Editors:Bijan Samali, Mario M. Attard & Chongmin Song

EditorsSamaliAttardSong

From Materials to Structures:Advancement through Innovation

From M

aterials to Structures:A

dvancement through Innovation

From Materials to Structures: Advancement through Innovation is a collection of peer-reviewed papers presented at the 22nd Australasian Conference on the Mechanics of Structures and Materials (ACMSM22) held in Sydney Australia, from 11-14 December 2012 by academics, researchers and practising engineers mainly from Australasia and the Asia-Pacific region. The topics under discussion include:

Biomechanics Composite structures and materials Computational mechanics Concrete, masonry, steel and timber structures Earthquake engineering and structural dynamics Fire engineering Geomechanics Foundation engineering Innovative and smart structures Pavement engineering Rehabilitation of structures Rock engineering Site investigation Soil improvement and reinforcement Structural health monitoring Structural optimisation Sustainable materials

From Materials to Structures: Advancement through Innovation will be a valuable reference for academics, researchers and practising engineers working in structural and material engineering and mechanics.

an informa business

FROMMATERIALS TO STRUCTURES: ADVANCEMENT THROUGH INNOVATION

This page intentionally left blankThis page intentionally left blank

PROCEEDINGS OF THE 22ND AUSTRALASIAN CONFERENCE ON THE MECHANICS OFSTRUCTURES AND MATERIALS, ACMSM 22, SYDNEY, AUSTRALIA, 1114 DECEMBER 2012

From Materials to Structures:Advancement through Innovation

Editors

Bijan SamaliUniversity of Technology Sydney, Sydney, Australia

Mario M. AttardUniversity of New SouthWales, Sydney, Australia

Chongmin SongUniversity of New SouthWales, Sydney, Australia

CRC Press/Balkema is an imprint of the Taylor & Francis Group, an informa business

2013 Taylor & Francis Group, London, UK

Typeset by MPS Limited, Chennai, IndiaPrinted and bound in Great Britain by CPI Group (UK) Ltd, Croydon, CR0 4YY.

All rights reserved. No part of this publication or the information contained herein may bereproduced, stored in a retrieval system, or transmitted in any form or by any means,electronic, mechanical, by photocopying, recording or otherwise, without written priorpermission from the publishers.

Although all care is taken to ensure integrity and the quality of this publication and theinformation herein, no responsibility is assumed by the publishers nor the author for anydamage to the property or persons as a result of operation or use of this publicationand/or the information contained herein.

Published by: CRC Press/BalkemaP.O. Box 11320, 2301 EH, Leiden, The Netherlandse-mail: Pub.NL@taylorandfrancis.comwww.crcpress.com www.taylorandfrancis.com

ISBN: 978-0-415-63318-5 (hardback+USB)ISBN: 978-0-203-52001-7 (eBook)

From Materials to Structures: Advancement through Innovation Samali, Attard & Song (Eds) 2013 Taylor & Francis Group, London, ISBN 978-0-415-63318-5

Table of contents

Preface XVScientific committee XVII

Keynote papers

Analysis of dynamic penetration of soils 3J.P. Carter & M. Nazem

Strategies for structural health monitoring of bridges: Japans experience and practice 15Y. Fujino & D. Siringoringo

Recycled and renewable materials as resources for electric arc furnace steelmaking 21M. Zaharia, N.F. Yunos &V. Sahajwalla

Terror, security, and money: The risks, benefits, and costs of critical infrastructure protection 29M.G. Stewart & J. Mueller

Composite structures & materials

Concrete filled fabricated VHS tube to mild steel plate triangular stub columns under axialcompression load 43F. Alatshan, F.R. Mashiri & B. Uy

Multi-objective design optimisation of GFRP sandwich beams 49Z.K. Awad, T. Aravinthan, Y. Zhuge & F. Gonzalez

Effect of construction sequence in the axial shortening behaviour of composite columns 55N. Baidya, L. Zhang, P. Mendis & S. Fragomeni

Structural evaluation of concrete expanded polystyrene sandwich panels for slab applications 61R.M. Bajracharya, W.P. Lokuge, W. Karunasena, K.T. Lau & A.S. Mosallam

Long-term deformation of composite concrete slabs under sustained loading 67A. Gholamhoseini, R.I. Gilbert, M.A. Bradford & Z.T. Chang

Residual strength of timber-concrete composite beams after long-term test 73M. Hailu, C. Gerber, R. Shrestha & K. Crews

Effect of boundary conditions on the creep response of sandwich beams with a viscoelastic soft core 79E. Hamed & M. Ramezani

A review of FRP composite truss systems and its connections 85R.M. Hizam, A.C. Manalo &W. Karunasena

Post-critical behaviour of sandwich cylindrical shells with variable thickness 91P. Jasion & K. Magnucki

Analytical solution of multi-layer composite beam including interlayer slip and uplift 95A. Kroflic, M. Saje & I. Planinc

Time dependent behaviour of two-layer composite beams 101A. Kroflic, M. Saje & I. Planinc

Effect of ply configuration on hollow square reinforced concrete columns confined withCarbon fibre-reinforced polymer (CFRP) 107T.D. Le, M.T. Lester & M.N.S. Hadi

V

Electroelastic analysis of interface cracks and corners in piezoelectric composites usingscaled boundary finite element method 113C. Li, H. Man, C. Song &W. Gao

Long-term in-plane analysis of concrete-filled steel tubular arches under a central concentrated load 119K. Luo, Y.L. Pi, M.A. Bradford &W. Gao

A comparison of various plate theories for functionally graded material sandwich plates 127S. Natarajan & M. Ganapathi

Crack propagation modeling in functionally graded materials using polygon elements modeledby the scaled boundary finite element method 133E.T. Ooi, S. Guo & C. Song

Effect of eccentric load on retrofitted reinforced concrete columns confined with FRP 139T.M. Pham, X. Lei & M.N.S. Hadi

Application of externally post-tensioned FRP bars for strengthening reinforced concrete members 145A. Rajabi, H.R. Valipour, B. Samali & S. Foster

Evaluation of effective width of GFRP-steel composite beams for structural construction 149S. Satasivam, Y. Bai & X.L. Zhao

Behaviour of composite steel-concrete beams under elevated temperatures 155K.Wilkins, O. Mirza & B. Uy

Experimental trends of FRP-to-concrete joints anchored with FRP anchors 161H.W. Zhang & S.T. Smith

Computational mechanics

Multi-scale nonlinear elastic analysis of thin-walled members including local effects 169R.E. Erkmen

Thin plate bending analysis using the generalized RKP-FSM 175M. Khezri, Z. Vrcelj & M.A. Bradford

On volume change dependency of the soil water characteristic curve in numericalmodeling of unsaturated soils 183A. Khoshghalb & N. Khalili

Improved nonlinear analysis methods for determining the initial shape of cable-supported bridges 189M.Y. Kim, D.J. Min & M.M. Attard

Efficient bending analysis for piezoelectric plates using scaled boundary finite-element method 195H. Man, C. Song, W. Gao & F. Tin-Loi

Evaluation of stress intensity factors on cracked functionally graded materials using polygonsmodelled by the scaled boundary finite element method 201E.T. Ooi, I. Chiong & C. Song

In-plane stability of variable cross-section columns with shear deformations 207L. Su & M.M. Attard

Limit analysis in the presence of plasticity and contact 213S. Tangaramvong, F. Tin-Loi & C.M. Song

Application of explicit finite element analysis in solving practical structural engineering problems 219T. Watts, K. Kayvani & A. Kucyper

Fluid flow through single fracture using Lattice Boltzmann Method 225P. Yin & G. Zhao

Concrete structures

The behaviour of fibre reinforced continuous concrete slabs under load an experimental study 233F.M. Abas, R.I. Gilbert, S.J. Foster & M.A. Bradford

VI

Application of flexible faade systems in reducing the lateral displacement of concreteframes subjected to seismic loads 241P. Abtahi, B. Samali, M. Zobec & T. Ngo

Investigation of ground flint glass as a supplementary cementitious material in autoclavedlime-silica binders 247K. Angus, P.S. Thomas, K. Vessalas & A.S. Ray

Finite element model calibration of an instrumented RC building based on seismic excitationincluding non-structural components and soil-structure-interaction 251F. Butt & P. Omenzetter

Modelling of reinforced concrete beam response to repeated loading including steel-concreteinterface damage 257A. Castel, R.I. Gilbert, S.J. Foster & G. Ranzi

Shear strengthening of RC beam with external FRP bonding: A state-of-the-art review 263R. Choudhury, T.G. Suntharavadivel, P. Keleher & A. Patil

Evaluation of longitudinal bond shear stress and bond-slip relationship in composite concrete slabsusing partial shear connection method 269A. Gholamhoseini, R.I. Gilbert, M.A. Bradford & Z.T. Chang

Evaluation of mechanical properties of carpet fibre reinforced concrete 275N. Ghosni, B. Samali & K. Vessalas

Bond failure in grouted post-tensioned slab tendons with little or no initial prestress 281R.I. Gilbert

Effect of supporting conditions on the long-term load capacity of high strength concrete panels 287Y. Huang & E. Hamed

A mathematical model for complete stress-strain curve prediction of permeable concrete 293M.K. Hussin, Y. Zhuge, F. Bullen &W.P. Lokuge

Effects of temperature, relative humidity and outdoor environment on FRP-concrete bond 299M.I. Kabir, R. Shrestha & B. Samali

A preliminary investigation of the strength and ductility of lapped splices of reinforcing bars in tension 305A.E. Kilpatrick & R.I. Gilbert

Feasibility study of autonomous deformation control of PC viaducts 313M. Kunieda, N. Chijiwa, K. Ohara & K. Maekawa

Anchorage of deformed reinforcing bars in tension: An outlook for advanced formulationof a bond-slip constitutive law 319M.H. Mazumder, R.I. Gilbert & Z.T. Chang

Structural performance of 45 year old corroded prestressed concrete beams 325T.M. Pape & R.E. Melchers

Fatigue behaviour of reinforced concrete beams with addition of steel fibres 333A. Parvez & S.J. Foster

An experimental study on the shrinkage and ultimate behaviour of post-tensioned composite slabs 339G. Ranzi, A. Ostinelli & B. Uy

