Discover millions of ebooks, audiobooks, and so much more with a free trial

Only $11.99/month after trial. Cancel anytime.

Designing and Building with UHPFRC
Designing and Building with UHPFRC
Designing and Building with UHPFRC
Ebook1,326 pages11 hours

Designing and Building with UHPFRC

Rating: 0 out of 5 stars

()

Read preview

About this ebook

This book contains the proceedings of the international workshop “Designing and Building with Ultra-High Performance Fibre-Reinforced Concrete (UHPFRC): State of the Art and Development”, organized by AFGC, the French Association for Civil Engineering and French branch of fib, in Marseille (France), November 17-18, 2009. This workshop was focused on the experience of a lot of recent UHPFRC realizations.

Through more than 50 papers, this book details the experience of many countries in UHPFRC construction and design, including projects from Japan, Germany, Australia, Austria, USA, Denmark, the Netherlands, Canada… and France. The projects are categorized as novel architectural solutions, new frontiers for bridges, new equipments and structural components, and extending the service life of structures. The last part presents major research results, durability and sustainability aspects, and the updated AFGC Recommendations on UHPFRC.

LanguageEnglish
PublisherWiley
Release dateJan 29, 2013
ISBN9781118587553
Designing and Building with UHPFRC

Related to Designing and Building with UHPFRC

Related ebooks

Construction For You

View More

Related articles

Reviews for Designing and Building with UHPFRC

Rating: 0 out of 5 stars
0 ratings

0 ratings0 reviews

What did you think?

Tap to rate

Review must be at least 10 words

    Book preview

    Designing and Building with UHPFRC - Jacques Resplendino

    UHPFRC’09 Sponsors

    AFGC Association française de génie civil − French Association for Civil Engineering, French branch of fib, IABSE and RILEM

    fib fédération internationale du béton International Federation for Structural Concrete

    MEEDDM / CGDD Ministère de l’écologie, de l’énergie, du développement durable et de la mer / Commissariat général au développement durable − French Ministry in charge of sustainable development

    EFB Ecole française du béton − French foundation for knowledge on concrete

    CERIB Centre d’études et de recherches de l’industrie du béton − Research center for precast concrete industry

    Eiffage TP

    Lafarge

    Vicat

    VSL

    The financial support of the UHPFRC’09 Sponsors enabled the organization of the international workshop in Marseille 17-18 November 2009, with over 330 attendees, thus collecting the valuable information provided in this book.

    PART I

    Genesis of a New Construction Material

    Chapter 1

    Introduction: What is a UHPFRC?

    After recalling the definition of ultra high performance fiber-reinforced concretes (UHPFRC), this chapter details the basic elements of their composition. It presents the main products available on the market and provides a brief history of the development of these concretes. Then the chapter details the main features of UHPFRC, highlighting what distinguishes them from conventional concrete: heat treatment, delayed effects, compressive and tensile strength, performance in terms of sustainability and quality of the faces. The chapter analyzes the changes in design and implementation technology necessitated by UHPFRC development and ends with a presentation of the interest in these materials as part of sustainable development.

    UHPFRC are materials with a cement matrix and a characteristic compressive strength between 150 MPa and 250 MPa. They contain steel fibers in order to achieve ductile behavior in tension and, if possible, overcome the use of passive reinforcement. UHPFRC differ from high performance and very high performance concretes in that:

    – the systematic use of fibers ensures that the material is not brittle and can allow us to avoid any classical active or passive reinforcements;

    – their compressive strength is generally greater than 150 MPa;

    – their mix-design with a high binder content leads to the absence of any capillary porosity;

    – the direct tensile strength of the matrix is systematically higher than 7 MPa.

    The aim of UHPFRC development is to achieve high tensile strengths through participation of the fibers, which provide tensile strength after the cement matrix has cracked. When the tensile strength is sufficiently high, it may be possible, depending on the way the structure works and the way the loads to which it is subjected, dispense with conventional reinforcement. In general, we remove any traditional passive reinforcement cage in order to keep just the main passive or active reinforcement bars required when the resistance to major forces cannot be provided by the fibers.

