Optimum Design of Multilayer Asphalt Surfacing Systems for Orthotropic Steel Deck Bridges

(1)

(2)

(3)

Optimum Design of Multilayer Asphalt

Surfacing Systems for Orthotropic Steel Deck

(4)

The research described in this thesis was performed in the section of Road and Railway Engineering, Faculty of Civil Engineering and Geosciences, Delft University of Technology, Stevinweg 1, 2628 CN Delft, The Netherlands. This work was supported by the Dutch Transport Research Centre (DVS) of the Ministry of Transport, Public Works and Water Management (RWS).

(5)

Optimum Design of Multilayer Asphalt

Surfacing Systems for Orthotropic Steel Deck

Bridges

PROEFSCHRIFT

ter verkrijging van de graad van doctor aan de Technische Universiteit Delft,

op gezag van de Rector Magnificus prof. Ir. K.C.A.M. Luyben, voorzitter van het College voor Promoties,

in het openbaar te verdedigen op 14 september 2015 om 10:00 uur

Door

Jinlong LI

Master of Engineering

Nanjing University of Science & Technology, China geboren te Jiangxi, China

(6)

This dissertation has been approved by Promotor: Prof. dr. A. Scarpas

Copromotor: Dr. X. Liu

Composition of the doctoral committee:

Rector Magnificus Delft University of Technology, Netherlands, chairperson Prof. dr. A. Scarpas Delft University of Technology, Netherlands, promotor Dr. X. Liu Delft University of Technology, Netherlands, copromotor

Independent members:

Prof. dr. ir. S.M.J.G. Erkens Delft University of Technology, Netherlands Prof. dr. I. L. Al-Qadi University of Illinois at Urbana-Champaign, USA Prof. dr.-ing. habil. M. Oeser RWTH Aachen University, Germany

Prof. dr. J. Yang Southeast University, China

Reserve member:

Prof. dr. ir. E. Schlangen Delft University of Technology, Netherlands

Published and distributed by: Jinlong Li

Email: jinlong.li@hotmail.com

Printed by CPI Koninklijke Wöhrmann ISBN 978-94-6203-874-5

All rights reserved. No part of the material protected by this copyright notice may be reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording or by any information storage and retrieval system, without written consent from the publisher.

(7)

(8)

(9)

Acknowledgements

This research project was funded by the Dutch Transport Research Centre (DVS) of the Ministry of Transport, Public Works and Water Management (RWS). The China Scholarship Council (CSC) sponsored my living expenses in the Netherlands during the first four years. Their financial support is highly appreciated.

First, I would like to thank Prof. J.G. Rots, director of Structural Engineering in the Faculty of Civil Engineering and Geosciences, for providing the necessary infrastructure for this thesis. I am especially grateful to my promoter Prof. Tom Scarpas who not only had helped me to get the opportunity to carry out this research here, but also encouraged, supported and advised me during this study. My sincere gratitude to Dr. Xueyan Liu, my daily supervisor, who has given me ample support, guidance and encouragement in all the key moments during my PhD.

I wish to express my sincere thanks to my colleague Ir. Cor Kasbergen for his great help to make the computer code computationally efficient, and to Ir. George Tzimiris for the excellent experimental results. I also have to acknowledge my other colleagues Alex, Santosh, Anupam, Alieh, Katerina, Stavros, Sayeda, Mirella, Fani and Tom J, for their friendship that encouraged an active research environment. I would feel much lonelier in the Netherlands without them.

I am indebted to Professor Chen Shiming, my promoter in Tongji University, for the vision and virtue which encouraged me to study abroad.

I would like to express my gratitude to all my other Chinese friends for so many memorable weekend parties and journeys.

I am particularly grateful to my beloved parents for their love, understanding, support and endurance throughout these years. Furthermore, I would like to thank my brother and sister in China, giving me the support and freedom to find my own way. My heartfelt appreciation goes to my wife who is a sea away from me, which enables me to focus on my research work. Certainly, this thesis is dedicated to them.

There are also many other people who supported and helped me directly or indirectly during my study in the Netherlands. I wish to express my deepest gratitude to all of them that are not named in this short page.

(10)

List of Symbols

Latin and Greek

υ Poisson’s ratio ε strain σ stress τ shear stress δ displacement Δ amplitude

C displacement constraint coefficient matrix

G strain energy release rate

H height

I1 first stress invariant of material constitutive model

J2 second stress invariant of material constitutive model

K stress intensity factor

K tangent global stiffness matrix

L local to global axis transformation matrix

P load

Pa atmospheric pressure

R radius

R force constraint coefficient matrix

T traction stress/force

θ angle

Subscript and superscript

a, ad adhesive eq equivalent max maximum min minimum n normal t tension

x,y,z axis direction

Abbreviations

MAT Membrane adhesion test SLBT Shaft loaded blister test MA Mastic asphalt

DAC Dense asphalt concrete 3D Three dimensional 5PB Five-point bending FE Finite element

(15)

FEA Finite element analysis GA Guss asphalt MB Membrane PA Porous asphalt ST Steel deck UM Upper membrane BM Bottom membrane

OSDB Orthotropic steel deck bridge SMA Stone mastic asphalt

(16)

(17)

1 Introduction

The word ‘orthotropic’, which is a combination of the words ‘orthogonal’ and ‘anisotropic’, was first used by German engineers in the 1950s. Therefore, an orthotropic deck has anisotropic structural properties at 90°. It consists of a deck plate supported in mutually perpendicular directions by a system of transverse crossbeams and longitudinal stiffeners. It is therefore a plate with dissimilar elastic properties in two directions. Structural steel is used mostly, although other metals such as aluminium or advanced composite materials can be used.

Since the first orthotropic steel deck bridge (OSDB) was opened in 1950 over the Neckar River in Mannheim, Germany, the OSDB has become a popular economical alternative when the following issues are important: lower mass, ductility, thinner or shallower sections, rapid bridge installation, and cold-weather construction (Gurney, 1992). Lower superstructure dead mass is the primary reason for the use of orthotropic decks in long-span bridges. Very thin decks can be built using the orthotropic system. In the Netherlands, the first orthotropic steel bridges were the Hartel Bridge and the Harmsen Bridge opened in 1968. Nowadays more than 1000 orthotropic steel bridges have been built in Europe, out of which 86 are in the Netherlands. OSDBs are also popular in Asia and especially in China and Japan. The application of OSDBs is developing fast in China (Wang, 2009).

