[A._Mammoli,_C._A._Brebbia,_A._Klemm]_Materials_Ch(BookFi.org).pdf

Home of the Transactions of the Wessex Institute.Papers presented at Materials Characterisation V are archived in the WIT eLibrary

in volume 72 of WIT Transactions on Engineering Sciences (ISSN 1743-3533).The WIT eLibrary provides the international scientific community with immediate

and permanent access to individual papers presented at WIT conferences.http://library.witpress.com

Materials Characterisation V

WITeLibrary

WIT Press publishes leading books in Science and Technology.Visit our website for new and current list of titles.

www.witpress.com

Computational Methods

and Experiments

FIFTH INTERNATIONAL CONFERENCE ONCOMPUTATIONAL METHODS AND EXPERIMENTS IN

MATERIALS CHARACTERISATION

MATERIALS CHARACTERISATION 2011

A.A. MammoliUniversity of New Mexico, USA

C.A. BrebbiaWessex Institute of Technology, UK

A. KlemmGlasgow Caledonian University, UK

INTERNATIONAL SCIENTIFIC ADVISORY COMMITTEE

Organised by

Wessex Institute of Technology, UKUniversity of New Mexico, USA

Sponsored by

WIT Transactions on Engineering Sciences

CONFERENCE CHAIRMEN

G. Badalians GholikandiA. BaytonA. Galybin

H. HuhG. MoriconiP. ProchazkaI. Sanchez

P. Viot

WIT Transactions

Editorial Board

Transactions Editor

Carlos BrebbiaWessex Institute of Technology

Ashurst Lodge, AshurstSouthampton SO40 7AA, UKEmail: [email protected]

B Abersek University of Maribor, SloveniaY N Abousleiman University of Oklahoma,

USAP L Aguilar University of Extremadura, SpainK S Al Jabri Sultan Qaboos University, OmanE Alarcon Universidad Politecnica de Madrid,

SpainA Aldama IMTA, MexicoC Alessandri Universita di Ferrara, ItalyD Almorza Gomar University of Cadiz,

SpainB Alzahabi Kettering University, USAJ A C Ambrosio IDMEC, PortugalA M Amer Cairo University, EgyptS A Anagnostopoulos University of Patras,

GreeceM Andretta Montecatini, ItalyE Angelino A.R.P.A. Lombardia, ItalyH Antes Technische Universitat Braunschweig,

GermanyM A Atherton South Bank University, UKA G Atkins University of Reading, UKD Aubry Ecole Centrale de Paris, FranceH Azegami Toyohashi University of

Technology, JapanA F M Azevedo University of Porto, PortugalJ Baish Bucknell University, USAJ M Baldasano Universitat Politecnica de

Catalunya, SpainJ G Bartzis Institute of Nuclear Technology,

GreeceA Bejan Duke University, USAM P Bekakos Democritus University of

Thrace, Greece

G Belingardi Politecnico di Torino, ItalyR Belmans Katholieke Universiteit Leuven,

BelgiumC D Bertram The University of New South

Wales, AustraliaD E Beskos University of Patras, GreeceS K Bhattacharyya Indian Institute of

Technology, IndiaE Blums Latvian Academy of Sciences, LatviaJ Boarder Cartref Consulting Systems, UKB Bobee Institut National de la Recherche

Scientifique, CanadaH Boileau ESIGEC, FranceJ J Bommer Imperial College London, UKM Bonnet Ecole Polytechnique, FranceC A Borrego University of Aveiro, PortugalA R Bretones University of Granada, SpainJ A Bryant University of Exeter, UKF-G Buchholz Universitat Gesanthochschule

Paderborn, GermanyM B Bush The University of Western

Australia, AustraliaF Butera Politecnico di Milano, ItalyJ Byrne University of Portsmouth, UKW Cantwell Liverpool University, UKD J Cartwright Bucknell University, USAP G Carydis National Technical University of

Athens, GreeceJ J Casares Long Universidad de Santiago de

Compostela, SpainM A Celia Princeton University, USAA Chakrabarti Indian Institute of Science,

IndiaA H-D Cheng University of Mississippi, USA

J Chilton University of Lincoln, UKC-L Chiu University of Pittsburgh, USAH Choi Kangnung National University, KoreaA Cieslak Technical University of Lodz,

PolandS Clement Transport System Centre, AustraliaM W Collins Brunel University, UKJ J Connor Massachusetts Institute of

Technology, USAM C Constantinou State University of New

York at Buffalo, USAD E Cormack University of Toronto, CanadaM Costantino Royal Bank of Scotland, UKD F Cutler Royal Botanic Gardens, UKW Czyczula Krakow University of

Technology, PolandM da Conceicao Cunha University of

Coimbra, PortugalL Dvid Kroly Rbert College, HungaryA Davies University of Hertfordshire, UKM Davis Temple University, USAA B de Almeida Instituto Superior Tecnico,

PortugalE R de Arantes e Oliveira Instituto Superior

Tecnico, PortugalL De Biase University of Milan, ItalyR de Borst Delft University of Technology,

NetherlandsG De Mey University of Ghent, BelgiumA De Montis Universita di Cagliari, ItalyA De Naeyer Universiteit Ghent, BelgiumW P De Wilde Vrije Universiteit Brussel,

BelgiumL Debnath University of Texas-Pan American,

USAN J Dedios Mimbela Universidad de

Cordoba, SpainG Degrande Katholieke Universiteit Leuven,

BelgiumS del Giudice University of Udine, ItalyG Deplano Universita di Cagliari, ItalyI Doltsinis University of Stuttgart, GermanyM Domaszewski Universite de Technologie

de Belfort-Montbeliard, FranceJ Dominguez University of Seville, SpainK Dorow Pacific Northwest National

Laboratory, USAW Dover University College London, UKC Dowlen South Bank University, UK

J P du Plessis University of Stellenbosch,South Africa

R Duffell University of Hertfordshire, UKA Ebel University of Cologne, GermanyE E Edoutos Democritus University of

Thrace, GreeceG K Egan Monash University, AustraliaK M Elawadly Alexandria University, EgyptK-H Elmer Universitat Hannover, GermanyD Elms University of Canterbury, New ZealandM E M El-Sayed Kettering University, USAD M Elsom Oxford Brookes University, UKF Erdogan Lehigh University, USAF P Escrig University of Seville, SpainD J Evans Nottingham Trent University, UKJ W Everett Rowan University, USAM Faghri University of Rhode Island, USAR A Falconer Cardiff University, UKM N Fardis University of Patras, GreeceP Fedelinski Silesian Technical University,

PolandH J S Fernando Arizona State University,

USAS Finger Carnegie Mellon University, USAJ I Frankel University of Tennessee, USAD M Fraser University of Cape Town, South

AfricaM J Fritzler University of Calgary, CanadaU Gabbert Otto-von-Guericke Universitat

Magdeburg, GermanyG Gambolati Universita di Padova, ItalyC J Gantes National Technical University of

Athens, GreeceL Gaul Universitat Stuttgart, GermanyA Genco University of Palermo, ItalyN Georgantzis Universitat Jaume I, SpainP Giudici Universita di Pavia, ItalyF Gomez Universidad Politecnica de Valencia,

SpainR Gomez Martin University of Granada,

SpainD Goulias University of Maryland, USAK G Goulias Pennsylvania State University,

USAF Grandori Politecnico di Milano, ItalyW E Grant Texas A & M University,

USAS Grilli University of Rhode Island, USA

R H J Grimshaw Loughborough University,UK

D Gross Technische Hochschule Darmstadt,Germany

R Grundmann Technische UniversitatDresden, Germany

A Gualtierotti IDHEAP, SwitzerlandR C Gupta National University of Singapore,

SingaporeJ M Hale University of Newcastle, UKK Hameyer Katholieke Universiteit Leuven,

BelgiumC Hanke Danish Technical University,

DenmarkK Hayami University of Toyko, JapanY Hayashi Nagoya University, JapanL Haydock Newage International Limited, UKA H Hendrickx Free University of Brussels,

BelgiumC Herman John Hopkins University, USAS Heslop University of Bristol, UKI Hideaki Nagoya University, JapanD A Hills University of Oxford, UKW F Huebner Southwest Research Institute,

USAJ A C Humphrey Bucknell University, USAM Y Hussaini Florida State University, USAW Hutchinson Edith Cowan University,

AustraliaT H Hyde University of Nottingham, UKM Iguchi Science University of Tokyo, JapanD B Ingham University of Leeds, UKL Int Panis VITO Expertisecentrum IMS,

BelgiumN Ishikawa National Defence Academy, JapanJ Jaafar UiTm, MalaysiaW Jager Technical University of Dresden,

GermanyY Jaluria Rutgers University, USAC M Jefferson University of the West of

England, UKP R Johnston Griffith University, AustraliaD R H Jones University of Cambridge, UKN Jones University of Liverpool, UKD Kaliampakos National Technical

University of Athens, GreeceN Kamiya Nagoya University, JapanD L Karabalis University of Patras, Greece

M Karlsson Linkoping University, SwedenT Katayama Doshisha University, JapanK L Katsifarakis Aristotle University of

Thessaloniki, GreeceJ T Katsikadelis National Technical

University of Athens, GreeceE Kausel Massachusetts Institute of

Technology, USAH Kawashima The University of Tokyo,

JapanB A Kazimee Washington State University,

USAS Kim University of Wisconsin-Madison, USAD Kirkland Nicholas Grimshaw & Partners