    1.1. The basis of UHPFRC mix-design

    In ordinary concretes, the ratio of water to binder (W/B) is in the order of 0.4 to 0.6. To obtain a high performance concrete, the W/B ratio has to be decreased to below 0.4 (0.3 to 0.35). This reduction is possible thanks to the addition of an admixture (superplasticizer), which allows deflocculating binder, and cement with minimal amounts of water. In a high-performance concrete, if we wish, we can complement the W/B reduction with an extension of the granular spectrum through ultra-thin micrometer-sized additions (generally consisting of silica fume), 5 to 10 times smaller than the size of the cement particles. To obtain UHPFRC, we should further decrease the W/B ratio to below 0.25 (about 0.16 to 0.2). This decrease is generally obtained through a significant increase in the binder amount, and a quantity of water substantially the same as in conventional concrete. We should also add a large quantity of ultra-thin addition (typically silica fume) amounting about 20% of the cement mass. The UHPFRC aggregate size is also very small compared to ordinary concrete – the largest grain size is of the order of millimeters – and particular attention is given to the nature of the aggregates, which must present sufficient mechanical strength to avoid being the weak points of the mixture. The resulting material is extremely compact and has quite remarkable mechanical and durability performance.

    With such a formulation we relatively easily obtain compressive strength above 150 MPa, but the material becomes fragile: explosive in compression with complete disappearance of any plastic domain. To avoid this problem, it is necessary to add fiber in order to restore ductile behavior in compression. For structural uses (when mechanical performance of the material is needed to ensure the structural strength) we use the steel fiber and fiber rate needed to ensure a non-brittle behavior in bending (about 2 to 3% by volume, i.e. 160 to 240 kg of steel per cubic meter of concrete, up to 10% by volume in high-performance formulations in which we seek to ensure perfect hardening behavior in pure tension). To implement such proportions of fiber, they are generally straight and not serrated to prevent urchin formation and to ensure a proper rheology of the UHPFRC. The fibers are small needles made of very high resistance steel. Their size is matched to that of aggregates in order to limit shock embarrassment aggregate-fiber: a length from 12 to 20 mm and diameter from 0.1 to 0.3 mm. Some UHPFRC include several sizes of fibers that may have complementary behaviors (microfiber improving the anchoring of larger fibers, and sewing the first networks of microcracks). The high fiber content used in UHPFRC gives them quite interesting tensile and shear strength (about 8 to 11 MPa in direct tension and 25 to 40 MPa in bending) that allows traditional passive reinforcement frames to be dispensed with. The steel reinforcement is limited to acting principal forces on large sections. To enable the design of structures without any reinforcements, it was necessary to develop rules for specific calculations. This objective absolutely necessary to produce a work in public domain was one of the main motivations that led to the drafting of the AFGC (Association Français de Gènie Civil) recommendations on UHPFRC.

    1.2. The main UHPFRC available on the market

    UHPFRC currently available on the market in France are:

    – the different DUCTAL® concretes, including RPC (Reactive Powder Concrete) from the research program between Bouygues, Lafarge and Rhodia in France and marketed by LAFARGE;

    – BSI/CERACEM® developed by the group with EIFFAGE SIKA;

    – BCV® developed by the cement and the Vicat group Vinci.

    Other UHPFRC have been used for various applications:

    – CEMTECmultiscale® developed by LCPC, applied to several works in Switzerland and Canada; and

    – laboratory materials developed by EDF, the CERIB (Centre d’Études et de Recherches de l’Industrie du Béton).

    Abroad, we can note CRC (Compact Reinforced Composite) technology developed by Aalborg Portland Cement (Densit) in Denmark. Some products are developed in Germany and Japan, knowing that it is mainly BSI/CERACEM® and Ductal® techology that are used in Western Europe, and Ductal® is the leading technology employed in Asia, Australia and North America (US and Canada).