Instantaneous and long-term behaviour of cracked reinforced concrete slabs prepared withdifferent curing conditions 345M.M. Rahman, A. Ostinelli, G. Ranzi & R.I. Gilbert

Effect of reinforcement confinement on concrete cover cracking in reinforced concrete structures 351H.B. Sabtu & M.G. Stewart

Lateral strain of confined concrete incorporating size effect 357A.K. Samani & M.M. Attard

Combining high-strength self-compacting and normal-strength concretes in reinforcedconcrete frame structures 363M. Soleymani Ashtiani, R.P. Dhakal & A.N. Scott

VII

Investigating the arching action in reinforced concrete beams 369N.F. Vesali, H. Valipour, B. Samali & S.J. Foster

FEM modelling and analysis of reinforced concrete section with lightweight blocks infill 375A.S. Wahyuni, V. Vimonsatit & H. Nikraz

Effective stiffness of reinforced concrete section with lightweight blocks infill 381A.S. Wahyuni, V. Vimonsatit & H. Nikraz

Dynamic analysis of structures

Assessment of key response quantities for design of a cable-stayed bridge subjected tosudden loss of cable(s) 387Y. Aoki, B. Samali, A. Saleh & H. Valipour

Ambient vibration tests and analysis of a multiple-span elevated bridge 393X. Chen, P. Omenzetter & S. Beskhyroun

A novel piezoelectric wafer-stack vibration energy harvester 399X.Z. Jiang, Y.C. Li & J.C. Li

Optimization-based interval dynamic response analysis of a bridge under a moving vehiclewith uncertain properties 405N. Liu, W. Gao, C.M. Song & N. Zhang

Automatic dynamic crack propagation modeling using polygon scaled boundary finite elements 411E.T. Ooi, M. Shi, C. Song, F. Tin-Loi & Z.J. Yang

Dynamic analysis for plate structures by the scaled boundary finite-element method 417T. Xiang, H. Man, C. Song &W. Gao

Earthquake & wind engineering

A newly developed analytical model of transient downburst wind loads 425E. Abdelaal, X. Ma & J.E. Mills

Dynamic behaviour of flexible facade systems in tall buildings subjected to wind loads 431A. Azad, B. Samali, T. Ngo & C. Nguyen

Seismic risk analysis based on historical events reported in Sri Lanka 437P. Gamage & S. Venkatesan

Influence of infilled masonry wall on vibration properties and dynamic responses ofbuilding structures to earthquake ground excitations 443H. Hao

Behavior of reinforced concrete rectangular aboveground tanks subjected to near-sourceseismic excitations 449M. Hosseini & Sh. Abizadeh

A comparative study on the seismic performance of moment resisting frame steel buildings,designed by IBC and Eurocode, based on nonlinear time history analyses 455M. Hosseini & F. Sheikhlou

Optimal story-wise distribution of viscous dampers in a five-story building 461B. Kashani Madani & M. Hosseini

Effects of vertical seismic loadings on safety evaluation for earth dams 469H.J. Li & Z.W. Yan

Smart structures embedded with MR dampers using non-affine Fuzzy Control 475Z. Movassaghi, B. Samali & Q.P. Ha

Seismic performance improvement of stone masonry buildings in mud mortar 479R. Pun, B. Samali & H. Valipour

Quasi-static testing protocol for simulating earthquake conditions in regions of low-moderate seismicity 485R. Shahi, N. Lam, E. Gad & J. Wilson

VIII

Dynamic analysis of structures with interval parameters under random process earthquake excitations 491C.W.Yang, C. Wang, W. Gao & C.M. Song

Fibre composites

Assessment of wollastonite microfibre on drying shrinkage behaviour of cement-based composites 499N.L. Galea, P. Hamedanimojarrad, K. Vessalas & P.S. Thomas

Experimental study on the bondline behavior between concrete and FRP materials 505S.A. Hadigheh, R.J. Gravina, S. Setunge & S.J. Kim

An experimental investigation of a thermal break composite faade mullion section 513S. Huang, J. Li, B. Samali & M. Zobec

Mechanical properties of bamboo fiber-polyester composites 519A.C. Manalo, W. Karunasena & K.T. Lau

Influence of hooked-end steel fibers on absorbed energy of slurry-infiltrated fiber concretein flexural test 525Y. Shafaei & O. Eren

Properties and behaviour of gomuti fibre composites under tensile and compressive load 531A. Ticoalu, T. Aravinthan & F. Cardona

Formula for SIF of cracked steel plates strengthened with CFRP plate 537Q.Q. Yu, X.L. Zhao, T. Chen, Z.G. Xiao & X.L. Gu

Fire engineering

Influence of in-situ pore pressures and temperatures on spalling of reinforced concrete wallssubjected to hydrocarbon fire 543M. Guerrieri & S. Fragomeni

Thermal performance of non-load bearing LSF walls using numerical studies 549P. Keerthan & M. Mahendran

Prediction of shear failure of hollowcore slabs exposed to fire 555J.K. Min, R.P. Dhakal, A.K. Abu, P.J. Moss & A.H. Buchanan

A review on fire protection for phase change materials in building applications 561Q. Nguyen, T. Ngo & P. Mendis

Self-strengthening of structural steel members using shape memory alloys in fire 567H. Sadiq, M.B. Wong, X.L. Zhao & R. Al-Mahaidi

Foundation and pavement engineering

Review of residential footing design on expansive soil in Australia 575A.M.A.N. Karunarathne, E.F. Gad, S. Sivanerupan & J.L. Wilson

Analysis of pile group behaviour due to excavation induced ground movements 581R. Nishanthan, D.S. Liyanapathirana & C.J. Leo

Inelastic lateral seismic response of building frames under influence of bedrock depth variationsincorporating soil-structure interaction 587H.R. Tabatabaiefar, B. Fatahi & B. Samali

Numerical and experimental investigations of stress wave propagation in utility poles undersoil influence 593N.Yan, J. Li, U. Dackermann & B. Samali

Geomechanics

Comparison of existing design methods for geosynthetic reinforced pile-supported embankments:Three-dimensional numerical modelling 601P. Ariyarathne, D.S. Liyanapathirana & C.J. Leo

IX

Efficient modeling of wave propagation in unbounded domains using the scaled boundaryfinite element method 607X. Chen, C. Birk & C. Song

Experimental investigation of desiccation of clayey soils 613Y. Gui, G. Zhao & N. Khalili

A geotechnical site investigation by surface waves 619P. Harutoonian, C.J. Leo, D.S. Liyanapathirana & K. Tokeshi

A constitutive permeability evolution model for fractured porous media 625J. Ma, N. Khalili & G. Zhao

Comparisons of seismic rock slope stability assessments between the Hoek-Brown andMohr-Coulomb failure criteria 629Z.G. Qian, A.J. Li, V. Kong & A.V. Lyamin

Permeability of the fractured rockmass A review 635K.K. Singh, D.N. Singh & P.G. Ranjith

Horizontal-to-vertical spectral ratio inversion using Monte Carlo approach and enhanced byRayleigh wave dispersion curve 641K. Tokeshi, P. Harutoonian, C.J. Leo, D.S. Liyanapathirana & R. Golaszewski

An investigation of arching mechanism of geosynthetic reinforced column supported embankments 647N.N.S. Yapage, D.S. Liyanapathirana, C.J. Leo, H.G. Poulos & R.B. Kelly

Mechanics of materials

Synergistic energy absorption in the in-plane static compression response of filled honeycombs 655R.J. DMello & A.M.Waas

Stress analysis of cemented wellbores in geosequestration of carbon dioxide 661M.G. Haider, J. Sanjayan & P.G. Ranjith

Metal surface profile and residual stress: Persuasion of adhesion 667M.S. Islam, L. Tong & P.J. Falzon

Strength of glass under concentrated force 673H. Jiang, N.T.K. Lam, L. Zhang & E.F. Gad

Orthotropic Simo and Pister hyperelasticity 679D.C. Kellermann & M.M. Attard

Pressure correction in water-bag testings to investigate post cracked behaviour of laminated glass 685R. Lumantarna, C. Nguyen, M. Zobec & T. Ngo

Biomechanical environment of early stage of bone healing under biological internal fixation 691S. Miramini, L. Zhang, P. Mendis & M. Richardson

Material structural design with isotropy constraint 697A. Radman, X. Huang &Y.M. Xie

Experimental study on mixed-mode fracture between concrete and rock 703H. Zhong, T. Ding & G. Lin

In-plane buckling analysis of funicular arches with pinned supports 709J. Zhu & M.M. Attard

New design and construction technologies

Structural performance under lateral loads of innovative prefabricated modular structures 717T. Gunawardena, T. Ngo, P. Mendis, L. Aye & J. Alfano

Displacement based design method for outrigger braced tall buildings 723N. Herath, P. Mendis, T. Ngo & N. Haritos

Innovative materials for next generation faade systems 729Q. Nguyen, P. Mendis, T. Ngo, P. Tran & C. Nguyen

X

Review of diaphragm actions in domestic structures 735I. Saifullah, E.F. Gad, J.L. Wilson, N.T.K. Lam & K.Watson

Study of blockage effect on scouring pattern downstream of a box culvert 741S. Sorourian, A. Keshavarzi, B. Samali & J. Ball

Shock and impact loading

Fundamentals of impact actions demonstrated by miniature experimentations 747M. Ali, J. Sun, N. Lam, L. Zhang & E. Gad

Numerical simulation of impact pile driving and its effect on far field 751S.D. Ekanayake, D.S. Liyanapathirana & C.J. Leo

Effects of energy level and impact repetitions on the impact fatigue behaviour and post-impactflexural properties of square FRP pultruded tubes 757E.J. Guades, T. Aravinthan, A.C. Manalo & M.M. Islam

A novel adaptive base isolator utilising magnetorheological elastomer 763Y.C. Li, J.C. Li & B. Samali

Numerical modelling of composite textile subjected to impact loading 769P. Tran, T. Ngo, E.C. Yang, P. Mendis &W. Humphries

Bio-inspired composite structure subjected to underwater impulsive loading 775P. Tran, T. Ngo & P. Mendis

Numerical simulation of concrete spalling under impact 781C.Wu & L. Shen

Simulation of pressure impulse diagrams for foam protected RC members 787C.Wu & H. Sheikh