Ltd, UKE Kita Nagoya University, JapanA S Kobayashi University of Washington,

USAT Kobayashi University of Tokyo, JapanD Koga Saga University, JapanS Kotake University of Tokyo, JapanA N Kounadis National Technical University

of Athens, GreeceW B Kratzig Ruhr Universitat Bochum,

GermanyT Krauthammer Penn State University, USAC-H Lai University of Greenwich, UKM Langseth Norwegian University of Science

and Technology, NorwayB S Larsen Technical University of Denmark,

DenmarkF Lattarulo Politecnico di Bari, ItalyA Lebedev Moscow State University, RussiaL J Leon University of Montreal, CanadaD Lewis Mississippi State University, USAS lghobashi University of California Irvine,

USAK-C Lin University of New Brunswick,

CanadaA A Liolios Democritus University of Thrace,

GreeceS Lomov Katholieke Universiteit Leuven,

BelgiumJ W S Longhurst University of the West of

England, UKG Loo The University of Auckland, New

ZealandJ Lourenco Universidade do Minho, PortugalJ E Luco University of California at San

Diego, USA

H Lui State Seismological Bureau Harbin,China

C J Lumsden University of Toronto, CanadaL Lundqvist Division of Transport and

Location Analysis, SwedenT Lyons Murdoch University, AustraliaY-W Mai University of Sydney, AustraliaM Majowiecki University of Bologna, ItalyD Malerba Universit degli Studi di Bari, ItalyG Manara University of Pisa, ItalyB N Mandal Indian Statistical Institute, India Mander University of Tartu, EstoniaH A Mang Technische Universitat Wien,

AustriaG D Manolis Aristotle University of

Thessaloniki, GreeceW J Mansur COPPE/UFRJ, BrazilN Marchettini University of Siena, ItalyJ D M Marsh Griffith University, AustraliaJ F Martin-Duque Universidad Complutense,

SpainT Matsui Nagoya University, JapanG Mattrisch DaimlerChrysler AG, GermanyF M Mazzolani University of Naples

Federico II, ItalyK McManis University of New Orleans, USAA C Mendes Universidade de Beira Interior,

PortugalR A Meric Research Institute for Basic

Sciences, TurkeyJ Mikielewicz Polish Academy of Sciences,

PolandN Milic-Frayling Microsoft Research Ltd,

UKR A W Mines University of Liverpool, UKC A Mitchell University of Sydney, AustraliaK Miura Kajima Corporation, JapanA Miyamoto Yamaguchi University, JapanT Miyoshi Kobe University, JapanG Molinari University of Genoa, ItalyT B Moodie University of Alberta, CanadaD B Murray Trinity College Dublin, IrelandG Nakhaeizadeh DaimlerChrysler AG,

GermanyM B Neace Mercer University, USAD Necsulescu University of Ottawa, CanadaF Neumann University of Vienna, AustriaS-I Nishida Saga University, Japan

H Nisitani Kyushu Sangyo University, JapanB Notaros University of Massachusetts, USAP ODonoghue University College Dublin,

IrelandR O ONeill Oak Ridge National Laboratory,

USAM Ohkusu Kyushu University, JapanG Oliveto Universit di Catania, ItalyR Olsen Camp Dresser & McKee Inc., USAE Oate Universitat Politecnica de Catalunya,

SpainK Onishi Ibaraki University, JapanP H Oosthuizen Queens University, CanadaE L Ortiz Imperial College London, UKE Outa Waseda University, JapanA S Papageorgiou Rensselaer Polytechnic

Institute, USAJ Park Seoul National University, KoreaG Passerini Universita delle Marche, ItalyB C Patten University of Georgia, USAG Pelosi University of Florence, ItalyG G Penelis Aristotle University of

Thessaloniki, GreeceW Perrie Bedford Institute of Oceanography,

CanadaR Pietrabissa Politecnico di Milano, ItalyH Pina Instituto Superior Tecnico, PortugalM F Platzer Naval Postgraduate School, USAD Poljak University of Split, CroatiaV Popov Wessex Institute of Technology, UKH Power University of Nottingham, UKD Prandle Proudman Oceanographic

Laboratory, UKM Predeleanu University Paris VI, FranceM R I Purvis University of Portsmouth, UKI S Putra Institute of Technology Bandung,

IndonesiaY A Pykh Russian Academy of Sciences,

RussiaF Rachidi EMC Group, SwitzerlandM Rahman Dalhousie University, CanadaK R Rajagopal Texas A & M University, USAT Rang Tallinn Technical University, EstoniaJ Rao Case Western Reserve University, USAA M Reinhorn State University of New York

at Buffalo, USAA D Rey McGill University, Canada

D N Riahi University of Illinois at Urbana-Champaign, USA

B Ribas Spanish National Centre forEnvironmental Health, Spain

K Richter Graz University of Technology,Austria

S Rinaldi Politecnico di Milano, ItalyF Robuste Universitat Politecnica de

Catalunya, SpainJ Roddick Flinders University, AustraliaA C Rodrigues Universidade Nova de Lisboa,

PortugalF Rodrigues Poly Institute of Porto, PortugalC W Roeder University of Washington, USAJ M Roesset Texas A & M University, USAW Roetzel Universitaet der Bundeswehr

Hamburg, GermanyV Roje University of Split, CroatiaR Rosset Laboratoire dAerologie, FranceJ L Rubio Centro de Investigaciones sobre

Desertificacion, SpainT J Rudolphi Iowa State University, USAS Russenchuck Magnet Group, SwitzerlandH Ryssel Fraunhofer Institut Integrierte

Schaltungen, GermanyS G Saad American University in Cairo, EgyptM Saiidi University of Nevada-Reno, USAR San Jose Technical University of Madrid,

SpainF J Sanchez-Sesma Instituto Mexicano del

Petroleo, MexicoB Sarler Nova Gorica Polytechnic, SloveniaS A Savidis Technische Universitat Berlin,

GermanyA Savini Universita de Pavia, ItalyG Schmid Ruhr-Universitat Bochum, GermanyR Schmidt RWTH Aachen, GermanyB Scholtes Universitaet of Kassel, GermanyW Schreiber University of Alabama, USAA P S Selvadurai McGill University, CanadaJ J Sendra University of Seville, SpainJ J Sharp Memorial University of

Newfoundland, CanadaQ Shen Massachusetts Institute of Technology,

USAX Shixiong Fudan University, ChinaG C Sih Lehigh University, USAL C Simoes University of Coimbra, Portugal

A C Singhal Arizona State University, USAP Skerget University of Maribor, SloveniaJ Sladek Slovak Academy of Sciences,

SlovakiaV Sladek Slovak Academy of Sciences,

SlovakiaA C M Sousa University of New Brunswick,

CanadaH Sozer Illinois Institute of Technology, USAD B Spalding CHAM, UKP D Spanos Rice University, USAT Speck Albert-Ludwigs-Universitaet Freiburg,

GermanyC C Spyrakos National Technical University

of Athens, GreeceI V Stangeeva St Petersburg University,

RussiaJ Stasiek Technical University of Gdansk,

PolandG E Swaters University of Alberta, CanadaS Syngellakis University of Southampton, UKJ Szmyd University of Mining and Metallurgy,

PolandS T Tadano Hokkaido University, JapanH Takemiya Okayama University, JapanI Takewaki Kyoto University, JapanC-L Tan Carleton University, CanadaE Taniguchi Kyoto University, JapanS Tanimura Aichi University of Technology,

JapanJ L Tassoulas University of Texas at Austin,

USAM A P Taylor University of South Australia,

AustraliaA Terranova Politecnico di Milano, ItalyA G Tijhuis Technische Universiteit

Eindhoven, NetherlandsT Tirabassi Institute FISBAT-CNR, ItalyS Tkachenko Otto-von-Guericke-University,

GermanyN Tosaka Nihon University, JapanT Tran-Cong University of Southern

Queensland, AustraliaR Tremblay Ecole Polytechnique, CanadaI Tsukrov University of New Hampshire, USAR Turra CINECA Interuniversity Computing

Centre, ItalyS G Tushinski Moscow State University,

Russia

J-L Uso Universitat Jaume I, SpainE Van den Bulck Katholieke Universiteit

Leuven, BelgiumD Van den Poel Ghent University, BelgiumR van der Heijden Radboud University,

NetherlandsR van Duin Delft University of Technology,

NetherlandsP Vas University of Aberdeen, UKR Verhoeven Ghent University, BelgiumA Viguri Universitat Jaume I, SpainY Villacampa Esteve Universidad de

Alicante, SpainF F V Vincent University of Bath, UKS Walker Imperial College, UKG Walters University of Exeter, UKB Weiss University of Vienna, AustriaH Westphal University of Magdeburg,

GermanyJ R Whiteman Brunel University, UK

Z-Y Yan Peking University, ChinaS Yanniotis Agricultural University of Athens,

GreeceA Yeh University of Hong Kong, ChinaJ Yoon Old Dominion University, USAK Yoshizato Hiroshima University, JapanT X Yu Hong Kong University of Science &

Technology, Hong KongM Zador Technical University of Budapest,

HungaryK Zakrzewski Politechnika Lodzka, PolandM Zamir University of Western Ontario,

CanadaR Zarnic University of Ljubljana, SloveniaG Zharkova Institute of Theoretical and

Applied Mechanics, RussiaN Zhong Maebashi Institute of Technology,

JapanH G Zimmermann Siemens AG, Germany

Materials Characterisation V

Editors

A.A. MammoliUniversity of New Mexico, USA

C.A. BrebbiaWessex Institute of Technology, UK

A. KlemmGlasgow Caledonian University, UK

Computational Methods and Experiments

Editors:A.A. MammoliUniversity of New Mexico, USAC.A. BrebbiaWessex Institute of Technology, UKA. KlemmGlasgow Caledonian University, UK

Published by

WIT PressAshurst Lodge, Ashurst, Southampton, SO40 7AA, UKTel: 44 (0) 238 029 3223; Fax: 44 (0) 238 029 2853E-Mail: [email protected]://www.witpress.com

For USA, Canada and Mexico

Computational Mechanics Inc25 Bridge Street, Billerica, MA 01821, USATel: 978 667 5841; Fax: 978 667 7582E-Mail: [email protected]://www.witpress.com

British Library Cataloguing-in-Publication Data

A Catalogue record for this book is availablefrom the British Library

ISBN: 978-1-84564-538-0ISSN: 1746-4471 (print)ISSN: 1743-3533 (on-line)

The texts of the papers in this volume were set individually by the authors or under theirsupervision. Only minor corrections to the text may have been carried out by thepublisher.