There are different types of OSDBs, most of them varying in the type of longitudinal stiffeners (Medani, 2006). There are two types of longitudinal stiffeners: open stiffeners such as flat bars, angles and bulb sections and closed stiffeners with a trapezoidal, V or rounded form (Gurney, 1992). Two typical layouts of orthotropic bridges are shown in Figure 1.1. Because the closed stiffeners have higher torsional and bending stiffness and provide better mechanical performance of steel bridges, the closed stiffer type is more recommended in recent decades.

In the Netherlands, an asphaltic surfacing structure for orthotropic steel bridge decks mostly consists of two structural layers. The upper layer consists of porous asphalt (PA) because of reasons related to noise reduction. For the lower layer, a choice between mastic asphalt (MA), or guss asphalt (GA), can be made. Mostly, various membrane layers are involved, functioning as bonding layer, isolation layer as well as adhesion layer.

The asphalt surfacing structure for OSDBs is a complicated and yet not properly solved technical problem. The high flexibility and large local deformations, wind and earthquake forces, temperatures and other natural factors make the problem even more complicated. Due to the special characteristics of OSDBs, fatigue cracking, rutting, delaminating and other damage types are commonly reported and these severely destroy the performance of steel bridges. Obviously, it is difficult to solve the problem of asphalt surfacing structures for OSDBs with

(18)

a simple or traditional method. Research on mechanical and structural design of the surfacing systems of OSDBs is urgent.

Figure 1.1 Two basic layouts of an OSDB with two-layer surfacing

1.1 Problems on the surfacing system of orthotropic steel

deck bridges

In the last three decades, several problems were reported in relation to asphaltic surfacing materials on OSDBs such as rutting, cracking, loss of bond between the surfacing material and the steel plate, Figure 1.2. The severity of the problems is increased by the considerable growth of traffic in terms of number of trucks, heavier wheel loads, wide-base tires etc. The various types of damage

(19)

identified by researchers in relation to steel deck bridges are mentioned in the followings.

Fatigue cracks Shear debonding

Rutting Other damages

Figure 1.2 Structural- dependent cracks in the surfacing on steel bridge

1.1.1 Fatigue cracking

Field evidence indicates that fatigue damage in the various surfacing layers is an important type of damage to steel deck surfacing structures (NPC, 1996)(Nishizawa et al., 2004). In other cases, fatigue problems have been identified at locations where the steel deck plate is supported by the underlying steel structure. All in all, fatigue damage in both the surfacing structure and the underlying steel deck is the most often reported type of damage in relation to steel deck bridges and their surfacing structure.

1.1.2 Shear debonding

Another frequent failure is shear failure of the adhesive layers between the steel deck and asphalt. A survey of the performance of 12 OSDBs in the USA mentions that the first problems that engineers are called to address is proper bonding between steel and surfacing layers (Touran, 1991). According to repairing practice, asphalt the alligators or potholes will lose bonding as water can penetrate down to the steel deck plate. Experiences also indicate that large area of the bridge surfacing will soon fail once the deck pavement system loses bond strength at interfaces.

(20)

Adhesive membrane overloading, causing immediate failure of the membrane when subjected to high shear stresses, was mentioned as a type of damage that occurs when adhesive membranes, normally applied for roofing purposes, are utilized. According to one researcher, this type of damage occurred in at least one larger bridge in the Netherlands (Huurman, 2008).

1.1.3 Rutting

Rutting in asphalt concrete surfacing layers is also a type of damage that is reported by a larger contingent of researchers. Although rutting is often encountered, it appears to be of less importance than shear debonding and fatigue cracking.

1.1.4 Other damages

Apart from the types mentioned in the above, various other types of damage have been reported so far, such as moisture damage and shoving of the surfacing structure.

Because of the rapid development of OSDBs all over the world, further and deeper study of the mechanical properties of the surfacing systems of steel bridges, and the search for a proper design method for surfacing systems of OSDBs, are of important theoretical significance and great practical value.

1.2 Merwedebrug - bridge of concern

The Merwedebrug (Figure 1.3), part of Highway A27 near Gorinchem, the Netherlands, was opened on March 15 1961. It has been playing a very important role in connecting the Randstad and North Brabant, and it is representative due to heavy load traffic every day. Our initial proposed research program came from the need of a surfacing structure for that bridge with prolonged service life.

(21)

Until 1994, the Merwedebrug bridge deck was surfaced with a two layer deck. The lower layer was Guss Asphalt (GA). The upper layer consisted of either GA or Dense Asphalt Concrete (DAC). A membrane was applied between the steel and the GA layer. It is not certain whether a membrane was also applied between the lower and upper asphalt layer. These decks on average required reconstruction every six years. Between reconstructions smaller repairs were frequently required to keep these structures operational (Huurman, 2010).

After research in 1994 a new type of surfacing structure was applied: membrane, GA, membrane, polymer modified Porous Asphalt, PA. Within half year this structure developed ravelling. With intensive repair work, the service life of the structure was stretched till 2000. According to experts, the additional cause of the poor performance was the 10 mm deck plate which is too thin for the current traffic.

In 2000 damage became so severe that the upper membrane and the PA were removed and replaced by PA 0/8. In 2001, the first damage appeared again. In 2003, the damage became too severe. Several suggestions for reconstruction were made. At the end the surfacing was reconstructed as follows.

 Right lane west side: PA repair by application of a PA repair system.  Right lane east side: replacement of PA by SBS modified PA 0/8.

After these repairs, in 2005, the PA surface layer was replaced with DAC. The applied structure consisted of a lower membrane, GA, an upper membrane and a surface layer of DAC. This structure performed poorly, mainly suffered from cracking and was kept open for traffic by a process of continuous repairs without achieving any improvement in performance. As time progressed the situation of the surfacing structure became worse. In 2009, it was decided that the structure required reconstruction.

Prior to the reconstruction works it was estimated that about 10% of GA needed to be replaced. During the works, which were performed during the night, it became apparent on May 15 2009, that the lower membrane had lost its bond over the larger part of the area to be reconstructed. Cracking and alligator cracking already indicated poor performance caused by poor membrane adhesion in the existing structure.

As expected, this structure performed very poorly. After one year of service, the structure required replacement of the right lane. Reconstruction works were performed at the end of May 2010. Some pictures made during the reconstruction works are shown below.

An investigation into the debris of removed pavement of the Merwedebrug (Figure 1.4) indicated an insufficient adhesion between the upper and lower asphalt layer, and clearly showed a poor bonding between the steel deck and pavement structure.

