPC inverted T-beams) The prestressed UHPC I-girders were

16

adopted for the first UHPC highway bridge in Wapello County, Iowa, North America, named

as Mars Hill bridge constructed in 2010 This bridge has a single-span with a 3- I beam of 33.5 m (see figure 2.6g)

UHPC has also been used as a cast-in-field connection for precast elements since 2004 because bridge-deck solution with UHPC joints can provide the strongest link in the bridge system and increase speed of construction The first project using this solution was completed

in 2006 for a highway bridge over the Canadian National Railway at Rainy Lake, Ontario Figure 2.6(h) describes UHPC waffle deck panel using UHPC joints for Mars Hill Bridge in Wapello County, Iowa, U.S.A built in 2010

More UHPC bridges can be found in the state of the art reported by the U.S Federal Highway Administration (FHWA) (Russell and Graybeal 2013) and Fehling et al (2014)

In addition to bridge applications, there are several examples of other structures in building using UHPC such as columns, beams, shells (roof systems), panels, wall panels, façade decoration and protection, stair and balconies, piles, louvers and sunshades The use of UHPC

in building allows engineers to design structures with novel and innovative shapes that was not possible before

A spectacular example of architecture taking advantage of the special benefits of UHPFRC was the toll-gate of the Millau Viaduct in France built in 2004, as shown in figure 2.7(a) An elegant roof in curved shape is a thin shell made of 53 match-cast prefabricated 2 m wide segments connected by an internal longitudinal prestressing

The first application of a precast UHPFRC lattice-style façade in the world system was the State Jean Bouin rugby Stadium in France, which is a Mega-Architectural project completed

in 2012, as seen in figure 2.7(b) The precast triangular elements with 8.2 m long, 2.4 m wide and 25 mm thick were produced for the waterproof roof and façade and jointed together by ribs on the side

UHPC is a suitable material for stairs and balconies, which require a lightweight design The use of fibers in UHPC and the combination between fibers and reinforcing steel can offer not only a very thin component but also slender and elegant shape Figure 2.7(c) shows ultra thin balconies for residential buildings using UHPFRC in Huize het Oosten, The Netherlands The thickness of slabs was 65 mm and the maximum load capacity of slabs was five times the maximum service load

It is well known that UHPC with extremely high compressive strength leads to significant reduction in the dimensions of columns in buildings The Shawnessy LRT Station in Calgary, Canada, was the first thin-shelled precast canopy roof systems using 24 ultra-thin, rectangular canopies supported by UHPC columns, as shown in figure 2.7(d) Similarly, a small canopy with UHPFRC columns as the supports and four UHPC segments with a maximum span of 8

m and a thickness of 25 mm was built in Hilversum, The Netherlands, as seen in figure 2.7(d) Some other examples of UHPC columns can be seen in the tree-like columns of the Musee des Civilisations de l’Europe et de la Mediterranee (MuCEM) in Marseille (2012), or the

17

heavily reinforced column of the construction of the 59-storey Park City Musashi skycraper in Japan, and the column formwork for the basement of the 66-storey Metapolis building in Dongtan, Korea

More variety of UHPC applications in building can be found in Fehling et al (2014)

(a) UHPFRC used for large scale flat slab curved roof - Peage du viaduc de Millau, France (Thibaux 2010)

(b) UHPFRC used for self supporting facade and roof elements-Stade Jean Bouin, France (Ricciotti et al 2010)

(c) UHPFRC used for ultra thin balconies for residential buildings - Huize het Oosten, Netherlands (Van Nalta,

and Baek Hansen 2012)

(d) Structural UHPC columns (Vicenzino et al 2005, Van Herwijnen 2005)

Figure 2.7: Examples of UHPC applications in constructions

2.1.5 Mechanical behaviour characterization of UHPC

2.1.5.1 Time development of compressive strength

The evolution of the compressive strengths of UHPC with time has been studied by several past researchers (e.g., Ma 2010; Kazemi and Lubell 2012; Graybeal 2005, 2007) The strength development is mainly affected by the components of cementitious material, the chemical properties of the admixture, placing and curing methods Ma (2010) investigated the effect of silicafume content on the strength development of UHPC cylinder specimens of 100 mm x

200 mm Figure 2.8(a) shows that the compressive strength of UHPC can reach more than

Shawnessy LRT Station, Canada

Hilversum, The Netherland

18

65% of f c28 (compressive strength at 28 days) after 3 days, then followed by a slow increase in the period of 7 to 14 days with the strength ranging between 80% and 90% of f c28 After 28 days, the compressive strength of UHPC is found to slightly increase As also observed in figure 2.8(a), there is no noticeable change in the development of compressive strength of UHPC when using two different mixtures of 18% and 30% silicafume An equation to estimate the development of compressive strength for UHPC was also suggested by Ma (2010) as follows:

t s

t

5 0

281

where f c (t) is the mean concrete compressive strength at an age of t days, f c , 28d is the mean

concrete compressive strength at 28 days, s is the coefficient depending on the type of cement,

s = 0.2 for rapid hardening high strength cement

Graybeal (2007) presented a very comprehensive test series on UHPFRC cylinders at various ages (between 1 and 57 days) after casting, wherein two curing regimes including steam-treated and un-treated curing after demolding were applied It was reported for un-treated UHPFRC that there is a high rate of strength development during the first week where 60% of

f c , 28d could be attained, then the rate of strength gain decreased until reaching a plateau at the

age of 30 days, as shown in figure 2.8(b) Graybeal (2007) also proposed the relationship between the compressive strength and time after 0.9 days for un-treated UHPFRC:

6 0

3

9.0exp

Figure 2.8: Development of UHPC compressive strength

With considering the high volume of steel fibers, Kazemi and Lubell (2012) performed the evolution of the compressive strengths of UHPC with time up to 180 days for cube specimens (50x50x50mm3) with steel fiber volumes varying from 0% and 5% The compressive strength was also computed through a function of time after casting:

4exp

19

where f cu (t), f cu, 60d are the compressive strength of cube specimen at time t days and at 60

days, respectively is the shape factor calibrated for two different fiber contents: =0.5 for

V f =0% and =0.6 for V f =5%

In order to determine the influence of curing methods on the development of compressive strength, Koh et al (2007) compared the strength development of UHPC cured at 20oC in the water and dry conditions with that at 90oC steam curing, as shown in figure 2.9 It can be seen that, during 7 days after casting, the rate of compressive strength development under 90oC steam curing is more rapid than that under 20oC curing Moreover, the compressive strength

at 7 days under 90oC steam curing is much higher than that under 20oC curing At 91 days, the compressive strength reached 190 MPa for 20oC water curing, which is similar to the strength expected after 90oC steam curing, while only 80% of 190 MPa was attained for 20oC dry curing

Figure 2.9: Development of UHPC compressive strength under different curing methods (Koh et al

2007)

According to Fehling et al (2014) and as also drawn by many researchers, the heat treatment

of UHPC can increase the strength of UHPC and accelerate the strength development These authors suggested that temperatures of approximately 80-90oC should be applied to UHPC specimens in 48 hours to get higher strengths than storing them in water for 20 days at 20oC

2.1.5.2 Compressive behavior of hardened UHPC

UHPC without fibers

The compressive behaviour of UHPC without fibers is distinctly different from conventional concretes such as NSC and HSC (see figure 2.11a) Some critical aspects for the mechanical response of UHPC without fibers are described as follows:

- Extremely high compressive strength with f c > 150 MPa

- High modulus of elasticity, in the range of 45-55 GPa

- Linear elastic portion corresponding to 70-80% of compressive strength for untreated specimens (Ma 2010), as shown in figure 2.10 However, the linear limit of stress-strain curve of specimens under steam treatment could reach 80-90% of compressive strength (Graybeal 2007)

- Poisson’s ratio remaining constant up to 70-80% of compressive strength According to Tue et al (2004a), Poisson’s ratio of UHPC using Basalt split as coarse aggregate is

approximately 0.21 below 0.7 f c and it increases up to 0.32 at the peak compressive

20

strength, as shown in figure 2.11(b) However, Liu et al (2012) and Yan and Feng (2008) pointed out that Poisson’s ratio of UHPC under heat-treatment at 90oC remains around 0.2 and there is no change till the failure stress On the other hand, Poisson’s ratio was reported to be between 0.18 and 0.19 for fine-grained UHPC (Fehling et al 2014)

- Extremely brittle failure with a loud explosion right after peak stress and no descending branch can be observed in the stress-strain curve

- The longitudinal strain typically ranges between 4 - 4.4‰ for fine-grained UHPC, while it can be expected around 3.5‰ for coarse-grained UHPC (Fehling et al 2014)

Figure 2.10: Stress-longitudinal strain diagram for UHPC without fibers in a uniaxial

compression test (Fehling et al 2014 )

UHPC with fibers (UHPFRC)

The addition of fibers to a UHPC matrix has very little influence on the ascending portion of the stress-strain diagram even increasing volume of steel fibers up to 2%, however the presence of fibers leads to not only less brittle compressive behaviour but also a significant improvement in post-peak softening branch The compressive response of UHPFRC, in general, and the post-peak behaviour as shown in figure 2.12, in particular, are affected mainly by:

- Fiber content and fiber orientation, and stiffness of fibers

- Fiber type (straight or hooked fibers, fiber cocktail) and fiber geometry (length l f, diameter

d f , aspect ratio l f /d f)

- Interaction of fibers and matrix (interfacial shear, fiber length to aggregate ratio)

Figure 2.11: (a) Stress-strain relationship of NSC, HSC and UHPC without fibers under uniaxial compression and (b) Poisson’s ratio development over compression stress (Tue et al 2004a)

21

The point of view as mentioned above was also drawn by many researchers (e.g Empelmann

et al 2008; Wille et al 2011; Kazemi and Lubell 2012; Hassan et al 2012; Graybeal 2005; Harish et al 2013; Sugano et al 2007; Wu et al 2016) Conventional test methods used for NSC were found to be unreliable for the measurement of the post-cracking behaviour of UHPFRC in compression Therefore, there is a large number of researchers who investigated

on the effect of fibers on the uniaxial compressive behaviour of UHPFRC by using suitable methods to measure the descending branch in stress-strain curve of UHPFRC