No responsibility is assumed by the Publisher, the Editors and Authors for any injury and/or damage to persons or property as a matter of products liability, negligence orotherwise, or from any use or operation of any methods, products, instructions or ideascontained in the material herein.

WIT Press 2011

Printed in Great Britain by Martins the Printers.

All rights reserved. No part of this publication may be reproduced, stored in a retrievalsystem, or transmitted in any form or by any means, electronic, mechanical, photocopying,recording, or otherwise, without the prior written permission of the Publisher.

Preface

The increasing demands for high-quality products from both industry and consumersare the driving force for the rapid developments in materials science and engineering.In principle materials science involves relating the desired properties and relativeperformance of a material to its microstructural features through characterization.The major determinants of the structure of a material and hence its properties areits constituent chemical elements and the way a material has been processed intoits final form.

Over the years, a variety of experimental techniques have been developed forcharacterizing the physical and chemical properties of materials. Unfortunatelydue to a number of simplifying assumptions and limitations on the use of individualmethods, it is not often possible to describe in a qualitative, reliable way themicrostructural features of many materials. Triangulation of different experimentalmethods as well as computer simulations may become essential to achieve athorough, comprehensive analysis. Simulations can contribute to the understandingof the phenomena and to provide a good basis for the development of durablematerials and components which can withstand ambient and extreme environmentalconditions.

The way forward in material characterisation is to develop new experimentaltechniques or apply existing methodologies adopted from other related disciplines.A very wide range of materials, starting with metals through polymers,semiconductors to composites, necessitates a whole spectrum of experimentaltechniques and numerical models, which are specific for material types. Some ofthese well established methodologies could potentially find applications in newfields. In this context a multidisciplinary approach in material characterisation andthe exchange of original ideas is indispensible.

The aim of the International Conference on Computational Methods andExperiments in Materials Characterisation held in Kos, Greece, was therefore tofacilitate such interdisciplinary interactions within the research community. The

resulting conference book has been arranged in several chapters addressing variousexperimental and numerical methods. The wide range of topics covers mechanicalcharacterisation and testing, corrosion problems and thermal analysis as well asrecycled materials, nano-composites and energy materials.

The editors would like to express their gratitude to all authors without whoseinvolvement this book could not have been produced. We wish to aknowledge thevaluable input of the members of the Scientific Advisory Committee in attractingand selecting many high quality contributions. We trust that this book will presentsome innovative ideas and will facilitate further developments in materials science.

The Editors,Kos, Greece 2011

Contents

Section 1: Micro and nano characterisation of cementitious materials (Special session organised by A. J. Klemm) Application of positron annihilation lifetime spectroscopy to nano-characterisation of polymer-modified mortars P. Guagliardo, A. J. Klemm, S. N. Samarin & J. F. Williams ............................. 3 Multi-technique investigation of calcium hydroxide crystals at the concrete surface E. Gueit, E. Darque-Ceretti, P. Tintillier & M. Horgnies ................................. 15 Characterization of the influence of the casting mould on the surface properties of concrete and on the adhesion of a protective coating M. Horgnies, P. Willieme, O. Gabet, S. Lombard & M. Dykman...................... 27 Section 2: Nano-materials HRTEM techniques applied to nanocrystal modeling: towards an atom-by-atom description D. G. Stroppa, L. A. Montoro, E. R. Leite & A. J. Ramirez ............................... 41 Ca(OH)2 nanoparticle characterization: microscopic investigation of their application on natural stones V. Daniele & G. Taglieri ................................................................................... 55 Nanocarbon composite materials with optical response on radioactive waste M. Vantsyan, G. Popova, E. Karpuzova, M. Bobrov, O. Plaksin & E. Dabek ........................................................................................................ 67

Section 3: Corrosion problems Evaluation of the fretting corrosion mechanisms on the head-cone interface of hip prostheses I. Caminha, C. R. M. Roesler, H. Keide, C. Barbosa, I. Abud & J. L. Nascimento ............................................................................................ 77 Improving corrosion performance by surface patterning M. Bigdeli Karimi, V. Stoilov & D. O. Northwood............................................ 85 Material characterisation to understand various modes of corrosion failures in large military vehicles of historical importance A. Saeed, Z. Khan, N. Garland & R. Smith........................................................ 95 Section 4: Computational models and experiments A multi-factor interaction model (MFIM) for damage initiation and progression C. C. Chamis.................................................................................................... 109 Analytical solution of a two-dimensional elastostatic problem of functionally graded materials via the Airy stress function H. Sakurai........................................................................................................ 119 Moment curvature analysis of concrete flexural members confined with CFRP grids A. Michael & P. Christou ................................................................................ 131 Application of effective media theory in the characterization of the hygrothermal performance of masonry Z. Pavlk, E. Vejmelkov, L. Fiala, M. Pavlkov & R. ern......................... 143 3D FIB reconstruction and characterisation of a SOFC electrode S. Chupin, N. Vivet, D. Rochais & E. Bruneton............................................... 155 Modelling of load transfer between porous matrix and short fibres in ceramic matrix composites J. G. P. Silva, D. Hotza, R. Janssen & H. A. Al-Qureshi................................. 165 Modeling aspects concerning the axial behavior of RC columns H. O. Koksal, T. Turgay, C. Karako & S. Ayenk.......................................... 175

Section 5: Innovative experiments Surface characterization of eucalyptus and ash wood veneers by XPS, TOF-SIMS, optic profilometry and contact angle measurements G. Vzquez, R. Ros, M. S. Freire, G. Antorrena & J. Gonzlez-lvarez........ 187 Interface resistances in heat and moisture transport: semi-scale experimental analysis Z. Pavlk, J. Mihulka, J. umr, M. Pavlkov & R. ern ............................. 199 Section 6: Mechanical characterisation and testing Tension/compression test of auto-body steel sheets with the variation of the pre-strain and the strain rate G. H. Bae & H. Huh ........................................................................................ 213 Definition of averaged elastic-plastic characteristics of sandwich panel structures I. I. Zakirov, V. N. Paimushin & I. M. Zakirov ................................................ 227 Hot deformation and mechanical properties of P/M Al special M. Tercelj, P. Cvahte, I. Perus & G. Kugler ................................................... 239 Coarsening kinetics of the bimodal distribution in DS GTD111TM superalloy V. S. K. G. Kelekanjeri, S. K. Sondhi, T. Vishwanath, F. Mastromatteo & B. Dasan ...................................................................................................... 251 Effect of the elastomer stiffness and coupling agents on rheological properties of magnetorheological elastomers A. Boczkowska & S. F. Awietjan...................................................................... 263 Optimization of magnetoelastic properties of pure nickel by means of heat treatments A. L. Morales, A. J. Nieto, J. M. Chicharro, P. Pintado, G. P. Rodrguez & G. Herranz................................................................................................... 275 Nanomechanical structure-property relations of dynamically loaded reactive powder concrete P. G. Allison, R. D. Moser, M. Q. Chandler, T. S. Rushing, B. A. Williams & T. K. Cummins ............................................................................................. 287 Dynamic strength of concrete under multiaxial compressive loading Y. P. Song & H. L. Wang ................................................................................. 299

Modelling and simulation of the rutting resistance of bituminous mixes: experimental and stochastic approaches A. E. Ouni, A. Dony & J. Colin........................................................................ 307 Laboratory tests on the cleanliness of soil materials used as subgrades in pavement structures A. Athanasopoulou & G. Kollaros................................................................... 315 Use of additives to improve the engineering properties of swelling soils in Thrace, Northern Greece A. Athanasopoulou & G. Kollaros................................................................... 327 Characteristics of a bolted joint with a shape memory alloy stud N. Ould-Brahim, A.-H. Bouzid & V. Brailovski............................................... 339 Section 7: Thermal analysis Experimental validation of a thermal model of adhesively bonded scarf repairs for CFRP composite materials incorporating cure kinetics C. C. N. Bestley, S. G. R. Brown & S. M. Alston ............................................. 351 Computational and experimental characterization of building envelopes based on autoclaved aerated concrete V. Ko, J. Vborn & R. ern....................................................................... 363 Section 8: Recycled materials Quantitative description of the morphology of polyurethane nanocomposites for medical applications J. Ryszkowska & B. Waniewski ...................................................................... 377 Description methods of the properties of composites from oxybiodegradable foil waste and wood J. Ryszkowska & K. Saasiska........................................................................ 387 The effect of slag composition on recycling of OFHC through the ESCM process S. Ketabchi, F. K. Ahadi, K. Hanaee & S. H. Alhoseini .................................. 397 Author Index .................................................................................................. 407