    1.3. Brief history of the development of UHPC

    Early research into UHPFRC was carried out by Professor Bache in the 1970s in Denmark through the development of CRC technology. This technology is still very active and probably accounts for the biggest worldwide production by volume of UHPFRC. However, this is a special technology in which a large percentage of metal fiber is added to a cement matrix to produce prefabricated building structures (balconies and staircases) that are reinforced by traditional means calculated without taking into account the mechanical participation of the fibers.

    In France, initial research on UHPFRC was in the 1990s under the leadership of Pierre Richard (Bouygues Group) through Reactive Powder Concrete (RPC) technology. The design was then optimized in the Lagarge research center at l’Isle d’Abeau in partnership with Bouyghes and Rhodia to develop Ductal®, the first marketed UHPFRC, which was launched in the late 90s. This technology is distributed both in Western Europe [BEH 03, HAN 06], Asia [BEH 03], Australia [CAV 03] and in the United States. It is used in all kind of structural applications, such as bridges and building, and also in non-structural cladding like street furniture and/or decorative and design objects.

    Under the leadership of EDF (the French power company), which has been a very influential contractor in the development of UHPC, technology BSI/CERACEM® was created by Eiffage in the late 90s and then developed with the assistance of the company SIKA. This technology gives rise to the realization of a growing number of applications, both in the field of new structures [DEM 08, ELG 08, HAJ 03, HAJ 04a, HAJ 04b, RES, THI 02] and in repairing or strengthening old structures [GEN 04], mainly in France and Western Europe. In the 2000s, the cement company Vicat with the support of the Vinci Group has developed the BCV, which is also the subject of significant structural achievements [RES 06]. With the experience and potential of these materials [RES 08], France was the first country to issue recommendations in 2002 [BET 02, HAJ 04c, RES 04] to formalize methods to characterize the performance of these materials and to give rules to design UHPFRC structure without any frame other fiber that constitutes the major innovation of this type of material.

    In the 2000s, several countries have engaged research on UHPFRC. The Japanese are very active. They published recommendations in 2004. They have made several outstanding designs [OKU 06] (footbridges, road and rail bridges) and currently have significant structural applications (components of high-rise buildings, airport structures in marine sites). In Australia, significant activity has developed based on the realization of structures [CAV 03] and the use of such materials in shields to protect against explosions. In 2005, the Germans started an ambitious program of research (€10,000,000) over six years involving a large number of universities under the guidance of the University of Kassel. They have little experience in the design and actual use of the material due to a lack of companies and building owners involved in the development of these technologies.

    1.4. The main features of UHPC

    This article does not purport to present all the mechanical characteristics of UHPFRC, but rather to highlight the specific performances that distinguish them from conventional concrete.

    1.4.1. Thermal treatment

    Some UHPFRC are subject to special treatment (called type 2 in the new recommendations) applied several hours after the concrete sets. This treatment consists of heating the element to a very high temperature (about 90ºC) and relative humidity close to saturation for approximately 48 hours. This treatment can significantly increase the durability and reduce delayed effects: no drying shrinkage and a very substantial reduction of creep. More traditional treatments, such as steaming, applied in the early hours can possibly anticipate the beginning of the setting and accelerate the initial set of UHPFRC. These treatments are performed at a moderate temperature (about 40 to 50°C) and have a much smaller influence on delayed effects. Many UHPFRC are implemented without any heat treatment, which does not prevent them from very quickly developing resistance (greater than 70 MPa after a few hours). These mixtures often have a dormant phase before the setting phase, which is longer than conventional concrete. When setting starts, the increase in resistance is very fast.