Figure 2.12: Stress-strain curve of UHPFRC (following Fehling et al 2004a)

Hassan et al (2012) described different test methods for the determination of the uniaxial compressive behaviour of UHPC and UHPFRC reinforced with steel fibers up to 2% It was reported that linearity limit of UHPFRC could reach approximately 90%-95% of its compressive strength Yoo et al (2013) stated that the compressive strength and elastic modulus are improved by a higher amount of fibers up to a volume fraction of 3%, while the cylinders with volume of fibers up to 4% have the lowest compressive strength and elastic modulus due to poor fiber dispersion, as illustrated in figure 2.13(a)

Wille et al (2011) reported that the addition of randomly distributed short steel fibers (l f /d f = 13/0.2) at the volume of 2.5 and 6% increased the compressive strength by 5.5 and 13% compared to UHPC, while following the research by Skazlic and Bjegovic (2008) an average improvement of 16% and 22% in compressive strength was observed for UHPFRC using 2%

and 5% short steel fibers (l f /d f = 13/0.15) Kazemi and Lubell (2012) found that the addition

of 2% to 5% steel fibers results in the increases in cube compressive strength of 3.7% to 25% over UHPC; however, these authors suggested that additional vibration should be used for high fiber contents to achieve better mechanical characteristics The effect of short steel fiber

(l f /d f = 13/0.2) volume fraction of 0%, 1%, 2%, 3% and 6% on the compressive strength of both 50 mm and 100 mm cubes under ambient condition (20oC) and high temperature condition (90oC) was investigated at 7 days and 28 days by Le et al (2007), as depicted in figure 2.13(b)

Similarly, El-Helou et al (2014) presented a series of compression tests on UHPC and UHPFRC cylinders in order to capture the full compressive stress-strain curves with taking into account of circumferential strain for the cases of UHPFRC using 2% and 4% steel fibers Figure 2.14 depicts the compression test setup for UHPFRC (2% steel fibers) cylinder specimen of 76 mm x 152 mm and the failure mode after testing The axial displacement was

Figure 2.13: Influence of steel fiber on the compressive strength and elastic modulus (Yoo et al 2013) and on only compressive strength with variation methods of curing (Le et al 2007)

Figure 2.14: Compression test setup for UHPFRC 2% volume of steel fibers and the failure mode

after testing (El-Helou et al 2014)

Figure 2.15: (a) Axial stress-axial strain and (b) axial stress-circumferential strain of UHPC and

UHPFRC using 2% and 4% steel fibers (El-Helou et al 2014)

23

Likewise, Empelmann et al (2004) studied the post-peak behaviour of UHPFRC with the addition of different steel fiber volumes and various UHPFRC mixes These authors stated that the incorporation of steel fibers results in a significant improvement in the post-peak

behaviour, however they recommended that longer steel fibers (l f  17 mm) should be used instead of shorter steel fibers (l f <17 mm) to reduce the amount of steel fibers, while the

compressive performance is still similar Regarding steel fiber types, Wu et al (2016a) investigated the effect of three shaped steel fibers (straight, corrugated and hooked-end) with different fiber contents by volumes of 0, 1, 2 and 3% on the compressive strength of UHPC It was indicated from the figure 2.16 that the compressive strength increases with the increase of fiber content and UHPC using straight steel fibers has higher compressive strength than other type of steel fibers at the same content

Figure 2.16: Effect of steel fiber shape and content on the compressive strength of UHPC (Wu et al

2016a)

In contrast to these above authors, Yan and Feng (2008), and Liu et al (2012) maintained that

it is not feasible to improve the compressive strength and ductility of UHPC by the incorporation of steel fibers even with the volume content exceeding 2%

It has been found that there is a large scatter of the measured descending portion of UHPFRC cylinders in compression reported by numerous researchers owing to many factors induced by fiber orientation, distribution, fiber alignment in formwork, concreting activities, measuring technique (e.g., Empelmann et al 2004; Kazemi and Lubell 2012; Hassan et al 2012; Harish

et al 2013; Sugano et al 2007) Therefore, the descending portion of the stress-strain curve of UHPFRC seems to be hardly possible to predict by proposed mathematical formulae

Beside the influence of steel fibers, the geometry of test specimens is found to be an important factor affecting the compressive strength It is recommended that cylindrical specimens with a slenderness ratio (length/diameter) of 2 should be used in compression test (Fehling et al 2014) Many researchers have investigated the possibility of using smaller size specimens to reduce requirements for grinding process and high capacity of testing machines The use of at least 100 mm diameter cylinders has been proposed to evaluate the compressive strength of UHPC and UHPFRC (AFGC 2002 and JSCE 2008) It has been shown that smaller cylinders and cube specimens tend to perform a slightly higher compressive strength (e.g., Ma 2010; Graybeal 2007; Kazemi and Lubell 2012) Graybeal (2007) tested on the cubes and cylinders with different sizes for UHPFRC using 2% short steel fibers The results revealed that the 75