Section 1 Micro and nano

characterisation of cementitious materials

(Special session organised by A. J. Klemm)

This page intentionally left blank

Application of positron annihilation lifetime spectroscopy to nano-characterisation of polymer-modified mortars

P. Guagliardo1, A. J. Klemm2, S. N. Samarin1 & J. F. Williams1 1ARC Centre of Excellence for Antimatter-Matter Studies, School of Physics, University of Western Australia, Australia 2School of Built and Natural Environment, Glasgow Caledonian University, UK

Abstract

Positron annihilation lifetime spectroscopy (PALS) has been applied to study the microstructural features of immature cement mortars. Two types of cement mortars containing superabsorbent polymers (SAPs) were studied, in addition to Ordinary Portland Cement (OPC). The ortho-positronium lifetimes for all samples were in the range of 1.70-1.73 ns, values that are close to that of free water (1.7 ns) and hence suggest the presence of water-filled pores. Periodic lifetime measurements showed that the intensity of this component decreased slightly over a period of four weeks, indicating water loss associated with the curing process, evaporation or a combination of the two. Keywords: cement mortar, superabsorbent polymer, positron annihilation lifetime spectroscopy, positronium, porosity, hydration, curing.

1 Introduction

Cement is a material of immense practical importance, but in spite of its almost ubiquitous use, its microstructural characterisation still proves to be problematic. The complexity of the problem is enhanced when the composition of cement is modified through the use of auxiliary agents. Ordinary Portland cement is a combination of gypsum (CS*H2) and clinker. Gypsum acts to prevent rapid setting of the cement paste while clinker is the hydraulic binder. The major constituents of clinker are tricalcium silicate (C3S), dicalcium silicate (C2S),

Materials Characterisation V 3

www.witpress.com, ISSN 1743-3533 (on-line) WIT Transactions on Engineering Sciences, Vol 72, 2011 WIT Press

doi:10.2495/MC110011

tricalcium aluminate and tetracalcium alumina-ferrite (C4AF); C = CaO, S = SiO2, A = Al2O3, F = Fe2O3, S* = SO3 and H = H2O (abbreviations commonly used in industrial nomenclature) [1, 2]. The presence of water among the active compounds of cement leads to a rearrangement of their structure and the initiation of the hardening process. The initial hardening is caused by the hydration of C3S, which forms a gel-like silicate and calcium hydrate phase referred to as C-S-H gel. These particles ultimately crystallise and bind together the particles of sand or stone into a hard mass. Other hydrates of the complexes described above are also formed, although the reaction rates may differ considerably. The final product is a hardened cement paste which is a mixture of unreacted cement particles, hydration products and pores. Two types of pores are usually distinguished - capillary pores, which comprise the water- or air-filled spaces between the hydrates, with sizes in the range of 10-1000 nm, and smaller pores referred to as gel pores, which are contained in the amorphous hydrate phase. Gel and capillary pores form a continuous network of pores throughout the material [3, 4]. The mechanical properties of cementitious materials are heavily influenced by porosity, with the volume and size distribution of pores controlling both the strength and durability of the material [5]. A detailed characterisation of the pore structure and the factors that affect it are thus crucial to advancing the design of these materials; however, classical porosimetry methods such as mercury intrusion porosimetry (MIP) and gas adsorption cannot always be relied upon to extract information on pores smaller than a few nanometres [6]. In this regard, positron annihilation lifetime spectroscopy (PALS) has considerable advantages over these classical methods. PALS is sensitive to both open and closed porosity, for example, whereas isolated pores are invisible to the aforementioned techniques. In addition, PALS is sensitive to pores in the size range of 0.3-30 nm (that is, in the size range of gel pores), and both pore sizes and relative concentrations can be measured (refer to the following section for a detailed explanation) [7]. In this work, PALS was used to study the features of the porosity of three cement mortars. Two types of cement mortars containing superabsorbent polymers (SAPs) were studied, in addition to a sample composed only of OPC. SAPs are cross-linked networks of hydrophilic polymers with a high capacity for water uptake - they can absorb and retain up to 500 times their weight in water. This makes them ideal for use in water-absorbing applications such as absorbent medical dressings and controlled release media [8]. During the mixing process the polymers absorb the pore solution immediately after their addition to the mortar, reaching saturation within minutes [9, 10]. They then swell to form spherical cavities filled with water. At later stages of the hydration process the water is released to the concrete matrix and the cavities remain as empty pores. For mortars with a low water-to-cement ratio, it is possible to replace part of the irregular capillary pores with larger spherical pores formed by saturated polymers. The dispersion and size of these pores can be estimated by the material attributes of polymers.

4 Materials Characterisation V


The samples were studied with PALS over a period of 4 weeks with measurements starting when the samples were aged 21 days, in order to elucidate the changes in nano-porosity caused by the addition of SAPs as well as the changes arising from the curing/hydration process. In addition, samples aged at 3 and 6 weeks were studied with MIP to determine changes in their pore distributions and bulk densities. MIP can provide information on larger pores (from 100 nm to 100 m) to which PALS is not sensitive.

2 Positron annihilation background

Positron annihilation techniques have been applied to study nano-porosity in a wide range of materials, such as zeolite, silicates and polymers [7]. In positron annihilation lifetime spectroscopy (PALS), positrons from a 22Na radioisotope are injected into the sample under study and a positron lifetime spectrum is recorded [11]. As a 1.274 MeV gamma-quantum is emitted during the decay of 22Na almost simultaneously with a positron, the lifetime is defined as the time between detection of the 1.274 MeV photon and the subsequent 511 keV photons created from the positrons annihilation with an electron. The measured spectrum is a histogram of the time periods between these two events. Ideally, it is a sum of decaying exponentials of the form

/ (1)

where is the intensity or weighting of the positron state with lifetime . An experimentally obtained spectrum differs from this form however, in that it is convolved with the time response function of the apparatus, usually approximated by a sum of Gaussians. Porosimetry is made possible by exploiting the phenomenon of positronium (Ps) formation, Ps being the hydrogen-like bound state of a positron and an electron [12]. In brief, positrons penetrate the material under investigation and rapidly lose energy, predominately through ionising collisions. Ps formation can then take place as a reaction between the positron and a secondary electron produced during the positrons thermalisation (these secondary electrons are often termed spur electrons [13]). Positronium is formed in two spin states with dramatically different annihilation characteristics. The singlet state (with anti-parallel orientation of spins) is termed para-positronium (p-Ps); it has a vacuum lifetime of 125 ps and decays via the emission of two gamma quanta. The triplet state (with parallel spins of electron and positron) is referred to as ortho-positronium (o-Ps) and has a much longer vacuum lifetime of 142 ns, decaying via three gamma emission. Due to its short lifetime, p-Ps is not significantly perturbed by the material; however, o-Ps, because of its intrinsically longer lifetime, interacts strongly with the material and its pore structure. In the presence of matter, o-Ps can decay into two gammas via a process known as pick-off annihilation - the positron in o-Ps annihilates with an electron of opposite spin in the material via a two-gamma process. As a result of pick-off



annihilation, the o-Ps lifetime in matter is significantly shorter than its vacuum value (and can be as low as ~ 1 ns) [7]. The basis for studying porosity is that o-Ps tends to localise in regions of low density, such as pores. This is because in the bulk of the material the positron and the electron in o-Ps experience repulsive Coulomb and exchange forces respectively. As a result, most of these materials (porous insulators) have a negative work function for positronium [7]. Porosimetry is possible because the annihilation rate (the inverse of the lifetime) in the pore is a function of the pore size smaller pores have a shorter lifetime compared to large pores, where the overlap of the Ps wave function with the pore walls is reduced. A semi-empirical quantum mechanical relation has been developed by Tao [14] and Eldrup et al. [15] to relate the o-Ps lifetimes to pore radii:

1 = 2 1-

RR+R +

12 sin

2RR+R (2)

where is the lifetime, R+R is the pore radius and R is empirically determined to be 0.16-0.17 nm. This model is applicable to sub-nanometre pores of a spherical geometry. Gidley et al. [16, 17] extended this model to include larger pore sizes and varying pore geometries (cylinders, cubes, channels and sheets).

3 Experimental

3.1 Materials and mixes

For the purpose of this research Portland cement (BS EN 197-1 CEM II/B-V 32,5) was mixed at 1:1 ratio with fine sand (the vast majority of particles were distributed below 0.6 mm). Throughout the investigation the total water-to-cement ratio of 0.45 was maintained. The mix compositions are presented in Table 1. The pastes were shaped into cylinders with diameters of about 25 mm and thicknesses in the range of 5-10 mm.

Table 1: Composition of cement mortars.

Sample designation OPC SAP-A SAP-B

Mix code R A B

(Water/Cement)total 0.45 0.45 0.45

(Water/Cement)effective 0.45 0.425 0.438

Sand/Cement 1 1 1

SAP content [%] (by cement weight) 0 0.25 0.25



The SAPs used in this study were cross-linked polymers provided by BASF. SAP-A is a copolymer of acrylamide and acrylic acid and SAP-B is a polymer based on acrylic acid. The products had absorption capacities of 200-250 ml/g in demineralised water, though the absorption in mortar depended on the product and was approximately 10 mL/g for SAP-A and 5 mL/g for SAP-B. Both materials were prepared by grinding and screening to sizes of 63-125 m, but there was also a minor (less than 10%) content of finer particles. The SAP particle shapes were irregular and their sizes after initial absorption were in the range of 135-270 m for SAP-A and 105-210 m for SAP-B.