(22)

Figure 1.4 Removed pavement debris of the Merwedebrug

Localized surface damage was also found on the Merwedebrug, as shown in Figure 1.5. This damage developed in a period of one year under traffic indicating a structure that was not properly bonded to the steel deck.

Figure 1.5 Localized surface damages on Merwedebrug

These observations indicate that a surfacing structure with enhanced composite action, i.e. balance membrane stiffness/strength, may well lead to surfacing structures with prolonged design life. Therefore, determination of the membrane bonding strength becomes important, and this is also the goal for the current project.

1.3 Project descriptions

The Ministry of Transport, Public Works and Water Management (RWS) in the Netherlands is facing a growing challenge in maintaining its network capacity. Even though the combined length of OSDBs in the primary road network is limited, the consequences of repairs of the steel deck plate or the overlaying surfacing structure to network capacity are dramatic. Unfortunately, the service life of asphaltic surfacing structures on OSDBs is limited to an average of 5 years.

(23)

Asphalt concrete surfacing structures have distinct advantages when compared to alternative surfacing structures: fast installation, good driving comfort, relatively cheap, and homogeneity in the road surface. This implies that alternative surfacing structures, for example the high strength concrete surfacing structures, are only applicable in special circumstances.

In conclusion, improvement of the performance of asphaltic surfacing structures on OSDBs is of the utmost importance.

Therefore, the Transport Research Centre (DVS) of RWS commissioned the Delft University of Technology to investigate the membrane performance. Earlier investigations have shown that the bonding strength of membrane layers to the surrounding materials has a strong influence on the structural response of orthotropic steel bridge decks. The most important requirement for the application of membrane materials on orthotropic steel bridge decks is that the membrane adhesive layer has to be able to provide sufficient bond to the surrounding materials.

In order to obtain insight into the response of membranes and their interaction with surrounding materials on orthotropic steel decks, a project of evaluation of the performance of modern surfacing systems on OSDBs has been undertaken. The focus is on membrane performance and the effect hereof on the structure as a whole.

The research was carried out in three stages at different scales named the material scale, the section scale and the bridge scale respectively. Each research stage is aimed to accomplish an optimization selection scheme at that scale. Through these three stages, a bottom-to-top optimum design and evaluation method of multilayer surfacing systems for OSDBs was established, Figure 1.6.

Figure 1.6 The bottom-to-top optimum design method with three stages

In the first stage, the Membrane Adhesion Test (MAT) for testing the membrane debonding strength was developed. Within this thesis, emphasis was placed on the theoretical background of MAT as well as FE simulations of the test. The experimental work of the project was carried out by another PhD student of our group. For illustration and verification purpose, only some important and representative test results were presented in this thesis. For elaborate test results readers are refered to the project report (X. Liu, T. Scarpas, 2012). The

Bridge scale

Section scale

(24)

response of each membrane under different temperatures was compared, and meanwhile, the response of different membranes under the same condition was also compared. The best performing membranes are recommended through membrane ranking for the further investigation tests in research stage 2 and 3. In Phase 2 of the project, four typical Dutch multilayer surfacing systems constructed with the five selected membrane products from Phase 1 were studied by means of five-point bending (5PB) beam tests and FE simulations. The findings of the 5PB beam tests will help us for the verification and calibration of the finite element predictions and for the further ranking of the best performance of the multilayer surfacing systems for Dutch OSDBs. In order to study the influence of the geometrical and structural parameters on the performance of the multilayer surfacing system, finite element simulations of 5PB beam tests were performed. The parametric studies were performed by means of the finite element system CAPA-3D developed at the Section of Structural Mechanics of TU Delft. The contribution to the overall system response of the mechanical properties of all the surfacing layers was studied. In the last stage of the work, Phase 3, finite element (FE) simulations of the Merwede bridge subjected to dual wheel stationary and moving loads are presented. Three cases of load locations have been investigated. All cases were simulated under 10 C and



5 C

respectively. Four different surfacing structures, as utilized in the 5PB beam tests were chosen for the FE simulations. The results of the simulations provide a useful guidance regarding the expected maximum strains in the four surfacing structures for two different temperatures.

1.4 Outline of the dissertation

In this dissertation, a literature review is introduced in Chapter 2. In this chapter, the commonly used methods for testing membrane debonding strength are briefly introduced.

The research work basically contains three stages, which consist of three study phases at different scales: 1) material scale, 2) section scale and 3) bridge scale. The work of Phase 1 is presented in Chapter 3, 4 and 5.

In Chapter 3, details of the theoretical backgroud of the Membrane Adhesion Test (MAT) developed for the project are presented. The MAT is used for evaluating the adhesive bonding strength of membranes with surrounding materials on orthotropic steel bridge decks.

In Chapter 4, the formulation of a specially developed contact element is provided. This element is utilized to model the adhesive bonding strength between the membrane layer and its surrounding materials. An adhesive traction-separation contact law is implemented in to the contact element in order to properly simulate the bonding effect between two material layers.

In Chapter 5, the experimental and FE simulation results of MAT are presented. The 3D finite element system CAPA-3D has been utilized as the numerical platform for this study.

(25)

Chapter 6 presents the research work of Phase 2. Four typical Dutch multilayer surfacing systems constructed with the five selected membrane products from Phase 1 were studied by means of five-point bending (5PB) beam tests and FE simulations. The findings of the 5PB beam tests will help us with the verification and calibration of the finite element predictions and for the further ranking of the best performance of the multilayer surfacing systems for Dutch OSDBs. In order to study the influence of the geometrical and structural parameters on the performance of the multilayer surfacing system, finite element simulations of 5PB beam tests were performed.

The last stage of the work, Phase 3, is reported in Chapter 7. Finite element simulations of the Merwede bridge subjected to dual wheel stationary and moving loads are presented. Three cases of load locations have been investigated. All cases were simulated under 10 o_{C and -5}o_{C respectively. Four different}

surfacing structures as utilized in the 5PB beam tests were chosen for the FE simulations. The results of the simulations provide a useful guidance regarding the expected maximum strains in the four surfacing structures for two different temperatures.

Conclusions and recommendations are described in Chapter 8.