    1.4.2. Shrinkage and creep

    Unlike ordinary concrete, the UHPFRC has a very low W/L ratio, which causes very high autogenous shrinkage (about 550 μm/m) and very small drying shrinkage (about 150 μm/m for UHPFRC without heat treatment, and no shrinkage after a type 2 heat treatment). The creep coefficient for UHPFRC without heat treatment is comparable to the coefficient obtained for HPC with silica fume (long-term creep coefficient between 0.8 and 1.0). After a type 2 heat treatment, the creep coefficient decreases dramatically (between 0.2 and 0.5). These characteristics have many implications on the performance and conditions of use of UHPFRC:

    – the formwork of restrained pieces should be designed in order to allow free retraction and avoid important internal stresses or cracks;

    – UHPFRC are ideally suited to precasting: after the end of the setting, subsequent delayed effects (shrinkage and creep) are very low compared to conventional concrete. This provides structural elements that do not budge. This is especially true when the elements are subjected to a type 2 heat treatment;

    – note that in the case of prestressing by pre-tensioning, a type 2 heat treatment has no influence towards limiting prestress losses: in fact, the heat treatment begins only several hours after the end of setting. It is made after the release of the cables and prestressing, so that the overall shrinkage causes prestress losses;

    – on the other hand, in the case of post-tensioned prestressed concrete, applying a type 2 heat treatment to UHPFRC before prestressing tends to very significantly reduce the prestress losses due to a combination of lack of drying shrinkage and dramatic decrease in creep strains.

    All these remarks are valid in the case of use of UHPFRC elements connected to a metal frame [TOU 05, TOU 08a, TOU 08b] or other material: the reduced delayed deformations of UHPFRC (shrinkage and creep) limit the internal forces due to the connection between the metal and the concrete.

    1.4.3. Compressive strength

    The UHPFRC behavior laws in compression are almost straight and have no real plastic range. The dispersions obtained on the compressive strengths are generally low due to the quality and homogeneity of the premix. Note that the very large compressive strengths (from 150 to 250 MPa) oblige us to reduce the size of the specimens compared to ordinary concretes (capacity presses are required) and special care is required in their creation. These requirements lead some firms to carry out checks using tests on cubes that are easier to make. This approach is accepted by the recommendations subject to having well-calibrated and justified correlation curves between strengths obtained on cubes and those obtained on cylinders in advance.

    1.4.4. Tensile strength

    The biggest challenge for UHPFRC, and what distinguishes them radically from the other concretes, is tensile strength. This concerns the strength of the material before cracking and the post-cracking resistance that involves mainly fiber strength. The tensile strength before cracking poses no real problem. Indeed, if control of the setting is correctly implemented (formulation, mixing time), it yields a low dispersion of the tensile strength of the cement matrix that governs the resistance before cracking. However, the fundamental issue of UHPC is to get control and ensure the strength after cracking in any structure knowing where this resistance depends on the orientation of the fibers, which depends on the conditions of implementation:

    – any potential flow during casting tends to orient the fibers in the direction of flow;

    – the fibers close to the walls tend to be oriented parallel to the formwork. This phenomenon only occurs for a depth less than or equal to the length of fiber. It has all the more influence on the tensile strength that the thickness of the structure is close to the fiber size;

    – a preferred orientation of fibers in the direction of gravity can sometimes occur due to the natural behavior of fibers in the viscous liquid phase of the concrete before setting.

    To better control the orientation of fibers during implementation, the general principle is to create the pieces by putting UHPFRC directly to its final position, to avoid flows. If we cannot remove flows, we try to compensate for and/or direct the flow in the direction of the main forces. In any event, to validate the process of implementation, we systematically perform a suitability test upstream of the actual structure. This suitability test consists of creating a specimen representative of the real structure, made of the same material and realized using the same procedures as those proposed for the execution of the actual structure. In that element we cut out several prisms in the direction of the main forces and perform bending tests in order to determine the actual tensile behavior law in the structure and correct the outcome of the theoretical law of laboratory specimens (the concept of the corrective coefficient K).

    For structures of substantial thickness for which the wall effect is limited, the size of specimens for tensile bending test must be relatively large compared to fiber size. Indeed, if the size of the prism is too small, the wall effect tends to comb the fibers in the direction of the prism (they are parallel to the edges) and it overestimates the actual strength of the laboratory material. Conversely, if the size of the specimens collected in the structure is too small, we obtain a very important dispersion on sawn samples that tends to underestimate the actual strength of the structure, or even render the statistical analysis of these samples impossible.