3.2 Positron lifetime and mercury intrusion measurements

For positron lifetime measurements, approximately 30 Ci of 22NaCl was deposited on 7m Kapton foil and covered with an identical foil. The edges of this foil-sandwich were then sealed, and this source foil was placed between two identical pieces of sample. The positron lifetimes were measured with a fast-fast coincidence system. The gamma-ray detectors consist of a truncated cone (31.8 mm diameter tapering to 12.7 mm with a height of 12.7 mm) BC418 scintillator coupled to a Burle 8850 photomultiplier tube. The time resolution of the system is approximately 220 ps, as determined from analysis of a spectrum of high-purity annealed nickel. The spectra comprise of at least 2 million counts and have been analysed using PAScual version 1.3.0 [18]. Mercury intrusion porosimetry (MIP) has been carried out with the use of a Porosimeter Autopore IV 9500 by Micromeritics, with a pressure range up to 60000 psi for all samples at ages 3 and 6 weeks.

4 Results and discussion

In Table 2 the fitted lifetimes and intensities for all samples are given at four different ages. In each case, the lifetime spectra could be fitted with three discrete components. In porous media a number of annihilation processes are possible because multiple positron and positronium states exist. It is likely that the two shorter lifetime components (1 and 2) in Table 2 are the averages of several different annihilation modes. These could include positron annihilation (of both free and trapped positrons), p-Ps self-annihilation and o-Ps pick-off annihilation in the bulk of the material, and annihilations in the source foil. Nanosecond lifetime components (3) on the other hand are typically associated with o-Ps pick-off annihilation in nano-pores. For all of the cement samples the average 3 over the course of the measurements was in the range of 1.70-1.73 ns. The presence of a nanosecond component is often an indicator of porosity or free volume, and these values would correspond to pores having diameters in the range of 0.51-0.52 nm (using the TE model, equation 2). However, these lifetimes are very close to that observed in pure water, approximately 1.7 ns [19]. Given that cement pastes are known to contain a significant fraction of water, and that the samples were tested at a relatively young age (testing began at 21 days after mixing), it seems likely that these



nanosecond components are associated with annihilation in water contained in capillary pores of the cement and hence cannot be directly related to a pore size. Consolati and Quasso [20, 21], Consolati et al. [22] and Salgueiro et al. [23] observed similar lifetimes in Portland cement pastes and attributed this component to water-filled pores.

Table 2: Lifetimes and intensities; standard deviations are given in brackets.

Age (weeks)

1 (ps) 2 (ps) 3 (ns) I1 (%) I2 (%) I3 (%) 2

SAP-

A 3 218 (9) 410 (10) 1.67 (0.02) 43 (4) 50 (3) 5.9 (0.1) 0.93

4 227 (9) 410 (10) 1.67 (0.03) 46 (5) 48 (5) 5.7 (0.2) 1.04 5 242 (7) 430 (10) 1.73 (0.03) 55 (4) 40 (4) 5.2 (0.2) 0.99 6 233 (8) 420 (10) 1.72 (0.03) 50 (4) 44 (4) 5.2 (0.2) 0.99

SAP-

B

3 219 (5) 409 (7) 1.73 (0.02) 45 (2) 49 (2) 5.0 (0.1) 1.15 4 229 (5) 421 (9) 1.76 (0.03) 49 (3) 46 (3) 4.0 (0.1) 1.08 5 221 (6) 409 (8) 1.71 (0.03) 46 (3) 50 (3) 4.3 (0.1) 1.04 6 237 (6) 420 (10) 1.63 (0.03) 53 (4) 43 (4) 4.2 (0.2) 1.08

OPC

3 224 (6) 402 (8) 1.71 (0.02) 47 (3) 48 (3) 4.6 (0.1) 1.01 4 239 (5) 430 (10 1.67 (0.02) 57 (3) 39 (2) 4.3 (0.1) 1.01 5 238 (5) 420 (10) 1.67 (0.03) 53 (3) 42 (3) 4.2 (0.1) 1.09 6 212 (9) 390 (10) 1.73 (0.03) 43 (3) 53 (4) 4.3 (0.1) 1.07

In Figure 1, o-Ps lifetimes (3) and intensities (I3) are plotted as a function of sample age. Lifetime spectra were recorded every 7 days for a 4 week period. Over the course of the measurements, I3 for the cement sample containing SAP-A is somewhat higher than for the cement containing SAP-B and the pure cement sample. If these lifetimes are indicative of water-filled pores then this suggests a higher concentration of pores in the sample containing SAP-A. As the results for SAP-B are closer to those of pure cement, this suggests that the presence of the polymer does not significantly alter the pore concentration. For the SAP-A sample, a gradual decrease in I3 is observed over weeks 3-5, whereas for the SAP-B and pure cement samples there is a decrease after the initial measurement (weeks 3-4), and then I3 is relatively constant. While these changes are small, they may be indicative of various stages of the hydration process, as the decrease in I3 is consistent with the reduction in total porosity that is known to occur as hydration progresses [3]. The gradual decline seen in the SAP-A sample suggests that this polymer releases its water more slowly compared to SAP-B. At the onset of the measurement (age 21 days) the hydration process is likely to be still underway, with both di- and tri-calcium silicates reacting with water to produce a network of calcium silicate hydrates. Water could then be consumed in the hydration process for at least the next two weeks. However, during this time it is difficult to say if the loss of water seen by the decrease in I3 is also associated with evaporation. As the samples lose water one might expect the lifetimes to increase as there would be a transition from filled to empty pores.



The fact that this does not occur suggests that pores collapse with water loss (the volume of capillary pores is known to decrease with hydration [3]). These results are consistent with the work of Consolati and Quasso [20, 21] who observed a systematic decrease in I3 with sample age, but they are in contrast to the work of Myllyl and Karras [24] where the opposite trend was seen. It should be noted however that a direct comparison with these studies is not possible due to differing experimental conditions (namely the composition and age of samples).

3 4 5 63.5

4.0

4.5

5.0

5.5

6.0

6.5

o-Ps

inte

nsity

, I3 (

%)

Sample age (weeks)

3 4 5 61.2

1.4

1.6

1.8

2.0

2.2

o-Ps

life

time,

3 (n

s)

Sample age (weeks) Figure 1: o-Ps intensities (top) and lifetimes (bottom) vs. sample age for

SAP-A (circles), SAP-B (triangles) and Ordinary Portland Cement (squares).

In addition to PALS, MIP measurements were carried out on samples aged 3 and 6 weeks; the resulting pore distributions are shown in Figure 2. It is apparent that the samples contain a wide distribution of pores, with the majority of pores



in the size range of 10-100 nm. The MIP results are in qualitative agreement with the PALS data in the sense that the sample containing SAP-B displays similar behaviour to pure cement - at 6 weeks their pore distributions almost coincide and the total porosity for both samples has decreased with aging. This decrease is also consistent with the reduction in I3 seen in the PALS data.

Figure 2: Pore size distributions determined by MIP at age 3 weeks (top) and 6 weeks (bottom); OPC diamonds, SAP-A squares, SAP-B triangles.

Conversely, the SAP-A sample shows a significant increase in the pore concentration within the 10-100 nm range at 6 weeks. This is in contrast to the decrease in I3 seen in PALS, although PALS may not be sensitive to the porosity in this size range. In spite of the differing trend, both PALS and MIP show that

0

0.05

0.1

0.15

0.2

0.25

1101001000100001000001000000

Log

diff

eren

tial i

ntru

sion

(mL/

g)

Pore size diameter (nm)

0

0.05

0.1

0.15

0.2

0.25

1101001000100001000001000000 Log

diff

eren

tial i

ntru

sion

(mL/

g)

Pore size diameter (nm)



the presence of SAP-A significantly alters the hydration process of the cement mortar. The bulk densities of the samples were also determined by MIP measurements (Table 3). It was found that the density of all samples increased with aging time. Again, the changes in density for the pure cement and SAP-B samples are similar. The change in density for SAP-A is much smaller, which is consistent with the slower rate of change of I3.

Table 3: Bulk densities determined by MIP at 3 and 6 weeks.

Sample Bulk density at 3 weeks (g/mL)

Bulk density at 6 weeks (g/mL)

OPC 1.83 1.89 SAP-A 1.82 1.86 SAP-B 1.84 1.91

In order to gain further insight into the nature of the porosity, the samples were heated at 116C for 4 hours under vacuum conditions (10-5 Torr) and lifetime measurements were repeated. Lifetime spectra were recorded directly after heating and the samples were sealed in a plastic bag containing a desiccant (to prevent water absorption) during the measurement. The resulting lifetimes and intensities are given in Table 4 and the corresponding pore sizes are given in Table 5. At least three types of water can be distinguished in a hydrating cement paste. Chemically-bound water is directly incorporated into the structure of the hydration products, physically-bound water is absorbed on the surfaces of cement particles and reaction products, and there is also free water contained in the capillary and gel pores [25]. The heat treatment will have removed a significant fraction of the latter two sources of water; however, much higher temperatures (above 950C) are required to remove chemically-bound water.

Table 4: Lifetimes and intensities obtained after heat treatment.