(26)

(27)

2 Multilayer Surfacing Systems on Orthotropic

Steel Deck Bridges

In previous chapter, the basic definition as well as a brief historical background of orthotropic steel deck bridges (OSDBs) are already introduced. As it is mentioned in Chapter 1, commonly utilized multilayer surfacing structures on OSDBs in the Netherlands have drawn increasing attention because of the frequent appearance of damage. In this chapter, several key issues related to multilayer surfacing system on OSDBs are reviewed and discussed, including:

 behaviour of involved materials;

 design methods of surfacing structures;

 laboratory and field experimental tests on adhesive membranes;

 application of fracture mechanics theory for evaluation of membrane materials.

2.1 Bridge deck surfacing systems

An OSDB is an efficient structural system in terms of rigidity, strength requirements, economic consideration and self-weight reduction. The steel deck part of an OSDB typically consists of a complex network of longitudinal stiffeners, transverse stiffeners or crossbeams, and the deck plate itself. The steel deck takes part in the structural resistance of the bridge, which results in an extremely lightweight and durable bridge deck concept (Zhang et al., 2012). An overlaying surfacing structure is essential for three main purposes (Gurney, 1992):

 Providing a running surface with suitable skid resistance and reduced noise emissions;

 Providing a flat running surface by varying its thickness to compensate for deflection of the steel deck plate;

 Protecting the steel deck plate by providing a waterproofing layer.

It is difficult to fulfil all of these functions by one surfacing material layer (Medani, 2006). A typical deck surfacing system consists of at least the following layers: a bonding layer, an isolation layer, an adhesion layer and a wearing course. The aforementioned functions are not always clearly defined. In some cases, one layer may achieve several functions or several layers fulfil the same goal. The specific requirements of the various surfacing layers are discussed comprehensively by Medani (2006).

Because of regional differences, such as annual average temperatures, amounts of rainfall, traffic volumes, availability of raw materials etc., different countries use different surfacing materials and different layer thicknesses. As those surfacing structures are all composed of multiple material layers, hereafter, the term ‘multilayer surfacing systems’ is utilized.

(28)

For this reason, in the framework of this project, an extensive investigation was carried out with the goal of acquiring a thorough understanding of the causes of damage in OSDBs.

2.2 Characteristics of surfacing systems on orthotropic steel

deck bridges

Bridge surfacing is essential in order to protect the bridge deck, prolong its service life, improve the driving comfort and safety, reduce vibrations and noise level, etc. There are two main types of surfacings based on paving materials: the cement concrete surfacing and the asphalt concrete surfacing. Because of advantages such as lighter weight, better deformation coordination and adhesion to bridge decks, ease of maintenance and repair and higher driving comfort, the asphalt concrete surfacing type is adopted mostly on OSDBs (Huang, Zhang, & Hu, 2002).

From the perspective of paving materials and construction methods, there are mainly four types of asphalt bridge surfacing materials utilized at present around the world (Zhen, 2010):

1) High-temperature mixing guss asphalt surfacings, commonly applied in Germany and Japan;

2) Mastic asphalt concrete surfacings, widely used in the United Kingdom; 3) Modified asphalt SMA surfacings, recently adopted in Germany and Japan; 4) Epoxy asphalt surfacings, widely accepted in the US and applied a lot in recent years in China.

In terms of layered structure types, there are three surfacing systems currently: single-layer system, layer system with the same material and double-layer system with different materials (Huang & Liu, 2005). A typical multidouble-layer surfacing system on OSDBs consists of the following structural layers: an optional overlay layer, top layer, under layer, upper and bottom interlayers and the steel deck plate, Figure 2.1. These layers should work together to provide a proper running surface for traffic.

(29)

Table 2.1 lists some common options for these structural layers, as well as their main functions. The listed division of functions is not always applicable because sometimes one layer can perform several functions.

Table 2.1 Means and functions of surfacing layers at present Structural

layers Means of application Functions Overlay SMA with small gradation, oepn graded friction course, sprayed material, e.g.

rubber bitumen sprayed with grit, ZOAB

For renewal or preservation, functional improvements such

as to provide safety and smoothness, to resist rutting

and reduce noise, etc. Top layer

Porous asphalt, mastic asphalt, guss asphalt, open synthetic wearing course, asphalt concrete, epoxy asphalt, rubber

modified asphalt, SMA, etc.

Skid resistance, flat surface, low sound levels, traffic loads spreading, fatigue resistance,

good stability

Upper interlayer

Preformed membrane, rubberized asphalt, bituminous membrane, polymeric membrane, liquid sprayed membrane with or without reinforcing

fabric, tar-epoxy sprayed with grit, rubber bitumen sprayed with grit, binder,

epoxy bitumen bonding, etc.

protection against corrosion, water proofing, sufficient bonding, shear resistance, good isolation of oil, water, minerals

and de-icing salts, fatigue resistance

Under

layer Mastic asphalt, guss asphalt, asphalt concrete, SMA, epoxy asphalt, etc.

to take and spread traffic loads, to prevent penetration of

oil, water, de-icing salts and minerals, to resist fatigue

cracking

Bottom interlayer

Preformed membrane, rubberized asphalt, bituminous membrane, polymeric membrane, liquid sprayed membrane with or without reinforcing

fabric, tar-epoxy sprayed with grit, rubber bitumen sprayed with grit, binder,

epoxy bitumen bonding, etc.

protection against corrosion, water proofing, sufficient bonding, shear resistance, good isolation of oil, water, minerals

and de-icing salts, fatigue resistance

Steel deck 10-18 mm steel plate

to provide a flat surface for bridge paving, to carry traffic

loads and dead weights, fatigue resistance

Within recently twenty years, more than twenty long-spanned steel bridges with orthotropic deck plates have been constructed to meet the requirements of repaid economy developments and travel demands. The deck system normally consists of a deck plate, 12-16 mm thick and longitudianl trapezoidal or rounded closed ribs or open stiffeners, supported with cross beams spaced from 3.2 to 4.5 m apart. Table 2.2 presents some well-known long-span OSDBs around the world, and the surfacing types (Hu, 2005; Zhen, 2010) .