Impact analyses simplified by the two-degrees-of-freedom models 793Y. Yang, N. Lam, L. Zhang & E. Gad

Steel and aluminium structures

Testing of steel-CFRP adhesive joints under freeze-thaw cycling 801A. Agarwal, S. Foster, E. Hamed & Z. Vrcelj

A new kinetic model for steel specific heat during phase transformation 807H. Fang, M.B. Wong &Y. Bai

Finite element modeling of a beam-column connection in industrial storage racking structures 813A. Firouzianhaji, A. Saleh & B. Samali

Shear tests of lipped channel beams with stiffened web openings 819P. Keerthan & M. Mahendran

The use of neural networks for identification of parameters of semi-rigid connections 825A. Kozowski & L. Ziemianski

Finite element modeling of existing cable net structures 829G.J. Lume

Theoretical research on cold-formed channel sections under bending 835S. Maduliat, P. Mendis, T.D. Ngo & M.R. Bambach

Analysis of a railway turnout system with a spot replacement sleeper 841A.C. Manalo, T. Aravinthan &W. Karunasena

Stability reinforcement of steel plates by heat-induced stress deformation fields 847N. Schillo, D. Schaefer & M. Feldmann

Numerical study of block shear strength of coped beams bolted with angles/tee-section 853K.S. Seak, A.C.C. Lam & M.C.H. Yam

XI

Re-evaluation of shear strength of high strength bolts in AS 4100 859R.H.R. Tide

Structural health monitoring

A comparative study on the performance of the damage detection methods in the frequency domain 867M.M. Alamdari, J. Li & B. Samali

A FRF-based damage detection method utilising wavelet decomposition 873M.M. Alamdari, J. Li & B. Samali

Adaptive multiple forgetting factor recursive least square (AMFF-RLS) for real-timestructural identification 879M. Askari, J. Li & B. Samali

Integrated bridge deterioration modeling for concrete elements incorporating Elman Neural Network 885G.P. Bu, J.H. Lee, H. Guan &Y.C. Loo

Transmissibility function analysis for boundary damage identification of a two-storey framedstructure using artificial neural networks 891U. Dackermann, J. Li & B. Samali

Numerical computation of dispersion relations in three-dimensional waveguides 897H. Gravenkamp, J. Prager, H. Man, C. Birk & C. Song

Long-term monitoring of vibration properties of structures with different materials andboundary conditions 903H. Hao

Deterioration prediction of concrete bridges with artificial neural network (ANN) derivedfrom discrete condition data 909M.S. Hasan, S. Setunge & D.W. Law

Structural damage detection using the Wiener filter 915M. Jayawardhana, X. Zhu & R. Liyanapathirana

Evaluation of thermal gradients and their effects in a bridge box girder using monitoring data 921P. Omenzetter, P. Chua, A. Issa & B. Sanders

Numerical modelling of Lamb waves in cracked plates using the scaled boundary finite element method 927A.A. Saputra, C. Birk, C. Song & H. Gravenkamp

Mechanical properties of epoxy grouts for structural repair 933M. Shamsuddoha, M.M. Islam, T. Aravinthan, A.C. Manalo & K.T. Lau

Damage detection in a timber bridge model 939D. Tran, S. Venkatesan & S. Fragomeni

Structural optimisation and reliability

Reliability analysis of steel pipeline welds subjected to long-term seawater exposure 947I.A. Chaves & R.E. Melchers

Statistical safety factor calibration of short concrete-filled steel tubular columns 953W.H. Kang, B. Uy, Z. Tao & S. Hicks

Model updating of a full-scale bridge structure using particle swarm optimization 959F. Shabbir & P. Omenzetter

Spatial reliability analysis of reinforced concrete columns subject to explosive blast loads 965Y. Shi & M.G. Stewart

Uncertainty in long-term analysis of concrete-filled steel tubular columns under sustained loading 971X. Shi, W. Gao, Y.L. Pi & M.A. Bradford

Topology optimization under displacement and softening constraints 977S. Tangaramvong, F. Tin-Loi &Y.M. Xie

XII

Damage severity estimation using Parallel Genetic Algorithm and power spectral density 983M. Varmazyar, N. Haritos & M. Kirley

Modal analysis of structures with mixed random and fuzzy parameters 991C.Wang, J.W. Feng, W. Gao & C.M. Song

Uncertain limit analysis of structures with interval parameters 997D.Wu, W. Gao, F. Tin-Loi & S. Tangaramvong

Sustainability of structures and materials

A novel acid resistant green mortar for high corrosive environments 1005G. Adam, S. Salek, B. Samali, P. Battista & M. McKinnon

Environmental impact assessment of post tensioned and reinforced concrete slab construction 1009D. Miller, J.H. Doh, H. Guan, M. Mulvey, S. Fragomeni, T. McCarthy & T. Peters

Deterioration of concrete structures in Australia under a changing climate 1015L. Peng & M.G. Stewart

High-strength self-compacting concrete for sustainable construction 1021R. Sri Ravindrarajah

Advancement in construction: Application of Rapid Quality Identification Techniquein stone for risk mitigation and sustainability 1027D. Sarma & M.D. Sarma

Damage potential to residential structures due to ground movement 1033D.Wagle, E.F. Gad, S. Sivanerupan & J.L. Wilson

An overview of sustainable concrete made with scrap rubber 1039O.Youssf & M.A. ElGawady

Timber engineering

Steel-timber hybrid structures Design performance and dynamic behaviour 1047C. Dickof, S.F. Stiemer & S. Tesfamariam

Review on long-term behaviour of timber-concrete composite floors 1053N. Khorsandnia, H.R. Valipour, R. Shrestha, C. Gerber & K. Crews

The predictive model for strength of inclined screws as shear connection in timber-concretecomposite floor 1059F. Moshiri, C. Gerber, H.R. Valipour, R. Shrestha & K.I. Crews

Dynamic performance of timber flooring systems 1065R. Rijal, B. Samali, R. Shrestha, G. Gerber & K. Crews

Stressed cross-laminated-timber for bridge applications 1071L. Shearman, C. Gerber & K. Crews

Behaviour of stress wave propagation in utility timber pole 1077M. Subhani, J. Li & B. Samali

Ultimate performance of timber connection with normal screws 1083Z. Zabihi, R. Shrethta, B. Samali & K. Crews

Concrete/material technology

Characterisation of cement mortar containing oil-contaminated aggregates 1091M.H. Almabrok, R.G. McLaughlan & K. Vessalas

An experimental investigation into the tensile strength of steel fibre reinforced concrete 1097A. Amin, S.J. Foster, D. Boillet & A. Muttoni

Comparison of the analytical models to determine modulus of rupture of self-compactingconcrete and conventional concrete 1105F. Aslani & S. Nejadi

XIII

Effects of fly ash on compressive strength of structural concrete 1113R.J. Case, K. Duan & T.G. Suntharavadivel

Brown coal fly ash geopolymer mortar 1119R. Dirgantara, D.W. Law, T.K. Molyneaux & D. Kong

Evaluation of fresh properties effect on the compressive strength of polypropylene fibrereinforced polymer modified concrete 1123N. Ghosni, K. Vessalas & B. Samali

Engineered cementitious composites incorporating recycled concrete fines 1129J. Li, P. Surya & E.H. Yang

Effect of different concentrations of lime water on mechanical properties of high volumefly ash concrete 1135X.H. Ling, S. Setunge & I. Patnaikuni

Recycled aggregate concrete prepared with water-washed aggregates An investigative study 1143C.Y. Lo, S.H. Chowdhury & J.H. Doh

Mechanical properties of polymer concrete with different types of resin 1147W.P. Lokuge & T. Aravinthan

Investigation on the mathematical models of chloride diffusion coefficient in concrete exposed tomarine environment 1153F. Nabavi, S. Nejadi & B. Samali

Investigation of flint glass for partial replacement of fine aggregate in fly ash cement-based mortars 1159A. Ngadimin, K. Vessalas, P. Thomas & P. Hamedanimojarrad

Effect of polyvinyl alcohol fibre and fly ash on flexural tensile properties of concrete 1165A. Noushini, K. Vessalas, N. Ghosni & B. Samali

Investigation into cavitation as a cause of rate-dependent fatigue loss in submerged concrete members 1171M. Sagan, C. Fujiyama & K. Maekawa

Mechanical properties of fibre reinforced reactive powder concrete after exposure to high temperatures 1177S. Sanchayan, N. Gowripalan & S.J. Foster

Assessment of bottom ash use as fine aggregate replacement in concrete 1183R. Satsangi, K. Vessalas & S. Russell

Assessment of compressive strength of elastomeric modified concrete incorporating waste tyre rubber 1187N. Sharifi, K. Vessalas & B. Samali

Acids attack on silica fume high-strength concrete 1193R. Sri Ravindrarajah

Drying shrinkage of concrete made from recycled concrete aggregate 1199B.A. Whiting, T.J. McCarthy & E. Lume

Author index 1205

XIV

From Materials to Structures: Advancement through Innovation Samali, Attard & Song (Eds) 2013 Taylor & Francis Group, London, ISBN 978-0-415-63318-5

Preface

The organisers of the 22ndAustralasian Conference on the Mechanics of Structures and Materials (ACMSM22)extend a warm welcome to all participants whose presence and contributions will no doubt be a key factor forthe success of this conference. This conference is hosted jointly by the Centre for Built Infrastructure Researchat the University of Technology Sydney (UTS) and the Centre for Infrastructure Engineering & Safety at theUniversity of NewSouthWales (UNSW).The theme for the 2012ACMSMconference isMaterials to Structures:Advancement through Innovation. The first Australasian conference on mechanics of structures and materialsbegan at the University of New SouthWales in 1967 as an initiative of the late Prof F.S. Shaw. Subsequently, theseconferences have been held biennially as a forum for exchanging the latest research in the field of mechanics ofstructures and materials by researchers in the Australasian region and beyond. The last conference, ACMSM21,was held at Victoria University in Melbourne in December 2010.

Over the forty five year span and twenty one conventions of the ACMSM conferences, there has been con-tinuous research growth in the understanding of infrastructure and the emergence of new and green materialshas added more impetus and relevance to these conferences. The ACMSM has become a biennial forum foracademics, researchers and practising structural and construction engineers, as well as materials scientists in theregion, fostering the exchange of ideas and detailing the research challenges in infrastructure development inour region.