Sample OPC SAP-A SAP-B 1 (ps) 263 (2) 214 (9) 219 (4) 2 (ps) 536 (7) 380 (9) 413 (5) 3 (ns) 4.3 (0.1) 1.32 (0.07) 1.62 (0.05) 4 (ns) - 12.6 (0.7) 14.9 (0.6) I1 (%) 75 (2) 36 (8) 44 (2) I2 (%) 24 (1) 60 (7) 52 (2) I3 (%) 1.3 (0.1) 3.3 (0.4) 2.5 (0.1) I4 (%) - 1.0 (0.1) 0.70 (0.01) 2 1.13 1.05 1.01



After heat treatment, 3 for OPC increases to 4.3 ns. This corresponds to a gel pore diameter of 0.86 nm using the TE model (equation 2). However, the intensity of this component decreases by a factor of about 3.3; this could suggest that a significant fraction of the pores collapse upon water removal. There is also a significant redistribution of intensity from I2 to I1 after heat treatment. These are the shorter components associated with multiple positron states in the material. The redistribution of intensity from the longer component (2) to the shorter one (1) indicates that the heat treatment has reduced the number of positron traps in the bulk material. This could be associated with an increase in the bulk density after heating.

Table 5: Pore diameters, d3 and d4, calculated from o-Ps lifetimes, and total o-Ps intensities (Io-Ps).

Sample OPC SAP-A SAP-B

d3 (nm) 0.86 0.42 0.50 d4 (nm) - 1.45 1.56

Io-Ps before (%) 4.3 5.2 4.2 Io-Ps after (%) 1.3 4.3 3.2

For the cement samples containing SAPs an additional low intensity, long lifetime component (4) appears in the spectra after heat treatment (and due to the presence of this long component it was necessary to fix the background to the left of the main coincidence peak to the average value in the analysis). 4 may be associated with pores formed by the polymers which have replaced some of the ordinary pore structure, and its emergence shows that the presence of SAP significantly alters the structure of the cement paste. For the sample containing SAP-A, 3 = 1.32 ns and 4 = 12.6 ns, with respective intensities of 3.3% and 1.0%. These lifetimes would correspond to pores having diameters of 0.42 and 1.45 nm respectively. There is also a redistribution of the weighting of the shorter components (1 and 2); however this is not as pronounced as for ordinary Portland cement. For the sample containing SAP-B, 3 = 1.62 ns and 4 = 14.9 ns, with intensities of 2.5% and 0.7% respectively. These lifetimes correspond to pores having diameters of 0.42 and 1.45 nm respectively. The value of 3 is comparable to that obtained before heat treatment (which was similar to the o-Ps lifetime in water). This could suggest that the heat treatment did not result in the complete removal of free water. For both SAP-A and SAP-B the total o-Ps intensity (I3 + I4) decreases after heat treatment, signalling a reduction in porosity; however, the reduction is larger for the ordinary Portland cement sample.



5 Conclusion

This study has shown that PALS can be used to monitor the hydration process in cement samples by using the o-Ps intensity as an indicator of the concentration of water-filled pores. In addition, PALS has shown that the addition of a SAP to the cement mixture alters the structure and hydration process of cement. The addition of SAP-A caused the hydration process to progress more gradually compared to OPC; however, the addition of SAP-B gave comparable results to that of OPC. Both PALS and MIP showed that the presence of SAP-A has a significant effect on the hydration process. In addition, MIP showed that the concentration of pores in the 10-100 nm range increases with aging for the SAP-A sample. For the samples with SAP added, heat treatment under evacuation (resulting in the removal of free water) gave rise to an additional long lifetime component and a higher overall o-Ps intensity compared to OPC. This shows that the SAP had modified the pore distribution, replacing some of the regular pore distribution with larger pores formed by saturated polymers.

References

[1] Bogue, R. H. The Chemistry of Portland Cement. (Reinhold, 1955). [2] Lea, F. M. The Chemistry of Cement and Concrete. (Arnold, 1970). [3] Neville, A. M. Properties of Concrete. 4th edn, (Longman, 1995). [4] Taylor, H. F. W. Cement Chemistry. (Academic, 1990). [5] Ghosh, S. N. Advances in cement technology (Pergamon, Oxford, 1983). [6] Gregg, S. J. & Sing, K. S. W. Adsorption, Surface Area and Porosity.

(Academic, 1982). [7] Jean, Y. C., Mallon, P. E. & Schrader, D. M. Principles and applications of

positron and positronium chemistry (World Scientific, 2003). [8] Buchholz, F. L. & Graham, A. T. Modern Superabsorbent Polymer

Technology (John Wiley & Sons, 1997). [9] Jensen, O. M. & Hansen, P. F. Water-entrained cement-based materials: I.

Principles and theoretical background. Cement and Concrete Research 31, 647-654, (2001).

[10] Jensen, O. M. & Hansen, P. F. Water-entrained cement-based materials: II. Experimental observations. Cement and Concrete Research 32, 973-978, (2002).

[11] Krause-Rehberg, R. & Leipner, H. S. Positron Annihilation in Semiconductors. Vol. 127 (Springer, 1999).

[12] Charlton, M. & Humberston, J. W. Positron Physics. (Cambridge University Press, 2001).

[13] Mogensen, O. E. Spur reaction model of positronium formation. The Journal of Chemical Physics 60 (1974).

[14] Tao, S. J. Positronium Annihilation in Molecular Substances. The Journal of Chemical Physics 56 (1972).



[15] Eldrup, M., Lightbody, D. & Sherwood, J. N. The temperature dependence of positron lifetimes in solid pivalic acid. Chemical Physics 63, 51-58, (1981).

[16] Gidley, D. W. et al. Positronium annihilation in mesoporous thin films. Physical Review B 60, R5157 (1999).

[17] Gidley, D. W. et al. Determination of pore-size distribution in low-dielectric thin films. Applied Physics Letters 76 (2000).

[18] Pascual-Izarra, C. et al. Advanced fitting algorithms for analysing positron annihilation lifetime spectra. Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment 603, 456-466, (2009).

[19] Patro, A. P. & Sen, P. Anomalous parapositronium lifetime in water at 20 degrees C. Journal of Physics C: Solid State Physics 5, 3273 (1972).

[20] Consolati, G. & Quasso, F. Evolution of porosity in a Portland cement paste studied through positron annihilation lifetime spectroscopy. Radiation Physics and Chemistry 68, 519-521 (2003).

[21] Consolati, G. & Quasso, F. A positron annihilation study on the hydration of cement pastes. Materials Chemistry and Physics 101, 264-268 (2007).

[22] Consolati, G., Dotelli, G. & Quasso, F. Positron lifetime spectroscopy as a probe of nanoporosity of cement-based materials. Radiation Physics and Chemistry 58, 727-731 (2000).

[23] Salgueiro, W., Somoza, A., Cabrera, O. & Consolati, G. Porosity study on free mineral addition cement paste. Cement and Concrete Research 34, 91-97, (2004).

[24] Myllyl, R. & Karras, M. Positron Annihilation Probing for the Hydratation Rate of Cement Paste. Applied Physics 7, 303-306 (1975).

[25] Sen Wang, P. U. et al. 1H nuclear magnetic resonance characterization of Portland cement: molecular diffusion of water studied by spin relaxation and relaxation time-weighted imaging. Journal of Material Science 33, 3065-3071 (1998).



Multi-technique investigation of calcium hydroxide crystals at the concrete surface

E. Gueit1, E. Darque-Ceretti1, P. Tintillier2 & M. Horgnies2 1MINES ParisTech, Center for Material Forming, Sophia Antipolis, France 2Lafarge Centre de Recherche, St Quentin Fallavier, France

Abstract

The durability and aesthetic qualities of high-performance concrete, which makes it particularly suitable for architectural applications, are constantly compromised by environmental aggressions. In this study, an innovative solution was developed to protect the concrete from these aggressions, which consists of growing a mineral coating on the concrete surface. The coating is composed of layered calcium hydroxide crystals, whose nucleation and growth are triggered byvarious non-ionic surfactants (the details of the process will not be presented). This paper describes the procedure used to investigate the structure of the formed crystals. Scanning Electron Microscopy and optical microtopography were used to determine the morphology of the crystals. Image analysis allowed the quantification of their amount, size and shape. The contribution and limits of each technique are discussed. Keywords: concrete, scanning electron microscopy, image analysis, optical microtopography, surface.

1 Introduction

The mechanical and aesthetic durability of concrete is often compromised by the constant environmental aggressions to which the structures are exposed (organic or inorganic particles, algae, micro-organisms, staining from various sources). It is possible to protect concrete from these attacks and increase its durability by applying organic coatings on the hardened surface, but this comes with operational and environmental costs.



doi:10.2495/MC110021

An innovative solution was proposed [1], which consists on covering and protecting high-performance concrete (HPC) with a mineral coating made of calcium hydroxide crystals (CH). The crystals growth happens during the concrete setting and is triggered by the presence of non-ionic surfactants at the concrete/formwork interface. One of the difficulties of this study was to determine the adequate characterization methods to properly investigate the influence of various surfactants on the amount, morphology and size of the crystals. Several techniques exist to observe, measure and study concrete hydrates, but not all of them are directly suitable for surface investigation. An original procedure had to be developed, combining several techniques. The structure and morphology of the crystals were assessed through Scanning Electron Microscopy (SEM) both on the concrete surface and on polished section. The SEM observations were completed by microtopography on the concrete surface. The quantification of the amount and size of the crystals was made by image analysis on binocular images of the concrete surface. The purpose of this paper is to illustrate what each of these techniques can bring to concrete surface studies.