(30)

Table 2.2 Some well-known long-span OSDBs and their surfacing systems OSDBs Construct_{ion year} Surfacing

Golden Gate Bridge (US) 1937 Concrete, resurfaced with epoxy asphalt _{mixture in 1986} Verrazano Narrows Bridge (US) 1964 Double layers 50mm epoxy asphalt _mixture

Humber Bridge (UK) 1981 SMA

Pont de Normandie Bridge (FR) 1995 Double layer SMA Qingma Bridge (HK) 1997 50mm SMA

Humen Bridge (CN) 1997 55-60mm double layer modified SMA13 Höga Kusten Bridge (SE) 1997 Double layer SMA

Akashi Kaikyō Bridge (JP) 1998 35mm guss asphalt+30mm modified _{asphalt mixture} Great Belt Bridge (DK) 1998 SMA

Jiangyin Yangtze River Bridge

(CN) 1999 SMA, resurfaced by 50mm guss asphalt in 2003 Haicang Bridge (CN) 1999 35mm SMA10+30mm SMA13

Tataro Bridge (JP) 1999 35mm guss asphalt+30mm modified _{dense gradation SMA} Shantou Queshi Bridge (CN) 1999 50mm SMA13+35mm SMA13

Nanjing 2nd Yangtse River

Bridge (CN) 2000 50mm epoxy asphalt mixture Wuhan Baishazhou Bridge (CN) 2000 45mm SMA13+35mm SMA10

Runyang Yangtze River Bridge

(CN) 2005 50mm double layer epoxy asphalt mixture Foshan Pingsheng Bridge (CN) 2006 50mm double layer epoxy asphalt _mixture

Sutong Bridge (CN) 2007 Epoxy asphalt Hangzhou Bay Bridge (CN) 2007 Epoxy asphalt Xihoumen Bridge (CN) 2008 Epoxy asphalt

Huangpu Bridge (CN) 2008 60mm double layer epoxy asphalt _mixture Thuan Phuoc Bridge (Viet Nam) 2009 Epoxy asphalt

BaLinghe Bridge (CN) 2009 Epoxy asphalt Jiaozhou Bay Bridge (CN) 2010 Epoxy asphalt E-Dong Yangze River Bridge

(CN) 2010 Epoxy asphalt

Chongqi Bridge (CN) 2011 Epoxy asphalt

A road surfacing is applied on some semi-rigid roadbed, while the bridge surfacing is laid directly onto a relatively flexible steel deck plate. Because of the flexibility of the OSD plate, with the effects of traffic loading, wind, temperature fluctuation and earthquakes etc., high localized stresses and complex deformations can occur in steel deck plates. Severe stress

(31)

concentrations are observed at welds between ribs and steel plates, and also between crossbeams and steel plates. Those factors will affect the bridge surfacing. Thus, higher standards for material strength, stability, durability and fatigue resistance are required for bridge surfacing systems. Together with other critical issues such as high temperatures of steel decks during summer, water proofing , rust protection and good adhesion between different surfacing layers, we can summarize the characteristics of bridge surfacing systems as follows:

1) The bridge surfacing is directly paved on the orthotropic steel deck. The deformations, displacements and vibrations of orthotropic steel bridges have direct impact on the behaviour of the bridge surfacing and request good adaptability of deformations from it.

2) Steel bridges have good thermal conductivity and could be easily affected by the ambient temperature. In addition to the temperature deformation of the surfacing layer itself, daily and seasonal changes in temperature will also affect the deformation of steel bridge decks.

3) The bridge surfacing should have excellent creep and shearing resistance at high temperatures.

4) Because of the supporting stiffeners, negative bending moments occur on top of stiffening ribs, crossbeams and longitudinal diaphragms under the vehicle load. The maximum tensile stress/strain appears at the top of the bridge surfacing. Fatigue cracks start from the surfacing top and develop to the bottom. Good resistance to fatigue cracking is of importance for surfacing materials. 5) The steel bridge surfacing should have sufficient density in order to prevent seepage of water which results in steel corrosion.

6) Sufficient bonding performance between structural layers is critical for multilayer surfacing systems on OSDBs.

7) OSDBs are usually important traffic hubs, or vital connections crossing rivers, lakes and seas. Maintenance work may hinder the whole traffic network. Therefore, it is crucial to prolong the service life of bridge surfacings.

2.3 Research work on surfacing materials

The majority of OSDBs around the world, are paved with asphalt mixtures. Asphalt surfacings are light and have good performance. Creep properties, the influence of temperature and fatigue properties (including reflection cracks) are some important issues associated with the durability of asphalt bridge surfacing systems. Earlier study on the performance of asphalt mixtures was focused on general characteristics of bitumen and asphalt mixtures.

Van Der Poel (1954) first brought out the definition of “stiffness modulus” in order to distinguish with the elastic modulus, because the rheological behaviour of bitumen is sensitive to loading time and temperatures.

Based on its 20 years of laboratory experience, the Shell Company worked out a method for testing the modulus of asphalt mixtures (Heukelom, 1966). They used the monograph method to obtain the modulus of asphalt, then took this

(32)

modulus into a formula to converge into the modulus of the asphalt mixture. This method is based on the concept that only by knowing the properties of asphalt can be known the properties of the asphalt mixture.

Hadly (1971) applied the direct tensile test to asphalt mixtures. Deacon measured the dynamic modulus of asphalt mixtures by fatigue bending tests. He applied sinusoidal cyclic loads onto small sample beams made from asphalt mixtures. Formulae were given to calculate the dynamic modulus at certain temperatures.

Based on the stress and strain characteristics of steel bridges under traffic loads and temperature changes, Huang and Ren (1994) suggested one kind of EVA modified asphalt binder with improved deformation ability under low temperatures and better stability at high temperatures. Based on material properties such as Marshall stability, flexural strength at low temperatures, creep properties at higher temperatures, fatigue strength under repeated loads etc., they comprehensively discussed and evaluated the performance of this modified asphalt mixture. This mixture was applied to bridges and main trunk roads later, which turned out to be quite good.

Monisimith (1994 & 1995) did fatigue tests on three kinds of asphalt mixtures and sixty-two sample beams, analysed the relationship between energy dissipation and the number of cyclic loading and established an equation that can relate the fatigue life to the total energy consumed. This equation even describes the elastic and viscous properties of the mixture, and the energy dissipation process.

Park, Kim and Schapery (1996) established the viscoelastic continuum model of an asphalt mixture by considering the influence of the damage rate. This model was able to determine the mechanical properties of an asphalt mixture under uniaxial loading and describes its correlation with time.

Xiong and Li (1997) did an experimental study on mixture properties such as dynamic stability, compressive strength, flexural strength and elastic modulus, especially the relations with variation of temperatures.