The peer reviewed papers contained in these proceedings were accepted for presentation atACMSM22, held attheAerial Function Centre, University ofTechnology, Sydney,Australia from 1114 December 2012.The almost200 papers were authored by academics, researchers and practising engineers from many countries around theworld and cover a broad range of structural engineering and materials research under the following topics:

Biomechanics Pavement engineering Composite structures and materials Rehabilitation of structures Computational mechanics Rock engineering Concrete, masonry, steel and timber structures Site investigation Earthquake engineering and structural dynamics Soil improvement and reinforcement Fire engineering Structural health monitoring Foundation engineering Structural optimisation Geomechanics Sustainable materials Innovative and smart structures

The abstracts submitted were initially reviewed by the organising committee and authors of those abstracts thatfell within the scope of the conference were asked to submit full papers for peer review. All the papers includedin these proceedings were subjected to rigorous review by the experts in relevant fields. This peer review processresulted in many papers being improved and some papers being rejected. The editors would like to acknowledgethe contributions made to the conference by the Scientific Committee who undertook the task of reviewing allthe submitted papers.

The editors would also like to thank all the keynote speakers, authors, participants and members of the localorganising committee especially Dr. David Kellermann, Dr. Michael Man and Dr. Ean Tat Ooi, for their effortand support for this conference.

On behalf of the ACMSM22 Organising Committee, we welcome you to exciting Sydney and hope that youfind the conference inspiring and enjoyable.

Bijan SamaliMario Attard

Chongmin SongSeptember 2012

XV

This page intentionally left blankThis page intentionally left blank

From Materials to Structures: Advancement through Innovation Samali, Attard & Song (Eds) 2013 Taylor & Francis Group, London, ISBN 978-0-415-63318-5

Scientific committee

SCIENTIFIC COMMITTEE

Abhi Ray, University of Technology, Sydney, AustraliaAdrian Russell, University of New South Wales, AustraliaAlfonso Nappi, University of Trieste, ItalyAli Saleh, University of Technology, Sydney, AustraliaAndrew Deeks, Durham University, UKArman Khoshghalb, University of New South Wales, AustraliaBehzad Fatahi, University of Technology, Sydney, AustraliaBijan Samali, University of Technology, Sydney, AustraliaBrian Uy, University of Western Sydney, AustraliaCarolin Birk, University of New South Wales, AustraliaChongmin Song, University of New South Wales, AustraliaDak Baweija, University of Technology, Sydney, AustraliaDavid A. Nethercot, Imperial College, UKDavid Thambiratnam, Queensland University of Technology, Brisbane, AustraliaEhab Hamed, University of New South Wales, AustraliaFrancis Tin-Loi, University of New South Wales, AustraliaGianluca Ranzi, University of Sydney, AustraliaGiles W. Hunt, Univeristy of Bath, UKHadi Khabbaz, University of Technology, Sydney, AustraliaHamid Valipour, University of Technology, Sydney, AustraliaHao Zhang, University of Sydney, AustraliaHong Guang, Griffith University, Griffith, AustraliaHossein Taiebat, University of New South Wales, AustraliaIchiro Ario, Hiroshima University, JapanJan G.M. van Mier, Institut fr Baustoffe (IfB), Zurich, SwitzerlandJay Sanjayan, Swinburne University of Technology, Hawthorn VIC, AustraliaJianchun Li, University of Technology, Sydney, AustraliaJohn Carter, University of Newcastle, Newcastle, AustraliaJohn Wilson, Swinburne University of Technology, Swinburne, AustraliaKaru Karunasena, University of Southern Queensland, Toowoomba, AustraliaKeith Crews, University of Technology, Sydney, AustraliaKenny Kwok, University of Western Sydney, AustraliaKim Rasmussen, University of Sydney, AustraliaKirk Vessalas, University of Technology, Sydney, AustraliaKoichi Maekawa, University of Tokyo, JapanLuming Shen, University of Sydney, AustraliaMario M. Attard, University of New South Wales, AustraliaMark A. Bradford, University of New South Wales, AustraliaMark Stewart, University of Newcastle, Newcastle, AustraliaMichaelYu Wang, The Chinese University of Hong Kong, Shatin, NT, Hong KongMoon-Young Kim, Sungkyunkwan University, Suwon, South KoreaMuhammad Hadi, University of Wollongong, Wollongong, AustraliaNasser Khalili, University of New South Wales, AustraliaPriyan Mendis, Melbourne University, Melbourne, AustraliaHong Hao, University of Western Australia, Perth, AustraliaRajesh Dhakal, University of Canterbury, Canterbury, New ZealandRaymond Ian Gilbert, University of New South Wales, AustraliaRob Melchers, University of Newcastle, Newcastle, AustraliaSritawat Kitipornchai, City University of Hong Kong, HKSam Fragomeni, University of Victoria, Melbourne, Australia

XVII

Shami Nejadi, University of Technology, Sydney, AustraliaSL Chan, The Hong Kong Polytechnic University, HKSri Ravindrarajah, University of Technology, Sydney, AustraliaSrikanth Venkatesan, University of Victoria, Melbourne, AustraliaStephen J. Foster, University of New South Wales, AustraliaSujeeva Setunge, RMIT University, Melbourne, AustraliaThiru Aravinthan, University of Southern Queensland, Toowoomba, AustraliaTommy Chan, Queensland University of Technology, Brisbane, AustraliaVeena Sahajwalla, University of New South Wales, AustraliaWei Gao, University of New South Wales, AustraliaWilfried Becker, TU Darmstadt, GermanyYew-Chaye Loo, Griffith University, Griffith, AustraliaYi-Min Xie, RMIT University, Melbourne, AustraliaYozo Fujino, University of Tokyo, JapanZhao Xiao, Monash University, Clayton, AustraliaZhenjunYang, University of Manchester, Manchester, UKZhong Tao, University of Western Sydney, AustraliaZora Vrcelj, University of New South Wales, Australia

LOCAL ORGANISING COMMITTEE

Bijan Samali, University of Technology, Sydney, Australia (chair)Mario Attard, University of New South Wales, Australia (co-chair)Chongmin Song, University of New South Wales, Australia (co-chair)Ean Tat Ooi, University of New South Wales, AustraliaMichael (Hou) Man, University of New South Wales, AustraliaDavid Kellermann, University of New South Wales, Australia

XVIII

Keynote papers

This page intentionally left blankThis page intentionally left blank

From Materials to Structures: Advancement through Innovation Samali, Attard & Song (Eds) 2013 Taylor & Francis Group, London, ISBN 978-0-415-63318-5

Analysis of dynamic penetration of soils

J.P. Carter & M. NazemCentre for Geotechnical and Material Modelling, The University of Newcastle, Callaghan NSW, Australia

ABSTRACT: The finite element analysis of dynamic penetration of soil deposits, by either free falling orpropelled objects, is one of the most sophisticated and challenging problems in geomechanics.A robust numericalmethod is described here for dealing with such complex and difficult problems. The approach is based onthe Arbitrary Lagrangian-Eulerian (ALE) method of analysis, whose main features and challenges are brieflydescribed. The ALE method is then employed to perform a parametric study of a variety of penetrometers inlayers of inhomogeneous clay deforming under undrained conditions. The effect of the mechanical properties ofthe clay soil on the penetration characteristics is discussed.

1 INTRODUCTION

Penetrometers are widely used to investigate themechanical properties of soils and to embed objectsin soil deposits, especially anchors and other itemsin the sea floor. Free falling penetrometers have beenemployed to provide information on the strength ofseabed sediments. Such tests can be envisaged as thedynamic equivalent of a static cone penetration test(CPT). They can provide useful data such as the totaldepth and time of penetration and the decelerationcharacteristics of the falling penetrometer. Potentially,these data can then be used to deduce strength param-eters for the soil in situ. Other types of projectiles arealso used to embed anchors and other objects deeplyin the sea floor. Large scale anchors of this type havebeen employed for ship moorings and the mooring offloating (offshore) petroleum production facilities.

To date, dynamic penetrometers have been usedfor offshore oil and gas industry applications such asdetermining soil strengths for pipeline feasibility stud-ies and anchoring systems, in military applicationsfor naval mine countermeasures and terminal ballis-tic studies, in extra-terrestrial exploration, and theyhave also been proposed and investigated for deepsea nuclear waste disposal (Chow 2011). The spec-ifications of a dynamic penetrometer, including itsshape, geometry, mass, and initial velocity, normallydepend on its application. For instance, torpedoes usedto anchor flexible risers are typically 1215 m long,weigh 240950 kN in air, and are 7621077 mm indiameter (Medeiros 2002), while the STING pen-etrometer, developed by Canadas Defence ResearchEstablishment Atlantic, to investigate the upper seabed consists of a 1 m long 19 mm diameter steel rodwith an enlarged tip at its base with diameters of 2570 mm (Mulhearn 2002). Numerical and experimentalstudies have shown that the penetration characteristics

of these penetrometers depend on their geometry andimpact energy as well as the mechanical propertiesof the soil layer, particularly its undrained properties.Scott (1970), Dayal andAllen (1973),True (1975), andBeard (1977) are among the pioneers who presentedthe earliest applications of free falling penetrometersin providing information on the upper few metres ofseabed sediments. At the time of these studies pow-erful numerical methods were not available to solvethe corresponding theoretical problem and researchershad to rely on empirical correction factors to interprettest data. Scott (1970) also demonstrated that techno-logical limitations prevented reliable data from beingobtained. Over the last decade, improvements in sen-sor technology have led to renewed interest in thedynamic penetration resistance of soil with severalnew penetrometer systems being developed to inves-tigate soil strength and bearing capacity. One of thefactors preventing the widespread use of this technol-ogy is the uncertainty in the method of interpretation.For example, experimental studies (Mulhearn, 2002,Abelev et al., 2009) with penetrometers of differentdiameters have indicated limitations of the widelyused empirical methods of interpretation developedby True (1975), and have demonstrated the need fora better understanding of the factors affecting thedynamic soil resistance.