2 Morphology of the crystals

The most common method to observe concrete hydrates morphology is Scanning Electron Microscopy [2]. Most of the time, the observations are made on polished section in secondary electron mode. When higher magnifications or no polishing are required, fresh fractures can also be observed. Surprisingly, it is very rare to find published picture of concrete surface directly observed by SEM. In this study, SEM observations were conducted on both surfaces and cross-sections of the concrete. Small cubes (1 cm x 1 cm x 1 cm) were cut from each concrete sample. For surface observation, the cubes were directly carbon-coated on the adequate face and observed in secondary electron mode. For cross-sections observations, the cubes were impregnated, polished and carbon-coated, and then observed in back-scattered electron mode. A SEM FEG Quanta 400 from FEI Company was used at an accelerating voltage of 15kV and current intensity of 1mA.

Figure 1: SEM observations of a standard high-performance concrete surface

in secondary electron mode (A) and in back-scattered electron mode on a polished section (B).



Figure 1 shows observations of both the surface (A) and the cross-section (B) of a standard high-performance concrete. No specific features are visible on the surface, except for a couple of scratch due to the formwork defects. As for the polished section, it reveals a classical concrete microstructure, with dark grey aggregates and light grey un-hydrated cement grains surrounded by cement paste. Figure 2 shows SEM observation of two cross-sections from two different samples. These observations reveal the presence at the concrete surface of a thin layered structure presenting various orientations and organization, sometimes well-aligned parallel to the surface (A), sometimes arranged in a more chaotic way in the first micrometers of the surface (B). At this level of observation, it is not possible to clearly indentify the nature of this unexpected phase. The limited magnification in back-scattered mode, the thin structure of the hydrates and the surface damages due to the polishing all complicate the interpretation. This is why it is necessary to make complementary observations of the surface itself.

Figure 2: SEM observation in back-scattered electron mode of two polished sections from two different concrete samples from the study.

Figure 3 shows the same samples as figure 2 observed from the surface in secondary electron mode. These micrographs allow a better understanding of the structure of the crystals. All of them share the same layered and flaky structure with different orientations. In all cases, the crystals are composed of thin leaves that grow around the nucleation point. For some crystals, these leaves are strictly parallel to the surface and grow as flower petals around the central point, reaching sometimes a perfect hexagonal shape. For other crystals, the leaves are strictly perpendicular to the surface, forming a very regular spherulitic structure. This very organized structure was not visible on the polished sections, where the leaves appeared randomly implanted in the surface. This is due to the polishing, which damaged the first micrometers of the surface and disturbed the structure. The characteristic hexagonal shape of the crystals, as well as their layered organization, allow identifying them as calcium hydroxide no other concrete phase would present this morphology. This is confirmed by EDS analysis, where mostly calcium is detected is these areas (spectrum not showed).



Figure 3: SEM observation in secondary electron mode of the surface of two different samples from the study.

3 Amount and size of the crystals

3.1 Microtopography

SEM observations were very useful to highlight the presence of an unusual phase and to identify it as CH crystals, but they are not sufficient to get a complete insight on the structure of the crystals. Even though the secondary electrons give information on the surface topography, it is difficult to evaluate from the micrographs how thick the crystals are or how deep they are embedded in the surface. As for the polished section, the crystals layers are too strongly delaminated by the polishing for the observation to be conclusive. This is why the observations were completed with profilometry of the surface. Mechanical and optical profilometry are sometimes used on concrete to assess its roughness and evaluate its behavior regarding adhesion problems [35]. In our case, the measurements were made using optical profilometry only, because a mechanical probe is likely to damage the very fragile CH crystals. The measurements were made on a confocal full-field 3D surface profilometer with a spot of 2 micrometers and a working distance of 4.5 mm. In these conditions, the vertical resolution is 0.01 micrometers and the lateral resolution 0.1 micrometer. Areas of 4 x 4 mm were scanned with a step of 10micrometer. The data were computed using the software MountainsMap. In the examples given below, there were only two steps of data treatments: the maps were straightened to compensate for the horizontality defects, and profiles were extracted. Figure 4 shows an example of profilometry on a sample where the crystals are oriented preferentially parallel to the surface. Two patterns appear on the 2D mapping: regular and slightly curved lines, which are due the formwork texture, and irregular spots, which are the CH crystals. Below the mapping is a roughness profile extracted along the dotted line. The narrow peaks correspond to the lines, the larger one at 1.7 mm corresponds to the CH crystal in the middle. The profile allows measuring the size of the crystal (400 micrometers) and the height between the concrete surface and the top of crystal (5 micrometers).



Figure 4: Example of a mapping of the concrete surface. The profile was

extracted from the mapping along the black dotted line.

Figure 5: Example of a mapping of the concrete surface. The profile was

extracted from the mapping along the black dotted line.



The same measurements can be made of figure 5. The mapping was acquired on the same sample, but in an area were the CH crystals have been pulled out of the surface during the demoulding phase, leaving hollows the size of the crystals. The depth of the hollows gives the height between the surface and the bottom of the crystals. The measurements from figures 4 and 5 give an estimation for the total thickness of the crystals of 20 micrometers. Of course, the operation has to be repeated a large number of times to obtain statistical results. This is only one example taken from one of the samples. It should be noted that profilometry gives no conclusive information for the samples where the crystals are oriented perpendicular to the surface. Figure 6 shows a mapping of such a sample (A): the spherulitic structures appear slightly lighter than the surface, but are difficult to distinguish. A zoom on the mapping (B) reveals that the crystals are indeed apparent, but that the resolution is far from being sufficient to separate the different flakes composing the structure.

Figure 6: Example of a mapping from a sample with perpendicular crystals.

The 3D mapping on the right is a zoom from the 2D mapping on the left.

3.2 Image analysis

The only way to know the amount of CH crystals as well as their dimension and geometrical parameters is to count and measure them one by one. Image analysis software are capable of automatically achieve this procedure on a large number of images, provided that they are correctly settled by the operator. Image analysis has been successfully used in previous studies [2, 6] to assess the amount of calcium hydroxide in concrete as well as the geometrical parameters of the particles (size, shape, etc.). In those cases, image analysis was conducted on SEM micrographs from polished sections. In our case, there is no need to work on sections as the crystals are on the concrete surface. Furthermore, the crystals are so large that SEM micrographs are not suitable for a proper counting. Even at low magnification, there are too few particles in the observation field. On the contrary, images taken under a binocular have the right scale to count and measure the particles.



The CH crystals are shinier than the cement paste, which makes them visible with the naked eye, but this is not sufficient for image analysis software to distinguish them. To overcome this problem, the samples were coloured with a black felt-tip pen prior to analysis. The black ink penetrates the cement paste, strongly colouring it, but is not absorbed by the crystals which then become clearly distinguishable. For each concrete samples, five small areas of the surface (1 cm2) were coloured and photographed under a binocular. The images were treated and analysed using the open-source software ImageJ [7]. The most difficult step in image analysis is to separate the studied particles from the background. In our case, the following procedure was used. Figure 7 shows an example of an image prior to treatment. The brightness and contrast of the images were optimized using the automatic function on the software. The optimization was not based on the whole image, but on the histogram analysis of a small area of one of the particle on the image. This step created a highly contrasted image where particles appeared red and the background black (figure 8).

Figure 7: Example of an image before treatment by the software imageJ.

Figure 8: Example of an image after optimization of the contrast and brightness.



The noise was removed using the remove outliers function of the software, which replaces each pixel by the median value of its neighbours in a given area (figure 9). In this case, a radius of 20 pixels was considered, and the removing was set to occur if the difference between the considered pixel and the median of its neighbour was higher than 10.

Figure 9: Example of an image after noise removal.

The images were then converted to binary and inverted. At this point, the background appeared white and the particles black (figure 10).

Figure 10: Example of an image after conversion to binary.

The holes in the particles were closed using the Fill holes function of the software (figure 11). Finally, the watershed function was used to automatically draw the outlines of the particles (figure 12). This step might be the main cause of errors in the measurements, as watershed segmentation works best for convex objects that do not overlap too much. In some cases, the software was not able to correctly separate clusters of particles, which were then counted as one big particle.



Figure 11: Example of an image after filling the holes.

Figure 12: Example of an image after the software automatically delimited the particles.

Once the image was treated, the particles were counted one by one. The smallest ones (with a surface smaller than 0.1 mm) were not considered as they are more likely residual noise than actual crystals. The particles overlapping with the boarder of the image were not considered either as their size would not be correctly measured. This means than the measured amount of crystals (from 1 to 20 particles per mm depending on the samples) is slightly smaller than reality. The following parameters were measured for each particle:

- The coordinates of the centre of the particles. These coordinates were used to calculate the distance between each crystal and its closest neighbour and to verify that the nucleation appeared randomly and homogeneously on the surface (figure 13).

- The Ferets diameter, which is the longest distance between two opposing points of the particle.

- The fraction area of the surface covered by the crystals. As these crystals are artificially grown to act as a mineral coating, this fraction area must be as high as possible. The results showed that even if the crystals seemed very large, they hardly cover 50% of the surface, except in one promising case where almost 100% of the surface was coated.