    Conversely, for thin elements in which the wall effect is fundamental (thickness of the element minus three times the fiber length, tending to favor a 2D orientation of the fibers), we directly test elements of the same thickness as the actual structure, both for prior laboratory tests and suitability tests.

    1.4.5. UHPFRC durability

    UHPFRC materials are by definition extremely close and compact. They have a special porous structure characterized by an absence of porosity and capillary porosity in very small non-interconnected scale (see Figure 1.1). This feature gives them quite exceptional performance in term of durability: water porosity, air permeability and chloride-ion diffusion factors are greatly reduced compared to ordinary or high-performance concrete. These performances make them suitable for all structures subjected to aggressive external environments [BEH 07]. Furthermore, their resistance to abrasion (Compagnie Nationale du Rhône – CNR test) and resistance to dynamic effects [SER 98, TOU 99] make it particularly relevant to applications for the protection of hydraulic works or structures subjected to shocks or explosions. Note that in shocks, the presence of small fibers limits the flying chips, which can sometimes pose safety risks for people.

    Figure 1.1. Distribution of pore sizes for different concretes (courtesy of CERIB data)

    1.4.6. Formed surfaces and esthetic aspects of UHPFRC

    The very compact characteristics of UHPFRC and the fineness of the material allow us to obtain surfaces of very high architectural quality, with the possibility of using a very fine matrix for the formwork. Yet esthetic problems may appear on unformed surfaces and due to trace oxidation related to the presence fibers at the surface of the walls in clear tint concrete, observable at short distances. Regarding unformed surface treatment, great care must be taken to curing. Several techniques can be used for leveling and/or smoothing unformed surfaces to avoid fibers that go beyond the surface. These treatments include the application of plastic or rubber spike rolls to fit the fibers and set the surface, avoiding adhesion problems related to the viscosity of the material. Techniques using slippery sheets can obtain similar results, while providing interesting solutions. For limiting the quantity of fibers appearing on the formed surfaces, efforts must be made on mix-proportioning and on the nature of the formwork to avoid any apparent fiber at the surface; such efforts thereby eliminate the risk of rust spots later. The joints between form panels should be completely sealed. If a significant number of fibers appear at the surface, special surface treatments can be implemented to prevent further corrosion of the fibers. Note that in architectural applications made with a white premix, CRC technology provides the implementation of stainless steel fibers. This technology is not common and has never been tested for a structural UHPFRC without passive reinforcement (CRC Danish technology combines UHPFRC with traditional passive reinforcement). Similarly, for certain precast applications we can make sandwich panels from steel fiber UHPFRC bordered by a surface layer made of an organic UHPC fiber. One of the main UHPFRC uses, for esthetic and architectural purposes, is that any shape can be created knowing that there are no longer geometrical constraints due to traditional reinforcement cages.

    1.5. UHPFRC: a material that needs to revolutionize the technologies of design and implementation

    The design and implementation of structures made of UHPFRC require removal of the reflexes that exist in traditional reinforced or prestressed concrete. The volumes of structures are further modified with widths that can bevery thin, which can be alarming to the uninitiated but which are often perfectly appropriate. The reduction in thickness promotes the fiber orientation and increases the resistance of the material. The absence of reinforcement makes the notion of minimum thickness required for steel reinforcement coating redundant. The ability to make very slender parts requires vigilance with regards to deformation and verification efforts, especially during transitional phases like handling or assembly. In addition, the more slender the elements are, the more reduced the dimensional tolerances have to be. Manufacturing processes and controls must be adjusted accordingly. The shape of the formwork must be completely reviewed. Traditional structures include angles allowing an easy shaping of the reinforcement frames. With UHPFRC rounded surfaces are better as they facilitate the implementation of the material and avoid discontinuities of flow.