For a variety of reasons, the numerical simulation ofthis type of penetration problem is probably one of themost sophisticated, difficult and challenging problemsin theoretical geomechanics. From the point of viewof the material, the analyst must consider the relativeincompressibility of the soil during the short testingtimes required for full penetration and the likely inho-mogeneity and rate dependence of the soil depositedon the seabed. This requires a robust stress-integrationalgorithm to accurately evaluate the stresses for a givenstrain field. Inertia effects may also be important in

3

cases where the penetration occurs rapidly. Moreover,the penetration of objects into layers of soil usuallyinvolves severe mesh distortion caused by the largedeformations. More importantly, the boundary condi-tions of the problem do not remain constant duringthe analysis since the contact surface between thepenetrometer and the soil changes continuously dur-ing penetration. Finally, an advanced time-integrationscheme must be employed to accurately predict thehighly nonlinear response of the soil.

This paper describes a robust numerical methodfor dealing with such complex and difficult prob-lems. The favoured approach is based on the ArbitraryLagrangian-Eulerian (ALE) method of analysis,whose main features and challenges are describedbriefly in the paper. Application of this method to thesimulation of the dynamic penetration of instrumentsinto undrained layers of non-uniform soil is discussed,and comparisons with experimental observations aremade.

2 ARBITRARY LAGRANGIAN-EULERIANMETHOD

2.1 Background

The Arbitrary Lagrangian-Eulerian (ALE) method,being well established in fluid mechanics and solidmechanics, has been shown to be robust and efficientin solving a wide range of static as well as dynamicgeotechnical problems involving large deformations.Nazem et al. (2006) presented a robust mesh optimisa-tion procedure used with the operator-split techniquefor solving large deformation problems in geome-chanics. This technique refines a mesh by relocatingthe nodal points on all material boundaries followedby a static analysis. Later, Nazem et al. (2008) pre-sented an ALE formulation for solving elastoplasticconsolidation problems involving large deformations.These authors compared the performance of the ALEmethod with the Updated Lagrangian (UL) methodand showed the efficiency and robustness of the ALEmethod by solving some classical problems such asthe consolidation of footings and cavity expansion.Sheng et al. (2009) addressed a successful applicationof the ALE method, incorporated with an automaticload stepping scheme and a smooth contact discreti-sation technique, in solving geotechnical problemsinvolving changing boundary conditions, such as pen-etration problems including the installation of piles.The application of the ALE method in the dynamicanalysis of geotechnical problems was also addressedby Nazem et al. (2009a).

2.2 Operator-split technique

The Arbitrary Lagrangian-Eulerian (ALE) methodhas been developed to eliminate the mesh distortionthat usually occurs in the Updated Lagrangian (UL)method by separating the material and mesh displace-ments. In a UL formulation, all variables are calculated

at the end of the last equilibrium configuration. Thisassumption necessitates updating the spatial coordi-nates of all material points according to the incremen-tal displacements at the end of each time step. Whilecapable of solving some problems with relatively largedeformations, the UL method often fails to succeeddue to the occurrence of excessive mesh distortionand the entanglement of elements. Mesh distortionusually starts with elements twisting or otherwise dis-torting out of their favourable shapes. For instance, ina domain consisting of triangular elements it is usuallypreferred that all triangles remain roughly equilateralduring the analysis. However, material displacements,especially in regions with higher deformation gra-dients, will gradually cause distortion of elements,decreasing the accuracy of the analysis or even ulti-mately resulting in a negative Jacobian. One way toovercome this shortcoming of the UL method is toseparate the material and mesh displacements, leadingto the development of the ALE method.

The equilibrium equation in theALE method can bewritten in two different forms. It is possible to writethe governing global system of equations in termsof two sets of unknown mesh and material displace-ments, leading to the so-called coupled ALE method.A supplementary set of equations in terms of the mate-rial and mesh displacements needs to be establishedthrough a mesh motion scheme and the two sets ofunknown displacements are then solved simultane-ously. Alternatively, in the decoupled ALE method orthe operator-split technique, the analysis can be per-formed in two separate steps: a UL step followed byan Eulerian step. In the UL step, the governing equa-tions are only solved for the material displacements inorder to fulfil the requirements for equilibrium. Thisstep usually results in a distorted mesh. In the subse-quent Eulerian step, the main goal is to minimise themesh distortion by refining the mesh. Mesh refine-ment can be achieved by generating a new mesh for theentire domain or by moving current nodal points intonew positions. Regardless of which strategy is adoptedfor mesh refinement, the topology of the domain, i.e.,the number of nodes, number of elements, and theconnectivity of elements should not be changed. TheEulerian step is particularly important if significantmesh distortion occurs during the UL step.

After mesh refinement, all state variables at nodalpoints as well as at Gauss points must be mapped fromthe distorted mesh to the new mesh. This remappingis usually performed using a first order expansion ofTaylors series as

where fr and f denote the time derivatives of an arbi-trary function f with respect to the mesh and materialcoordinates respectively, vi is the material velocity,and vri represents the mesh velocity. Other importantaspects of the ALE method are discussed as follows.

4

Figure 1. Node-to-segment contact discretisation.

2.3 Updated Lagrangian formulation

During each increment of the operator split technique,the analysis starts with a UL procedure at time t. Themain goal of the UL step is to find the incrementaldisplacements, velocities and accelerations which willsatisfy the equilibrium at time t+t. The matrix formof equilibrium is usually derived from the principle ofvirtual work. The weak form of this principle statesthat for any virtual displacement u, equilibrium isachieved provided

where k is the total number of bodies in contact, denotes the Cauchy stress tensor, is the variation ofstrain due to virtual displacement, u, u and u representmaterial displacements, velocities and accelerations,respectively, and c are the material density and damp-ing, b is the body force, q is the surface load acting onarea S of volumeV , gN and gT are the virtual normaland tangential gap displacements, tN and tT denote thenormal and tangential forces at the contact surface Sc.The solution of equation (2) requires the discretisationof the domain as well as the contact surfaces. In thisstudy we adopt the so-called node-to-segment contactdiscretisation as shown in Figure 1, in which a node onthe slave surface may come into contact with an arbi-trary segment of the master contact surface. With thisassumption the last two terms in equation (2), repre-senting the virtual work due to normal and tangentialcontact forces, can be written in the following form

in which nc is the total number of slave nodes, andFcN and F

cT represent the nodal forces of the contact

element.

The equation of motion for a solid can be obtainedby linearisation of the principle of virtual displace-ment and is written in the following matrix form(e.g., Wriggers 2006)

where M and C represent respectively the mass anddamping matrices, R is the stress divergence term, udenotes the displacement vector, and Fext is the time-dependent external force vector. Note that the rightsuperscript denotes the time when the quantities aremeasured and a superimposed dot represents the timederivative of a variable.The solution of the momentumequation requires a step-by-step integration scheme inthe time domain. Explicit and implicit methods areavailable for this purpose. In explicit methods the solu-tion at time t+t depends only upon known variablesat time t. These methods, such as the central differencemethod, do not require a factorization of the effectivestiffness matrix and are easy to implement. However,explicit methods are conditionally stable, i.e., the sizeof the time steps must be smaller than some criticaltime step. Implicit methods, on the other hand, requirethe solution of a nonlinear equation at each time stepand have to be combined with another procedure suchas the Newton-Raphson method. The main advantageof implicit methods is that they can be formulatedto be unconditionally stable, allowing the analystto use a bigger time step than used in the explicitmethods.

In this study we adopt the implicit generalised-method presented by Chung and Hulbert (1993)to integrate the momentum equation over the timedomain. This method has been shown to be very effi-cient in solving dynamic problems in geomechanics(e.g., Kontoe et al.2008 and Nazem et al.2009a). In thegeneralised- method, two integration parameters, fand m, are introduced into the momentum equation tocompute the inertia forces at time t+ (1 m)t andthe internal and damping forces at time t+ (1 f )t,respectively. The accelerations and velocities areapproximated by Newmark equations according to

where and are the Newmark integration parametersand t represents the time step.

By introducing a tangential stiffness matrix

5

we can use the Newton-Raphson method to solve themomentum equation in the following form for the ithiteration

and where Fint , the internal force vector, is obtainedaccording to the Cauchy stress tensor, , and the nodalforces at contact surfaces are obtained from

Note that the tangential stiffness matrix in equation (7)includes the contribution of the material stiffness,Kep,the stiffness due to geometrical nonlinearity, Knl , andthe stiffness due to normal and tangential contact, KNsand KTs, i.e.,

In the problems studied here the penetrometer isidealised as a rigid body, i.e., its size and shaperemain unchanged during penetration. To achieve thiscondition all nodal points on the penetrometer areprescribed to undergo equal vertical displacementswhile horizontal displacement is prohibited.

2.4 Definition of contact

To describe the contact at the interface between twobodies, constitutive equations must be provided forboth the tangential and normal directions. Among sev-eral strategies available in contact mechanics we usethe penalty method to formulate these constitutiverelations. The normal contact in the penalty methodis described by

where N is a penalty parameter for the normal contact.In general, the response in the tangential direction

is captured by the so-called stick and slip actions. In

the former, no tangential relative movement occursbetween the bodies whilst the latter represents a rela-tive displacement of bodies in the tangential direction.This assumption facilitates splitting the relative tan-gential velocity between the bodies, gT , into a stickpart gstT and a slip g

slT part, as in the following rate

form

The stick part can be used to obtain the tangentialcomponent of traction by

where T is a penalty parameter for tangential con-tact. Note that the assumption in (12) is analogousto the theory of plasticity in which the incrementalstrains are divided into an elastic and a plastic part.Following this analogy, we must provide a slip criterionfunction. This can be achieved by writing the classicalCoulomb friction criterion in the following form

where is the friction coefficient.For the numerical examples considered later in this

paper, the interface between the penetrometer and thesoil is assumed, for simplicity, to be perfectly smooth.This situation corresponds to the special case where= 0.

2.5 Material behaviour

The undrained behaviour of the soil to be penetrated isrepresented by an elastoplastic Tresca material modelwith an associated flow rule. In a large deformationanalysis, the stress-strain relations must be frame inde-pendent. This requirement, known as the principle ofobjectivity, can be satisfied by introducing an objec-tive stress rate, such as the Jaumann stress rate J ,into the stress-strain relations as

where Cep represents the elastoplastic constitutivematrix and is the strain vector. Among several avail-able options to integrate equation (15), Nazem et al.(2009b) showed that it is slightly more efficient toapply rigid body corrections during integration of theconstitutive equations. This strategy is adopted in thisstudy.