- The circularity of the particle, which is defined by 4.area / perimeter. A perfect circle has a circularity of 1, whereas a very elongated particle has a circularity close to 0 (figure 14). This geometrical parameter is very convenient to distinguish the crystals that grow parallel to the surface with regular shapes (high circularity) from the crystals that grow perpendicular to the surface, which have a low circularity.

Figure 13: Example of spatial distribution of the particles on the concrete surface, showing that the nucleation occurs rather homogeneously on the surface.

Figure 14: Two examples of the circularity distribution. The curve marked

with squares corresponds to a sample where the crystals have a low circularity because they grow perpendicular to the surface. The curve marked with dots corresponds to a sample where the crystals have a high circularity because they grow parallel to the surface and adopt very regular shapes.



The results of the image analysis allowed a better understanding of the influence of various surfactants on the nucleation (amount of particles) and growth (size and orientation of the particles) of the CH crystals. Two types of sample could not be properly analyzed though, one because it exhibited excessive roughness that induced too much background noise on the images, and one because the crystals overlapped too much to be clearly separated. In those cases, a more sophisticated procedure has to be developed. In particular, software capable of separating overlapping particles exists, but they were not tested yet in this study.

4 Conclusion

Three techniques scanning electron microscopy, microtopography and image analysis were successfully used to assess the morphology and geometrical parameters of calcium hydroxide crystals at the concrete surface. These techniques are well-known and developed, but not necessarily widely used in concrete research. Yet, they have proved very efficient in this case to make a preliminary study of a new phenomenon the massive growth of CH crystals in presence of surfactants, bringing complementary information and results. Of course, they absolutely do not make further investigation any less necessary. For example, a proper crystallographic study would be essential to fully understand the growing mechanisms and the action of the surfactants. In the field of materials characterization, it is important to be creative and to combine and adapt existing techniques. This is particularly true in the field of concrete research, and even more when it comes to concrete surface, a topic which is slowly emerging and where a lot of fascinating research still waits to be done.

References

[1] Gueit, E., Darque-Ceretti, E., Tintillier, P. & Horgnies, M., Surfactant-induced growth of calcium hydroxide at the concrete/formwork interface as a mineral coating for concrete, Manuscript submitted for publication.

[2] Skalny, J., Gebauer, I. & Odler, I., (eds). Calcium Hydroxide in Concrete, The American Ceramic Society: Westerville, 2001.

[3] Garbacz, A., Courard, L., & Kostana, K., Characterization of concrete surface roughness and its relation to adhesion in repair systems, Materials Characterization, 56, pp. 281-289, 2006.

[4] PER09 Perez, F., Bissonette, B. & Courard, L., Combination of mechanical and optical profilometry techniques for concrete surface roughness characterisation. Magazine of Concrete Research, 61(6), pp. 389-400, 2009.

[5] Ramirez, A.M., Demeestere, K., De Belie, N., Mntyl, T., & Levnen, E, Titanium dioxide coated cementitious materials for air purifying purposes: Preparation, characterization and toluene removal potential, Building and Environment, 45, pp. 832-838, 2010.



[6] Gallucci, E. & Scrivener, K., Crystallisation of calcium hydroxide in early model and ordinary cementitious systems, Cement andConcrete Reseach,37, pp. 492-501, 2007.

[7] NIH, http:\\rsbweb.nih.gov/ij



Characterization of the influence of the casting mould on the surface properties of concrete and on the adhesion of a protective coating

M. Horgnies, P. Willieme, O. Gabet, S. Lombard & M. Dykman Lafarge Centre de Recherche, St Quentin-Fallavier, France

Abstract

Protective coatings are deposited on concrete to improve aesthetics and to prevent ageing. However, their adhesion on concrete depends on several interlinked parameters. In this study, the surfaces of concrete are characterized according to the process of casting and post-treatment used (sandblasting) by using Scanning Electron Microscopy (SEM), Fourier Transformed-Infrared (FT-IR) spectroscopy and profilometry. The surface properties are correlated to the adhesion force of a polyurea (PU) coating. The development of a specific peel test (a strengthened and porous membrane is introduced into the layer of liquid coating before its crosslinking) ensures a reproducible debonding of the coating/concrete system and allows measuring the fracture energy. Moreover, the interface after debonding is analyzed by FT-IR to highlight the presence of concrete/coating residues and to determine the locus of failure. Results underline that the nature of casting mould influences the concrete surface and modifies the adhesion of PU coating. The mould made of polyoxymethylene (POM) induces a micro-tearing of the extreme surface of concrete during demoulding. By increasing the roughness and the open porosity of the concrete surface, this tearing enhances the adhesion of the coating. On the contrary, the smooth concrete surface, induced by the use of a polyvinylchloride (PVC) mould, decreases the anchorage of the coating. Finally, the sandblasting of the surface, by increasing the roughness and the interface area, is an interesting treatment to promote the adhesion of PU coating, whatever the mould used for the casting. Keywords: concrete, coating, roughness, FT-IR, SEM, peel test, adhesion.



doi:10.2495/MC110031

1 Introduction

The staining of concrete could occur due to its specific microstructure that retains the liquid and dust particles. The deposition of a coating is then important to close the surface porosity and protect concrete against acid rains, settlement of algae and lichensetc [1, 2]. PU coatings are commonly used due to their high resistance against chemical and mechanical aggressions [3, 4]. However, the surface properties of concrete depend on several interlinked parameters as chemical composition, intrinsic porosity and roughness. The aim of this study concerns the influence of the casting process and post-treatment on the adhesion between PU coating and concrete surface. The influences of the casting conditions on the hardened concrete surface were characterized by SEM, FT-IR and profilometry. Secondly, this study was undertaken to determine if the surface properties of concrete could influence the adhesion of PU coating. Fracture energies were measured by a specific 90-peel test. This method was retained because it is appropriate to characterize the adhesion of thin films [58]. Some publications have already showed the use of strengthened membrane or mesh sheet that were incorporated into the bulk of soft material to characterize [9]. Concerning our system, the introduction of a polymer membrane into the bulk of the coating was necessary to strengthen the system and measure a reproducible adhesion of the PU coating. After the peel tests, the FT-IR analyses of the debonded faces were undertaken to detect the residues of concrete or coating and determine the loci of failure. FT-IR spectroscopy allows detecting organic compounds of coatings [10, 11] and several components of concrete [12].

2 Material and methods

2.1 Material

2.1.1 Substrates made of hardened concrete A high-performance concrete was prepared by mixing 31% of white Portland cement (CEM I 52.5 PMES from Lafarge), 9% of limestone filler (DURCAL 1), 7% of silica fumes (MST), 43.5% of sand (BE01) and 1.5% of admixture. A water to cement ratio (W/C) of 0.26 was used. The samples were prepared by pouring the fresh concrete mixture into horizontal and rectangular formwork (15x12x1 cm) made of PVC or POM. The concrete samples were removed from their formworks after 18 hours and were stored during 28 days under ambient conditions (25C; 50% relative humidity) to complete their hydration. Some concrete samples were then sandblasted after demoulding (by using a powder of corindon) to increase their roughness. The sandblasted samples were cleaned by air flow to remove the dust before the deposition of coating.

2.1.2 Coating and conditions of deposition The PU coating was composed of 50% of isocyanates diluted into 45% of solvent (butyl acetate). A catalyst (dibutyltain laurate, DBTL) was added into the mix to initiate the reaction with water (present in concrete or in atmosphere).



The isocyanates units reacted with water to produce a polyurea-based film [13]. The coatings were sprayed (120 g/m) with air pressure of 3 atmospheres on the concrete surface. The drying period was about 3 days under ambient conditions (25C, 50% relative humidity) before peeling.

2.2 Methods of characterization

2.2.1 Scanning electron microscopy (SEM) Samples were characterized by using a high-resolution field-effect gun digital scanning electron microscope (SEM FEG Quanta 400 from FEI Company; using an accelerating voltage of 15 keV and a current intensity of 1 nA). Images of the cross-sections were obtained after being polished.

2.2.2 Profilometry The roughness of concrete samples was measured with a Surftest SJ-201 M mechanical profilometer (Mitutoyo) in order to calculate the arithmetic mean of the profile deviations from the mean line (Ra). The Ra value was obtained by compiling the arithmetic mean of 5 profiles of 12.5 mm.

2.2.3 Fourier transform-infrared spectroscopy (FT-IR) The FT-IR spectrometer Nicolet iS10 (Thermo Fisher Scientific Inc.) was equipped with a deuterated triglycine sulfate (DTGS) detector and controlled by OMNIC software. The Attenuated Total Reflexion (ATR) mode was mainly used in this study. FT-IR (in ATR mode) characterized the sample over a thickness of a few m. The sampling area analyzed was approximately 1 mm. The crystal used was made of diamond and 16 scans were routinely recorded over the range 4,000-650 cm-1 with a spectral resolution of 4 cm-1. The background was collected at ambient atmosphere before analyzing each sample. Spectra were corrected with a linear baseline. No specific preparations of the samples of concrete and coating were performed before FT-IR analyses: they were studied just after demoulding or after debonding.

2.2.4 Specific peel tests of concrete/coating system Peel test allows measuring the debonding force (F). According to the peel angle () and the width of the adhesive coating (w), the fracture energy (G) could be calculated according to [14]:

cos1

wF

G

In the specific case of 90 peel angle, G is equal to the peeling force (F) divided by the width (w) of the adhesive coating. The coating was strengthen

Documents

[A._Mammoli,_C._A._Brebbia,_A._Klemm]_Materials_Ch(BookFi.org).pdf