Wang and Tan (1998) managed to obtain the relationship between dynamic stability of surfacing and axle load. They presented a design method for a heavy loaded asphalt pavement. Zhang, Zhu and Tan (1998) proposed a statistical prediction equation of fatigue properties for asphalt mixes based on cumulative flow energy consumption obtained by creep tests. Jiang (1998) suggested an optimum proportion for SMA mixture based on the available types of aggregates and bitumen in China. Marshall test, rutting test, split test and the triaxial test were adopted to get the technical indicators of this SMA mixture. A comparison was done between the suggested SMA and other dense-graded SMA.

Tan and Zhao (1999) argued that the asphalt material is a typical viscoelastic material, its deformation were decided by Hooke elastic properties and Newton viscous properties. Based on the viscoelastic properties of an asphalt material, they decomposed its deformation under repeated loading and obtained the proportion between viscous and elastic. Zhang and Li (1999) reviewed the origin and application of SMA, as well as its difference from traditional dense-graded

(33)

mixtures. They also reviewed the design and test methods in the US and Europe at that time. Zeng and Chen (1999) concluded different requirements of bridge surfacing layers, argued that three key issues should be analysed during the design of a bridge surfacing structure and asphalt mixtures: the bridge structure, traffic loads and environmental conditions based on climate data. Additionally, they also introduced a design method of SMA mixture to satisfy the aforementioned requirements.

Wang (2000) discussed the selection of asphalt mixtures and the usage of bitumen by testing and empirical formula, as well as the correlation between experimental properties and paving performance.

Li, Sun and Ding (2001) studied a series of requirements for bridge surfacings, including high temperature stability, fatigue performance, resistance to shear deformations and drainage properties. They suggested some new materials and technologies to meet equirements such as crack resistance, stability, toughness, elasticity, aging resistance and drainage. Zhang, Han and Wang (2001) analysed the working conditions of bridge surfacings, discussed the design of water proofing and adhesion materials, compared their performances and talked about the selection of modified asphalt materials and bonding materials etc.

Epoxy Asphalt Concrete is a polymer concrete with a 45 year history as an extremely durable bridge deck surfacing. It was originally developed by Shell Oil Company in the late 1950’s as a jet fuel and jet blast resistant specialty pavement for airfield applications. The first time for it to be applied to orthotropic steel bridge traces back to 1976, the San Mateo-Hayward bridge across San Francisco Bay in the USA. Since 1967 over 450 million pounds (225,000 tons) have been installed on bridge decks a total of over 120 million square feet (ChemCo Systems, 2013). Epoxy Asphalt Concrete is a polymer concrete that is composed of a slow curing, Epoxy Asphalt binder mixed together with standard asphalt concrete aggregates in the pug mill of an asphalt plant. The Epoxy Asphalt binder is a two-phase chemical system in which the continuous phase is a thermoset epoxy and the discontinuous phase is a mixture of specialized asphalts. A hot spray application of an Epoxy Asphalt bond (tack) coat precedes the laying of the Epoxy Asphalt Concrete. Epoxy Asphalt concrete is applied and compacted with conventional asphalt concrete paving equipment. The pavement is quickly ready for traffic in its partially cured state once it has cooled down to ambient temperature. It develops full strength over two to four weeks depending on ambient temperatures.

2.4 Studies on structural analysis of bridge surfacing

structures

The classical method for calculating elastic supported continuous orthotropic steel plates is an analytical solution given by Pelikan and Essliger (1957), named the Pelikan-Esslinger method (P-E method). The P-E method is widely accepted to be a reliable analytical algorithm. Troitsky and Azad (1977) combined the statics method and eccentric stiffener theory together to study an orthotropic steel bridge deck, and compared with results by using the P-E method. The results showed their method to be suitable for design of orthotropic steel deck plates. With the development of computer technology, more accurate

(34)

numerical methods became the mainstream tendency in calculating orthotropic steel bridge structures, e.g. finite difference method (Wang & Li, 1982), finite strip method (Zhang et al., 1985) and finite element method (Reddy, 1993). The bridge surfacing system is composed of an orthotropic steel deck part and a bridge surfacing layer. Early studies on OSDBs are mainly focused on the properties of bridge paving materials, not enough considerations are given to the effects of the steel bridge part onto the mechanical properties of the surfacing layer. In recent years, bridge engineers gradually noticed the importance of matching the steel bridge part and the upper surfacing layer as a whole. Variation of orthotropic steel deck plate parameters has an important influence on the performance of the bridge paving layer. The concept of a comprehensive research work on steel bridge deck, paving materials as well as surfacing structure design as a whole is arousing attention to both the academic and engineering world.

The earliest research work on surfacing systems on OSDBs started in Germany, followed by the United States, Britain, Denmark, the Netherlands, Japan and some other countries and regions. Most of the research work are by using theoretical analysis, numerical calculation method or experimental study, to determine local stresses, displacements and reliability of bridge structures, so as to provide valid data for steel bridge surfacing design.

Based on the plane cross-section assumption, Metcalf (1967) proposed a composite beam theory, in which a simply supported beam, with a point force P applied in the middle was used as the approximate model. He presented the relational curve between the stiffness modulus ratio n and the deflection f，also the curve between n and surfacing strains.

Koroneose (1971) studied the load-shake effect of OSDB surfacing systems. He found that the viscoelastic effect of the surfacing system changed most significantly in the temperature range of 10-20 ℃, temperature had significant effects on the resonance behavior of surfacing layer under load impacts.

Cullimore (1983) simplified the surfacing system into a cantilever beam model and used the stress function Φ, neglected the dead load, established the stress function equation. He defined the boundary conditions by polynomials and calculated the stresses of surfacing layers.

Gunther, Bild and Sedlacek (1987) calculated the response of asphalt pavement on OSDBs by a simplified model and compared the results with measured in-situ data. Factors that were related with surfacing durability, such as the deck thickness, stiffeners close to main girder, asphalt pavement stiffness, material properties and pavement strength were discussed. Soliman (1987) did theoretical and experimental research on several steel deck surfacing systems. Temperature distribution through the depth of the surfacing layer was proposed and the temperature stress could be estimated.

Kolstein (1997) extended Metcalf’s work by considering a complete debonding condition between steel plate and surfacing layer. Nakanishi and Kensetsu (2000) took a two-span continuous beam as calculation model and did experimental study as well. They defined an adhesion coefficient t valued from 0

(35)

to 1 to simulate the bonding conditions between surfacing layer and steel deck. Strains under different conditions were compared. Shear stress distributions inside the surfacing layers were illustrated. Kennedy, Asce and Pfeil, Battista (2005), Zhu and Law (2000) did finite simulations and tests onto OSDB surfacing systems, special attention was paid on fatigue cracking issue under both static and dynamic traffic loads.