It is noted that the undrained shear strength of cohe-sive soils often increases with the rate of straining.This phenomenon, termed the strain rate effect, usu-ally introduces uncertainties in predicting the shearstrength of soils in dynamic penetration tests. Grahamet al. (1983) proposed a simple equation whichprescribes approximately 520% increase in shearstrength for each order of magnitude increase in the

6

rate of shear strain. This effect can be expressed by thefollowing widely used equation

where su is the undrained shear strength of the clay,su,ref is a reference undrained shear strength measuredat a reference strain rate of ref , denotes the shearstrain rate and is the rate of increase of strengthper log cycle of time. Typically ref = 0.01 per hourfor clays (Einav and Randolph 2006), and this valuewas adopted in the current study. For simplicity, theshear strength of a material can be assumed to beconstant while integrating the constitutive equationsduring an individual time-step of a nonlinear finiteelement analysis. However, the shear strength param-eters must be updated according to the known shearstrain rates at the end of each increment. Note that appearing in equation (16) should be evaluated usinga frame-independent quantity such as the maximumplastic shear strain.

In cone penetration tests it is found that the pen-etration resistance in clays increases as the rate ofpenetration (v) or the diameter of the penetrometerdecreases. In static tests this is mainly due to partialconsolidation occurring in advance of the cone tip,which in turn increases the effective stresses aroundthe cone followed by an increase of the sleeve fric-tion. Usually a normalised velocity,V , is used to assessthis degree of consolidation according to

where cv represents the coefficient of consolidationof the soil (e.g., Lehane et al. 2009). Values of thenormalised velocity between 10 (OLoughlin et al.2004) and 100 (Brown and Hyde, 2008) have beenrecommended to ensure that the soil behaviour isundrained. In the problems studied here the value ofnormalised velocity is generally above 10,000 dueto the fast penetration of the object (see Section 4),and the relatively short penetration time (0.10.9 s).Therefore, we assume undrained conditions throughoutthe penetration, and thus neglect the shear strengthincrease due to partial consolidation.

2.6 Mesh refinement

In the ALE operator split technique the material dis-placements are obtained at the end of the UL stepand will normally result in a distorted mesh. Refiningthe mesh at the beginning of each Euler step is veryimportant since a distorted mesh can lead to inaccurateresults. Most mesh refinement techniques are based onspecial mesh-generation algorithms, which must con-sider various parameters such as the dimensions of theproblem, the type of elements to be generated and theregularity of the domain. Developing such algorithmsfor any arbitrary domain is usually difficult and costly.

Moreover, these algorithms do not necessarily pre-serve the number of nodes and the number of elementsin a mesh and they may cause significant changes in thetopology. A robust method for mesh refinement basedon the use of a simple elastic analysis was presentedby Nazem et al. (2006). The method has been imple-mented for two-dimensional plane strain problems aswell as axi-symmetric problems. To obtain the meshdisplacements, we first re-discretise all the boundariesof the problem, which include the boundaries of eachdiscrete body, the material interfaces and the loadingboundaries, resulting in prescribed values of the meshdisplacements for the nodes on these boundaries. Eachboundary node is then relocated along the boundaryas necessary. With the known total displacements ofthese nodes on the boundaries, an elastic analysis isthen performed using the prescribed boundary dis-placements to obtain the mesh displacements for allthe internal nodes and hence the optimal new mesh.Animportant advantage of this mesh optimisation methodis its independence of element topology and problemdimensions. The method does not require any meshgeneration algorithm, does not change the topology ofthe problem, and hence can be easily implemented inexisting finite element codes. This method is poten-tially a good candidate for three-dimensional meshoptimisation.

The relocation of nodes along the boundaries playsa key role in the mesh refinement scheme proposed byNazem et al. (2006). This relocation demands an effi-cient mathematical representation of the boundarieswhich strongly depends on the physical problem beingsimulated. Nazem et al. (2008) proposed an efficientmethod for nodal relocation based on a quadratic splinetechnique.This method was successfully employed forsolving many geotechnical problems such as consol-idation and bearing capacity of soils under a footing.However, it is not suitable for simulating boundariesin a FFP problem due to the linear shape of the pen-etrating object, which consequently introduces abruptchanges in the slope of the straight lines defining theboundary. Instead, we introduce a simpler, but moreefficient, algorithm for nodal relocation along theboundary of the soil in contact with the penetrometer,which is based on the assumption of linear boundarysegments. This procedure is depicted schematically inFigure 2. The boundary between the soil and the pene-trating object at the beginning of a time step (Figure 2a)may become distorted after the UL step, as shown inFigure 2b. Such cumulative distortion, if not prevented,can potentially cause significant distortion of elementsin that region ultimately indicated by a negative Jaco-bian. Thus, nodal relocation on the boundary can beperformed as shown in Figure 2c. Note that the nodeson a curved boundary are only permitted to move in thetangential direction of the boundary, i.e., the normalcomponent of the convective velocity on a boundaryis zero, but not necessarily the tangential component.However, on a multi-linear boundary the nodes maybe allowed to move along a line defining a linear seg-ment.The nodal relocation on linear segments is shown

7

Figure 2. Nodal relocation on the boundary.

Figure 3. Linear nodal relocation.

schematically in Figure 3, where node i must be movedinto a new position, i, along line 1 or line 2 such thatl= li. Note that li represents a normalised length ofthe segment and can be calculated at the beginning ofa time step before the mesh is distorted. The Carte-sian coordinates of the new location can be calculatedexplicitly. For brevity, the full mathematical details arenot provided here but the procedure for relocating allnodes along a boundary is described in Nazem et al.(2012).

2.7 Energy absorbing boundaries

One of the well-known issues in computationaldynamic analysis of soil-structure-interaction (SSI)problems is how to simulate an infinite medium.Employing the finite element method, for instance,with boundaries that are not infinitely distant, one mustguarantee that the outgoing waves from the source(usually a structure) do not reflect back from the finiteboundaries toward the source, since in reality these

waves should propagate to infinity and dissipate at a fardistance from the source. If it is allowed to occur, suchreflection will most probably affect the accuracy ofthe numerical results. To assure no waves are reflectedback from truncated computational boundaries, it iscommon to use artificial boundaries which absorb theenergy of incom ing waves.

A simple, but efficient, boundary was developedby Lysmer and Kuhlemeyer (1969) which is knownas the standard viscous boundary in the litera-ture. The standard viscous boundary is probably themost popular artificial boundary since it possesses anacceptable dissipation characteristic at a low compu-tational cost. Kellezi (1998) suggests that absorbingboundaries must not be located closer than (1.21.5)s(where s is the length of the shear waves) from theexcitation source. A recent study by Kontoe (2006)showed that the standard viscous boundary is capableof absorbing dilatational waves (P-waves) as well asshear waves (S-waves) in the analysis of plane strainand axisymmetric problems. When required, this typeof boundary was used in the problems solved in thisstudy.

3 PENETRATION INTO NON-UNIFORM SOIL

In previous studies (e.g., Nazem et al. 2012), we con-sidered a smooth free-falling object penetrating into auniform soil layer, in which the material properties donot change in any direction. However, in many naturalsettings the shear strength of soils actually increaseswith depth. This variation can be expressed by a linearequation as

where su,ref (y) represents the reference undrainedshear strength of soil at depth y, ks indicates theincrease in shear strength per unit depth, and su,ref (0)is the reference shear strength of soil at the groundsurface.

The effect of the shear strength increase parame-ter ks, on soil penetration was studied by analyzing50 different problems with parameters summarized inTable 1. Note that in all cases the ratio between theshear modulus of the soil, G, and its undrained shearstrength, su,ref , is assumed to be 67 at all integrationpoints, and the rate parameter, , is 0.2. Dependingon the site, the rate of strength increase with depthks, usually varies between 13 kPa/m. Adopting thesetypical values for a free-falling penetrometer (FFP),the effect of ks on penetration characteristics will notbe significant, mainly due to the small diameter (size)of the penetrating object. Alternatively, a normalizedparameter, ksd/su,ref (0), is adopted in this study, whichquantifies the significance the rate of shear strengthincrease on the penetration.

The finite element mesh and boundary conditionsof this axi-symmetric problem are depicted in 4. Thefinite element mesh includes 10,252 nodal points

8

Figure 4. Finite element mesh of free falling penetrometer.

Table 1. Parameter values adopted in Section 3.

Unit Values

d m 0.04, 0.06,0.08, 0.1m kg 0.5, 1.0vo m/sec 5, 10su,ref (0) kPa 1.0G/su,ref 67L 0.2ksd/su,ref (0) 0, 0.25, 0.5, 1.0

and 4,988 6-node triangular elements each contain-ing 6 integration points. For simplicity, the materialdamping and friction between the soil and penetrom-eter are assumed to be zero. Note that the right and

Figure 5. Initial impact velocity of the penetrometer,normalised by 0.25d3su,ref (0), versus the total depth ofpenetration, normalized by d.

bottom boundaries are able to dissipate the energy ofoncoming waves.

In a previous study, Nazem et al. (2012) showedthat an appropriate strategy for studying the dynamicpenetration problem is to plot the initial impactvelocity of the penetrometer, 0.5mv20, normalised by0.25d3su,ref (0), versus the total depth of penetra-tion, normalized by d. Adopting this strategy, thenormalised impact energy of the penetrometer is plot-ted versus the normalised depth of penetration inFigure 5. According to Figure 5, the depth of penetra-tion of penetrometers with an identical impact energydecreases as the normalised parameter ksd/su,ref (0)increases. Figure 5 also shows that there is a linearrelation between the normalised depth of penetrationand the normalised impact energy, but the slope of suchlines depends on the value of ksd/su,ref (0). This depen-dency, for the specimens analysed in this example,can be eliminated by plotting the initial impact energy,normalised by 0.25d3su,ref (d), versus the normaliseddepth of penetration, as represented in Figure 6. Notethat su,ref (d) indicates the reference shear strength ofthe soil at depth d. Similar independencies for othermaterial properties and penetrometer parameters areyet to be investigated.