    The lack of reinforcement allows us to design all forms of structures as long as the geometry is coherent and provides a good resistance to the effort. The formwork technologies should evolve through solutions made in plastic or polystyrene automatically cut to create any 3D shapes. The development of prefabrication should go hand-in-hand with research on connections between elements to achieve the most efficient transmission of forces, tightness and durability of the assembly. UHPFRC as an almost inert material should be developed in the design of composite structures and/or as a connection solution between a steel frame and a traditional part in reinforced concrete. A major hindrance to the development of the material is probably related to the methods of calculations, which are not traditional for engineers. Strength calculations of fiber-reinforced concrete are poorly understood. It is the same for methods of tensile strength testing in bending, which are little known by laboratories and require specific digital tools (inverse method).

    1.6. UHPFRC and sustainable development issues

    A UHPFRC contains about twice the cement volume as conventional concrete, and thus produces twice as much CO2 and consumes twice as much energy in production. Yet experience using UHPFRC shows that if used appropriately, the quantities of material used in a structure can be divided by two or three. A UHPFRC structure therefore provides a slight gain in terms of initial CO2 footprint and energy compared to a conventional solution. It also offers a significant gain in terms of durability, lightness and global economy of material. It is therefore important to incorporate an anticipation of sustainability earnings enabled by UHPFRC solutions; this is particularly relevant when searching for long-life or evolutive structures and when taking into account economic cost, image, operating constraints and the environmental costs of all operations of maintenance required by traditional solutions.

    1.7. Conclusion and outlook

    The design and construction of UHPFRC structures require us to dispose of the reflexes attached to traditional reinforced or prestressed concrete. UHPFRC are not revolutionary materials, in the sense that their cost of production and implementation is still high and requires optimization to ensure a significant financial gain. Niches exist and powerful applications tend to increasingly grow as far as durability, esthetics, timeliness for erection, scalability, possible layouts and material gain are concerned. These uses are growing and UHPFRC can bring real innovative responses. Researches and important projects currently under way will further strengthen their medium-term development and contribute to further demonstrate their structural and architectural potential.

    1.8. Bibliography

    [BEH 03] BEHLOUL M., LEE K.C. Ductal® Seonyu footbridge, Structural Concrete, vol. 4, no. 4, pp. 195-201, 2003.

    [BEH 06] BEHLOUL M. et al. Ductal® prestressed girders for a traffic bridge in Mayenne, France, 7th int. Conference on Short and Medium Bridges, Montréal, Canada, 2006.

    [BEH 07] BEHLOUL M., ARSENAULT J. Ductal®: a durable material for durable structures, 5th Int. Conf. on Concrete Under Severe Conditions, CONSEC’07, Tours, France, 2007.

    [BET 02] Bétons Fibrés à Ultra-Hautes Performances, Recommandations Provisoires, AFGC-Sétra, 2002.

    [CAV 03] CAVILL B. & CHIRGWIN G. The world’s first Ductal road bridge Shepherds gully creek bridge, NSW, 21st Biennial Conference of Australian Concrete Institute, Brisbane, 2003.

    [DEM 08] DE MATTEIS D. et al. A fifth French bridge including UHPFRC components, the widening of the Pinel Bridge, in Rouen (France), Int. Symp. on UHPC, Kassel, 2008.

    [ELG 08] EL GOURARI Y. et al. La poutre ITE, une alternative aux poutrelles enrobées – Le triplement du pont Pinel à Rouen, Revue Travaux, n°849, February 2008.

    [GEN 04] GENES P. et al. Restructuration d’un immeuble parisien, Revue Travaux, November 2004.

    [HAJ 03] HAJAR Z. et al. Construction of the first road bridges made of UHPC, 3rd International Symposium on HPC, Orlando, 2003.

    [HAJ 04a] HAJAR Z. et al. Design and Construction of the world first Ultra-High Performance Concrete road bridges, International Symposium on UHPC, Kassel, 2004.

    [HAJ 04b] HAJAR Z. et al. Realization of an Ultra High Performance Fibre Reinforced Concrete thin shell structure covering the toll-gate station of Millau Viaduct, fib Symp., Avignon, 2004.