Nishizawa, Himeno, Nomura and Uchida (2001) set up a FE model of an OSDB surfacing system which contained finite strip elements, finite prism elements and finite connection elements. In this model, the rib structure, bonding layer and paving layer were regarded as independent units.

2.5 Research on adhesion performance of interlayers

Goodman (1968) proposed a zero-thickness contact element which was able to simulate the cleftiness inside rocks. Desai (1984) proposed a thin contact element with a certain thickness. The mechanism of two-node contact element is simple, and it is easy to simulate in finite element analysis, but it can only roughly model the deformation at the contact interface. The zero-thickness Goodman element has a clear concept and can better reflect the development of contact shear stress and deformation at the interlayer, and nonlinear characteristics of contact interlayers could be simulated too. By direct shear tests, its parameter could be easily determined, and the shear contact behaviour could be considered to some extent. Its disadvantage is the large normal stiffness value in order to prevent excessive overshooting, which will often result in large normal stress errors. Also, the contact interface may fluctuate.

Contact was first brought out to study the interaction of soil and its surrounding materials (Clough & Duncan, 1973; Zong-Ze, Hong & Guo-hua, 1995). The embryonic form of a contact element is a two-node element. Two nodes are set at two sides of the same location at a contact interface (Zhang & Ge, 2005). The element is composed of normal and tangential springs with stiffness coefficients. When the interface cracked, the stiffness coefficients are set to be infinitely small to simulate the non-connection reaction between two sides; large stiffness coefficient values model the fully adhesive condition.

Surfacing systems with different surfacing materials or adhesive layers may have different forms of shear failure (Zaman, Desai & Drumm, 1984). Much work done to asphalt surfacing structures was based on the assumption that the surfacing overlay was fully bonded to the steel deck plate, without considering non-perfect bonding of interlayers. There were also several researchers that adopted the Goodman zero-thickness contact element to simulate the interlayer between the asphalt layer and the steel deck plate, neither perfectly smooth nor completely bonded (Huang, Wang & Chen, 1999; Xiao & Zha, 2000). Nishizawa et al. (2001) created a SLPE model to simulate steel bridge surfacing system, using prism element to model the asphalt surfacing layer, shell element to model a steel deck plate, and describing interlayers by Goodman contact elements.

Shen and Cao (1996) also found shear deformation occurring between surfacing layer and steel deck plate via experimental tests of a bridge surfacing system

(36)

with 12 mm steel deck plate, polymer modified asphalt binder and 60 mm polymer modified asphalt concrete. Hameau (2001) tested the strain distributions of a multilayer surfacing system with a 10 mm steel deck plate, 3mm rubber asphalt waterproofing and adhesive layer, and a 60 mm mastic asphalt pavement layer. Large shear slippage was found between the asphalt surfacing layer and the steel deck plate. The strain distribution inside the mastic asphalt layer was nonlinear. Huang and Li (2001) studied strain distributions of steel bridge surfacing structure with a 14 mm steel deck plate, epoxy asphalt adhesive layer and a 50 mm epoxy asphalt concrete layer. They argued that the strain distribution through the thickness of the surfacing layer was linear.

2.6 Membranes for orthotropic steel deck bridges

In the Dutch primary and secondary road network, many of the bridges with larger spans are OSDBs. In many cases two layered surfacing structures are placed. In these cases a layer of Porous Asphalt (PA) is placed over a layer of Guss Asphalt (GA) placed on the steel deck bridge. Membranes are placed between the various layers for two main reasons: 1) to provide a watertight seal and 2) to provide some kind of durable bond between layers. The second reason for membrane application is vaguely formulated. The reason for this is that there is no consensus on the mechanical function of membranes. In general, two views exist: 1) membranes are bonded to sliding layers that act to disconnect the various structural layers; 2) membranes are shear bonding layers that act to promote the composite action of the structure as a whole.

In this section, the literature review will be divided in two main parts. The first part briefly introduces the types and functions of available membranes. The second part introduces testing and research on membranes.

2.6.1 Types and functions of available membranes 2.6.1.1 Types of membranes

A membrane in bridge surfacing systems is defined as a thin impermeable layer that is used in conjunction with asphalt wearing surface to protect the deck plate from the penetration of moisture and deicing salts. Most Canadian provinces and many European countries require the use of membranes on new bridge decks. About 60% of the U.S. state agencies use them with greater usage on existing bridge decks than new bridges (Russell, 2012).

In literature, several groups of membranes could be identified based on certain distinctions, such as membranes with or without inlays (reinforced), preformed membranes (produced in a factory) or in-place formed membranes (liquid applied membranes).

Preformed membranes involve the application of a primer to the clean bridge deck to improve the adhesion of the membrane to the deck. Some preformed membranes include a self-adhesive backing on the membrane sheet. These sheets can be rolled into place and then bonded to the deck primer using a roller. Others are bonded to the deck by heating the membrane using either a hand

(37)

torch or a machine. After the membrane is installed, a tack coat is applied to the top surface to increase bond with the asphalt overlay.

In-place formed membranes may be placed using either spray equipment or rollers and squeegees. These membranes are applied either hot or cold depending on the manufacturer requirements and they may or may not contain a reinforcing fabric.

Materials used to produce the membranes by various manufactures are rubberized asphalt, polymer-modified asphalt, modified bitumen, polymeric membrane, or bitumen and polymers (Russell, 2012).

In this dissertation, any layer between the steel deck and the asphalt layer, or between two asphalt layers, is termed a membrane. No distinction is made between preformed membranes prefabricated in a factory and in-place formed membranes. Similarly both membrane types with and without an inlay (reinforcing fabric) are considered membranes.

2.6.1.2 Functions of membranes

The following summarize the function of membranes in asphalt concrete surfacing structures.

1) Corrosion protection. The steel deck plate should be protected against corrosion. This function is achieved by several systems, where in most cases corrosion and waterproofing are more or less combined. From a literature survey, primers are commonly used to protect the steel deck from corrosion. Nunn and Morris (1974) reported a zinc-sprayed steel deck with one coat of etch primer. Primer bitumen and bitumen tack coat was used for the same purpose. Smith and Cullimore (1987) presented a cold setting coal tar epoxy system on the steel deck and lightly dressed with sand prior to setting. A three-coat system was applied on the shot blasted the steel top plate to satisfy corrosion and waterproofing requirements (Vincent, 2004). Héritier et al. (2005) discussed a primary bituminous fixing intended to protect metal from corrosion. Corrosion protection of the steel deck is trivial but essential (Huurman & van de Ven, 2008).