To study the deceleration characteristics of the pen-etrometer, its variation of velocity versus penetrationis plotted in Figure 7 for specimens where m= 0.5 kg,d = 0.04 m, v0 = 10 m/sec, and ksd/su,ref (0) = 0.0,0.25, 0.5, and 1.0. The plots in Figure 7 indicatethat the velocity of the object reduces approximatelyquadratically as it decelerates and penetrates furtherinto the non-uniform. Note that the quadratic reduc-tion of velocity with penetration has also been reportedin experimental studies as well as numerical analysisresults of penetration into a uniform clay layer (Nazemet al. 2012). It is also notable that for all cases studiedhere the total time of penetration is less than 0.3 sec.

9

Figure 6. Initial impact velocity of the penetrometer,normalised by 0.25d3su,ref (d), versus the total depth ofpenetration, normalized by d.

Figure 7. The velocity of the penetrometer versus thepenetration in diameter.

4 TORPEDO ANCHORS

An interesting application of dynamic penetrationproblems is the numerical analysis of installationof torpedo anchors, which have proved to be effi-cient for deep-water anchoring systems due to theireasy and economic installation. Torpedo anchorsare usually 1015 m long, 0.751.2 m in diameter,and may weigh between 2501,000 kN (Randolph &Gourvenec 2011).Assuming that the installation phaseis relatively fast, we consider only undrained soilbehaviour here. For simplicity, material damping andthe friction between the soil and anchor are ignored.The impact velocity of the anchor is assumed to be

Table 2. Values of parameters used in torpedo anchoranalysis.

Parameter Unit Values

d m 1.0m kg 1,000vo m/sec 5, 10su,ref (0) kPa 1.0G/su,ref 67 0.25, 0.5ksd/su,ref (0) 0, 1.0, 2.5

Table 3. Summary of numerical results for torpedo anchors.

ks m v0 p tpNo (kPa/m) (kg) (m/sec) (m) (sec)

1 0 0.25 2,000 5 5.34 1.852 0 0.25 2,000 10 14.34 2.893 1 0.25 2,000 5 1.96 0.534 1 0.25 2,000 10 3.19 0.505 1 0.25 5,000 5 4.15 0.936 1 0.25 5,000 10 6.07 0.837 1 0.25 10,000 5 7.67 1.438 1 0.25 10,000 10 10.24 1.269 2.5 0.25 2,000 5 1.49 0.3810 2.5 0.25 2,000 10 2.31 0.3511 2.5 0.25 5,000 5 2.69 0.6312 2.5 0.25 5,000 10 3.87 0.5413 2.5 0.25 10,000 5 4.24 0.8814 2.5 0.25 10,000 10 5.87 0.7715 0 0.5 2,000 5 2.21 0.7116 0 0.5 2,000 10 6.93 1.4117 1 0.5 2,000 10 2.46 0.3918 1 0.5 5,000 5 3.01 0.7419 1 0.5 5,000 10 4.71 0.6620 1 0.5 10,000 5 5.56 1.1621 1 0.5 10,000 10 7.93 1.0322 2.5 0.5 2,000 5 1.22 0.3123 2.5 0.5 2,000 10 1.85 0.2824 2.5 0.5 5,000 5 2.06 0.5025 2.5 0.5 5,000 10 3.11 0.4426 2.5 0.5 10,000 5 3.23 0.7227 2.5 0.5 10,000 10 4.66 0.62

less than its terminal velocity in a fluid, i.e., the tor-pedo may accelerate after impacting the soil due to thegravitational force.

To investigate the soil behaviour, 27 specimens areanalysed with the properties summarised in Table 2.The finite element mesh and the boundary conditionsare as those used in the previous example, and areshown in Figure 4.

The numerical results for all specimens, in termsof the total depth of penetration, p, and the total timeof penetration, tp, are summarised in Table 3. Accord-ing to Table 3 the total penetration time of a torpedoanchor is comparatively higher than the penetrationtime of a miniature FFP. In addition, the strength rateand shear strength increase with depth parameters,

10

Figure 8. The velocity of torpedo anchor versus time.

Figure 9. The velocity of torpedo anchor versus the pene-tration normalised by diameter.

and ks, significantly influence the total depth and timeof penetration.

To study the penetration properties of a torpedoanchor in more detail, the velocity variations of Spec-imens 17 and 23 are plotted versus time as well asthe normalised penetration in Figures 8 and 9, respec-tively. For these two specimens all input parameters,except ks, are identical.According to Table 3, the valueof ks in Specimens 17 and 23 are 1.0 and 2.5 kPa/m,respectively. Figures 8 and 9 indicate that the torpedoaccelerates at the early stages of penetration. Thisbehaviour was observed in all specimens, indicatingthat, unlike miniature FFPs, the acceleration of thetorpedo changes approximately linearly with time.

To study the soil behaviour due to torpedo pene-tration we consider two different cases where ks = 0(uniform soil) and ks > 0 (non-uniform). For a uniformsoil, Nazem et al. (2012) showed that the dynamic

Figure 10. Normalised dynamic soil resistance versus nor-malised penetration of torpedo into a uniform soil layer.

soil resistance due to penetration of a FFP can beestimated according to:

where

Using the values in Table 3, the dynamic penetra-tion factors, Ndp, for specimens 2 and 16 are estimatedas 31.75 and 53.01, respectively. On the other hand,the dynamic soil resistance versus the normalised pen-etration predicted by numerical analysis is plotted inFigure 10 for the same specimens.According to Figure10, the predicted normalised dynamic soil resistance isin good agreement with the value predicted by Equa-tion (19) for Specimen 2, and is slightly lower than53.01 for Specimen 16. This suggests that the valida-tion of Equations (19) and (20) for torpedo anchorspenetration into an uniform soil layer requires furtherinvestigation.

Unlike uniform soils, the dynamic resistance of anon-uniform layer of soil is not likely to convergeto a constant value as the normalised penetration,p/d, approaches infinity. For Specimens 8 and 14,where ks is respectively 1.0 and 2.5 kPa/m, the nor-malised dynamic soil resistance is plotted versus thenormalised penetration of the torpedo in Figure 11.It is evident that the soil resistance due to penetra-tion increases approximately linearly with penetration.This observation also indicates that the force actingon the anchor due to soil resistance changes with

11

Figure 11. Normalised dynamic soil resistance versus nor-malised penetration of torpedo into a non-uniform soillayer.

time as the object penetrates into soil, and hencethe deceleration of torpedo is not constant duringpenetration.

To represent the efficiency of the numerical methodin handling the mesh distortion, the finite elementmesh at the end of analysis of Specimen 8 is depictedin Figure 12.

5 CONCLUSIONS

The finite element solution of problems involvingdynamic penetration of soil deposits, by either freefalling or propelled objects, has been studied. Arobust numerical method was described for dealingwith such complex and difficult numerical problems.

The approach was based on the ArbitraryLagrangian-Eulerian (ALE) method of analysis andit allowed for features such as large deformations,rate dependent non-linear soil behaviour and mov-ing boundaries and contacts that are inherent in suchproblems.

The ALE method was employed to perform a para-metric study of a variety of penetrometers in layersof inhomogeneous clay deforming under undrainedconditions. The effect of the mechanical propertiesof the clay soil on the penetration characteristics hasbeen discussed. For smaller penetrometers it was foundthat the deceleration after impact was almost constantover most of the penetration. Higher initial energiesproduced deeper penetrations, as expected. Simpleclosed-form expressions were fitted to a sub-class ofthe results described.

For much larger objects penetrating typical softseabed deposits, such as torpedo anchors with a masson the order of 1 tonne, it was noticeable that the tor-pedo continued to accelerate under gravity just afterimpact during initial penetration, but ultimately cameto rest under deceleration that was not constant, butposition dependent.

Figure 12. Deformed mesh at the end of analysis,Specimen 8.

Overall, the efficiency and dependability of theproposed computation method has been demonstratedby analysing one of the most complicated, geo-metrically and materially non-linear, rate dependentproblems conceivable, at least within the realm ofgeotechnical engineering.

REFERENCES

Abelev,A., Simeonov, J. &Valent, P. 2009, Numerical investi-gation of dynamic free-fall penetrometers in soft cohesivemarine sediments using a finite element approach,Oceans2009, IEEE/MTS conference, Biloxi.

Beard, R.M. 1977, Expendable doppler penetrometer,Technical Report, Naval Construction Battalion Center,Port Hueneme, California.

Brown, M.J. & Hyde, A.F.L. 2008, High penetration rateCPT to determine damping parameters for rapid load

12

pile testing. 3rd International Conference of Geotechni-cal and Geophysical Site Characterisation, Taiwan. Pub.Balkema. 657663.

Chow, S.H. 2012. Rate effects in free-falling penetrometertests in clay. Forthcoming PhD Thesis, The University ofSydney, Australia.

Chung, J. & Hulbert, G.M. 1993, A time integration algo-rithm for structural dynamics with improved numeri-cal dissipation: the generalized-a method. Journal ofApplied Mechanics, 60, 3715.

Dayal, U. & Allen, J.H. 1973, Instrumental Impact ConePenetrometer. Canadian Geotechnical Journal, 10, 3,397409.

Einav, I. & Randolph, M. 2006, Effect of strain rate onmobilised strength and thickness of curved shear bandsGotechnique, 56, 501504.

Graham, J., Crooks, J.H.A. & Bell, A.L. 1983, Time effectson the stressstrain behaviour of natural soft clays.Gotechnique, 33, 327340.

Hu, Y. & Randolph. M.F. 1998, A practical numericalapproach for large deformation problems in soils. Inter-national Journal for Numerical and Analytical Methodsin Geomechanics, 22, 327350.

Kellezi, L. 1998, Dynamic soil-structure interaction. Trans-mitting boundary for transient analysis. PhD Thesis,Series R No. 50, Department of Structural Engineering& Materials, DTU, Denmark.

Kontoe, S., Zdrakovic, L. & Potts, D.M. 2008, An assess-ment of time integration schemes for dynamic geotech-nical problems, Computers and Geotechnics, DOI:10.1016/j.compgeo.2007.05.001.

Kontoe, S. 2006, Development of time integration schemesand advanced boundary conditions for dynamic geotech-nical analysis, PhD thesis, Imperial College, Universityof London.