2) Watertight seal. Most researchers indicate that adhesive membranes are amongst others applied to provide a watertight seal protecting any underlying layer (steel, primer or asphalt concrete) against water infiltration.

3) Water discharge. Adhesive membranes can also be used as water discharging layers on bridge decks. This is more often observed in relation to concrete decks where a membrane may be used to discharge moisture entrapped during construction.

4) Mechanical functions. The mechanical function of a membrane always plays an important role. Even in situations where the membrane is applied for other reasons, the mechanical interaction with the whole structure is very important. The importance of relating the application of a membrane to the mechanical behaviour of the whole structure was overemphasized by many researchers (Touran & Okereke, 1991; Huurman, Medani, Scarpas & Kasbergen, 2003; Medani, 2006).

(38)

5) Resistance against thermal and mechanical attacks during paving. A last aspect that needs to be mentioned is the resistance against damage during construction especially when the temperature is high with mastics and hot mix asphalt overlays. Damage may occur due to compaction and other interactions. 2.6.2 Tests on adhesive membranes

2.6.2.1 Tensile bond strength tests

Tensile bond strength tests are also called pull-off tests, Figure 2.2. The pull-off adhesion strength and mode of failure of a coating from a substrate are important performance properties that are used in specifications. This test method covers procedures for evaluating the pull-off adhesion strength of a coating on concrete. The test determines the greatest perpendicular force that a surface area can bear before a plug of material is detached. Failure will occur along the weakest plane within the system comprised of the test fixture, adhesive and substrate.

Figure 2.2 Tensile bond strength test (left: schematic diagram; right: in situ)

The tests are commonly used to determine the effects of temperature on the strength of adhesive membranes. However, for this test method, the possible debonding position can be at the top interface of the membrane or at the bottom interface or even across three materials layers. This will cause some ambiguity as to the actual debonding strength and mechanism.

2.6.2.2 Blister test

The blister test was described by Dannenberg in 1961 as a means of quantifying the adhesion of relatively thin plates to flat and rigid substrates. The test specimen consists of a perforated substrate with a thin flexible overcoating. A fluid is injected at the interface through the perforation, thereby causing a progressive debonding of the overlayer, Figure 2.3. Gent and Lewandowski (1987) have discussed how the adhesion energy can be calculated from the geometry of the blister and the fluid pressure. Recently, the test has been reanalysed and applied to predict the bonding between bituminous sealant and

(39)

Figure 2.3 Schematic diagram of the blister geometry

In the blister test, hydraulic liquid with pressure p is injected into the centre of the substrate, the adhesive lifts off the substrate. A blister forms, whose radius stays fixed, until a critical pressure is reached. At this critical value, the circumference of the blister increases in size and the pressure drops, which indicates adhesive failure. The blister test can measure interfacial fracture energy, which is a fundamental property of the interface. For elastic material, interfacial fracture energy is constant and directly related to adhesion strength. For viscoelastic material, interfacial fracture energy is expected to be time-dependent, although still geometry-intime-dependent, like the relaxation modulus. In earlier tests, deflections and debondings were caused by applying hydrostatic pressure. One disadvantage of the hydrostatic pressured blister test, is that the strain energy release rate increases as blister radius increases and debondings become unstable (Lai and Dillard, 1994). Moreover, pressurized blister tests require sophisticated experimental equipment to monitor the simultaneous change in blister dimension and dissolved gases may invalidate such tests (Wan, 1999).

2.6.2.3 Shaft loaded blister test

The shaft loaded blister test (SLBT), first reported by Williams (1969), introduces the crack driving force via a central load acting on a spherically capped shaft, Figure 2.4. The shaft loaded blister test offers an alternative to pressured blister tests because a universal test machine can drive the shaft that induces displacements. The configuration is of particular interest due to the simpler experimental setup and has received a considerable amount of attention in the last 20 years.

Even though the blister test is widely used to measure mechanical properties of thin films, its application is limited because of the complicated fluid control system and the difficulty of simultaneously measuring both blister height and pressure. Therefore, The SLBT is a very good replacement of the traditional blister test, which a transverse load is applied to a thin film by a mechanical system where the load-shaft displacement data is easily obtainable

(40)

experimentally. One drawback needs to be mentioned about this test is that, even with a spherically capped shaft, still there is a severe stress concentration around the loading point, which would affect the test accuracy.

Figure 2.4 Schematic diagram of the shaft loaded blister test

2.6.2.4 Peeling test

The schematic diagram of a peeling test is shown in Figure 2.5. This test method is primarily intended for determining the relative peel resistance between flexible adherents by means of a specimen using a tension testing machine. The unbounded ends of the test specimen are clamped in the test grips of the tension testing machine and a load of a constant head speed is applied. A recording of the load versus the head movement or load versus distance peeled is made. Peel resistance over a specified length of the bond line after the initial peak is determined.

Figure 2.5 Schematic diagram of peeling test

The peeling test meets many of the criteria of the ideal adhesion test. Sample preparation is typically simple and straightforward. Another advantage of this test is that the rate of the delamination and the locus of failure can be controlled precisely. This stems from the fact that a very high stress concentration exists at the point where the delamination starts.

(41)

A drawback apply to the peeling test is that, large stress concentration or deformation occurs around the clamp area, making interpretation of the results unclear.

(42)

Optimum Design of Multilayer Asphalt Surfacing Systems for Orthotropic Steel Deck Bridges

Optimum Design of Multilayer Asphalt

Surfacing Systems for Orthotropic Steel Deck

Optimum Design of Multilayer Asphalt

Surfacing Systems for Orthotropic Steel Deck

Bridges

Acknowledgements

Table of Contents

List of Symbols

1 Introduction

1.1 Problems on the surfacing system of orthotropic steel

deck bridges

1.2 Merwedebrug - bridge of concern

1.3 Project descriptions



5 C

1.4 Outline of the dissertation

2 Multilayer Surfacing Systems on Orthotropic

Steel Deck Bridges

2.1 Bridge deck surfacing systems

2.2 Characteristics of surfacing systems on orthotropic steel

deck bridges

2.3 Research work on surfacing materials

2.4 Studies on structural analysis of bridge surfacing

structures

2.5 Research on adhesion performance of interlayers

2.6 Membranes for orthotropic steel deck bridges