Lehane, B.M., OLoughlin, C.D., Gaudin, C. & Randolph,M.F. 2009, Rate effects on penetrometer resistance inkaolin. Gotechnique, 59, 4152.

Lysmer J.R. & Kuhlemeyer L. 1969, Finite dynamic modelfor infinite media. Journal of the Engineering MechanicsDivision, ASCE, 95(EM4), 85977.

Medeiros, C.J. 2002, Low cost anchor system for flex-ible risers in deep waters. Proceedings of the 34thAnnual Offshore Technology Conference, Houston, Texas,Paper OTC 14151.

Mulhearn, P.J. 2002, Influence of penetrometer probe tipgeometry on bearing strength estimation for mine burialprediction, DSTO, TR 1285.

Nazem, M., Sheng, D. & Carter, J.P. 2006, Stress integra-tion and mesh refinement in numerical solutions to largedeformations in geomechanics. International Journal forNumerical Methods in Engineering, 65, 10021027.

Nazem, M., Sheng, D., Carter, J.P. & Sloan, S.W.2008, Arbitrary-Lagrangian-Eulerian method for large-deformation consolidation problems in geomechanics.International Journal for Analytical and NumericalMethods in Geomechanics, 32, 10231050.

Nazem, M., Carter, J.P. & Airey, D. 2009a, ArbitraryLagrangian-Eulerian Method for dynamic analysis ofGeotechnical Problems. Computers and Geotechnics, 36(4), 549557.

Nazem, M., Sheng, D., Carter, J.P. & Sloan, S.W. 2009b,Alternative stress-integration schemes for large deforma-tion problems of solid mechanics.Finite Elements inAnal-ysis and Design, 45, 934943.

Nazem, M., Carter, J.P., Airey, D.W. & Chow, S.H.2012, Dynamic analysis of a smooth penetrom-eter free-falling into uniform clay. Gotechnique,[http://dx.doi.org/10.1680 /geot.10.P.055].

OLoughlin, C.D., Randolph, M.F. & Richardson, M.D. 2004,Experimental and theoretical studies of deep penetratinganchors,Proceedings of OffshoreTechnology Conference,Houston, TX, USA.

Randolph, M. & Gourvence, S. 2011. Offshore GeotechnicalEngineering, New York, Spon Press.

Scott, R.F. 1970, In-place ocean soil strength by accelerom-eter, Proc ASCE, J. Soil Mechanics and Foundations, 96,1, 419444.

Sheng, D., Nazem, M. & Carter, J.P. 2009, Some computa-tional aspects for solving deep penetration problems ingeomechanics. Computational Mechanics, 44, 549561.

True, D.G. 1975, Penetration of Projectiles into Sea BedFloor Soils, Civil Engineering Laboratory, TechnicalReport R-822, Port Hueneme, CA.

Wriggers, P. 2006, Computational Contact Mechanics,Springer-Verlag, Austria.

13

This page intentionally left blankThis page intentionally left blank

From Materials to Structures: Advancement through Innovation Samali, Attard & Song (Eds) 2013 Taylor & Francis Group, London, ISBN 978-0-415-63318-5

Strategies for structural health monitoring of bridges:Japans experience and practice

Yozo Fujino & Dionysius SiringoringoThe University of Tokyo, Tokyo, Japan

ABSTRACT: The needs for structural health monitoring in Japan originally come from its geographical condi-tions such as severe environment for deterioration and frequent natural disasters. Monitoring of the environmentand loading conditions with respect to natural hazards and severe environmental conditions have been conductedfor several decades. In the last decade, bridge monitoring has extended its function as an instrument for efficientstock management. Accumulation of bridge stock built within the same period around three to four decades agomeans that many bridges in Japan are expected to have serious deterioration problems in the next few decades. Inthis paper, we describe the needs for bridge structural health monitoring in Japan and the concept of monitoringas an essential part of risk reduction. The paper outlines strategies implemented for bridge monitoring in Japan,which categorized into three main groups according to the purpose of monitoring: natural hazard and environ-ment condition, effective stock management, and failure prevention. In the end, the paper describes examples ofimplementation of the strategies and the lessons learned past from monitoring experiences.

1 INTRODUCTION

Reliability and durability of civil infrastructure is animportant factor for sustainable economic growth, pro-ductivity, and well-being of a nation. Factors such asconstruction defects, structural deterioration, materialdegradation and aging, harsh environmental condi-tions, changing and increasing loading, as well asextreme events such as natural disasters may con-tribute to the failure of infrastructure systems tovarious degrees, from non-optimal performance toa total breakdown. Realising the importance of asustainable and reliable infrastructure system amidthe constant threat of natural disasters and struc-tural degradation, development and implementation ofcivil infrastructure maintenance and monitoring pro-grams has attracted the interest of many researchers,universities and industries in Japan.

This paper describes the need for bridge monitoringand the strategies currently being implemented to meetthese demands. It also provides examples of currentpractices and research activities in bridge monitoringin Japan.

2 BACKGROUND OF BRIDGE MONITORINGIN JAPAN

Development of bridge monitoring in Japan has beenhighly influenced by geographical and socioeconomi-cal conditions in Japan. Geographically, Japan is proneto natural disasters such as earthquakes and typhoonsand severe environmental conditions such as strong

winds and high humidity. Natural disasters are majorconcern for civil engineering construction and main-tenance. At the same time, evaluation of structuralperformance and/or damage of existing bridges arebecoming more and more important for rational andefficient stock management.

2.1 Natural disaster and environmental conditions

Japan Ministry of Land, Infrastructure and Transport(MLIT) have identified several factors that are threat-ening the sustainability of infrastructure, and on thetop of the list is natural disaster. From 1970 to 2004,the total infrastructure loss due to natural disasters wasapproximately US$165 billion, or 15% of the worldstotal infrastructure losses caused by natural disasters.In 1995, the Kobe earthquake alone killed more than6000 people and heavily damaged major infrastructuresystems. The financial loss was estimated at aroundUS$131.5 billion. The earthquake caused extensivedamage to highway bridges in the national highwayand local expressways networks. Bridge collapses andnear-collapse were observed at nine sites, while otherextensive damage occurred in at least 16 sites. To dealwith a similar scale of earthquakes in the future, thebridge seismic design code was later revised. Con-struction of new bridges are subjected to the new code,while existing bridges are undergoing a seismic assess-ment and retrofit programme to meet the requirementsof the new code. The first phase of the programmefocused mainly on retrofits of ordinary bridges withemphasis on bridge piers and girder restrainers. Inthe second phase, many large bridges are currently

15

being retrofitted. Fujino et al. [1], for instance, outlinesthe retrofit plan/design of three large cable-supportedbridges in Tokyo.

Another geographical aspect that influences Japansinfrastructure, and bridges in particular, is the coun-trys topography. As an archipelago country, Japanconsists of mountainous islands with the populationmainly concentrated near the seashore. Therefore,transportation infrastructures including many bridgesare built near the coastline, and some cross the chan-nels, such as the worlds longest bridge, the Akashi-Kaikyo Bridge. Due to this condition, some bridgesare situated in areas of severe environmental condi-tion and subjected to deterioration caused by highchloride intrusion and humidity. Consequently, main-tenance and monitoring systems must be employed tocontrol the environmental effects and to prevent furtherdeterioration.

On average, Japan spends around US$2.5 to 3.5billion annually for research activities related to dis-aster mitigations. This includes developing compre-hensive hazard monitoring systems nationwide. Forseismic hazard monitoring, the nationwide system(K-Net)[2] that constitutes a network of 1034 wide-band seismographs installed on the ground surfacewas developed. To monitor atmospheric related haz-ards, a high-resolution surface monitoring system(AMeDAS) involving wind, temperature and othermeteorological data was developed by the JapanMeteorological Agency (JMA).

2.2 Bridge stock management: condition andchallenge

The peak of Japans infrastructure developmentoccurred between 1955 and 1975. During this time,rapid economy growth accelerated public spendingon infrastructure, and as a result, Japans infrastruc-ture stocks have increased significantly. Infrastructureproblems have compounded ever since, for severalreasons such as failure to recognise the importanceof maintenance, underinvestment in maintenance, andlack of good management system.

The rush to expand infrastructure during the rapideconomic growth has led to cost cutting that some-times create problems in structural durability. Bridgesprovide an excellent example of the problem. Deteri-oration is an issue for many highway bridges in Japan,which were mainly built around the 1970s. Aware-ness of the current condition of bridges has emergedrecently, triggered by the discovery of steel memberfractures in the Kisogawa Ohashi Bridge, Mie Prefec-ture, in June 2007 and in the Honjo Ohashi Bridge,Akita Prefecture, in August 2007. The collapse of theI-35W interstate bridge in Minnesota, United States,which followed later in August 2007, has only inten-sified the concern, based on the similarity of the twobridges in Japan with the one that collapsed. Manybelieve these two incidents are the tip of the iceberg ofthe problem. Figure 1 shows the comparison of bridgeconstruction between Japan and the U.S. [3]. It can be

Figure 1. Comparison of period of bridge constructionbetween Japan and the United States.

seen that on average, Japanese bridges are 10 yearsyounger, but the construction period is more concen-trated. It is anticipated that older bridges will constitutehalf of all road bridges by the year 2020.

This concentration may induce simultaneous dete-rioration of large portions of the stock that would leadto a high social cost. Adding to the problem of ageingbridges is the drastic increase in traffic density. Thedaily traffic volume on major highways can reach ashigh as 15,000 vehicles per lane, of which the ratio ofheavy trucks is also high, in some cases 30% [3]. Thehigh intensity and frequency of loading generated bythis high traffic volume generates many problems inbridges. The most typical problem is fatigue damagein concrete slabs and steel welded girders.

3 CONCEPT OF MONITORING FOR RISKREDUCTION

The concept of bridge monitoring or structural healthmonitoring in general can be viewed as an integratedstrategy for risk reduction. In this context, risk isdefined as a function of two quantities: hazard andstructure vulnerability, in the form of:

Hazard, defined as the probability of occurrence ofunexpected events that may endanger the structure,is spatially and temporally dependent. Conventionalpractice assumes that the structural vulnerability istime-invariant, and a function of the structural config-uration such as system redundancy, ductility, materi-als, and quality of co