    [HAJ 04c] HAJAR Z. et al. UHPC – First recommendations and examples of application, fib Symposium: Concrete Structures: The Challenge of Creativity, Avignon, 2004.

    [HAN 06] HANOTEAU J. et al. Ductal: a new material, the bridge of St Pierre La Cour, in The French Technology of Concrete, AFGC, 2006.

    [OKU 06] OKUMA H et al. The first highway bridge applying ultra high strength fibre reinforced concrete in Japan, 7th Int. Conf. on Short & Medium Span Bridges, Montréal, 2006.

    [RES 04] RESPLENDINO J. First recommendations for Ultra-High-Performance Concretes and examples of application, Int. Symp. on UHPC, Kassel, 2004.

    [RES 02] RESPLENDINO J. et al. Ouvrages innovants de Bourg-lès-Valence, Revue Travaux, no. 783, pp. 42-47, February 2002.

    [RES 05] RESPLENDINO J. & BOUTEILLE S. Les derniers développements dans l’utilisation des Bétons Fibrés Ultra Performants en France, GC2005, Paris, 2005.

    [RES 06] RESPLENDINO J. et al. Construction of an overpass on the A51 Motorway, made of a pre-stressed box beam built with UHPFRC, in The French Technology of Concrete, AFGC, 2006.

    [RES 07a] RESPLENDINO J. et al. Construction d’un Passage supérieur sur l’autoroute A51, en caisson précontraint Réalisé en Béton Fibré ultra performant (BFUP), GC2007, Paris, 2007.

    [RES 07b] RESPLENDINO J. & BOUTEILLE S. Etude de solutions de tabliers de ponts réalisés en Béton Fibré Ultra Performant (BFUP) et en Matériau Composite (Fibre de verre), GC2007, Paris, 2007.

    [RES 08] RESPLENDINO J. Ultra-High Performance Concretes – recent realizations and research programs on UHPFRC bridges in France, Int. Symp. on UHPC, Kassel, 2008.

    [SER 98] SERCOMBE J., Modélisation du comportement du béton en dynamique rapide. Application au calcul des conteneurs à haute intégrité, ERLPC OA30, LCPC, Paris, 1998.

    [THI 02] THIBAUX T. & TANNER J.A. Construction of the first French road bridges in ultra high performance concrete, in La Technique Française du Béton, AFGC, 2002.

    [TOU 99] TOUTLEMONDE F. et al. Développement d’un conteneur pour l’entreposage de déchets nucléaires: résistance au choc, Revue Française de Génie Civil, vol. 3, no. 7-8, 1999.

    [TOU 05] TOUTLEMONDE F. et al. Innovative design of ultra-high performance fiberreinforced concrete ribbed slab: experimental validation and preliminary detailed analyses, 7th Int. Symp. on Utilization of HSC/HPC, Washington D.C., USA, 2005.

    [TOU 08a] TOUTLEMONDE F. et al. Validation of connection details between a UHPFRC ribbed slab and steel girders for a new composite bridge deck type, 8HSC-HPC, Tokyo, 2008.

    [TOU 08b] TOUTLEMONDE F. et al. Experimental validation of a ribbed UHPFRC bridge deck, 2nd Int. Symp. on Ultra High Performance Concrete, Kassel, 2008.

    Chapter written by Jacques RESPLENDINO.

    Chapter 2

    UHPFRC: Science-driven Material Engineering Can Provide Sustainable Solutions

    For over a century, the progress made in concrete technology has been very limited – the massive reconstruction program following the Second World War resulted in a substantial drop in concrete quality. It was not until the end of the 1980s that the spectacular rise in performance could be implemented in industrial practice (specification and design), initially in terms of mechanical strength but, very soon after, in terms of other properties (processing, durability, appearance, etc.). This technological revolution clearly was a result of scientific approaches that rapidly superseded the basically empirical approaches that hitherto were the only ones available.

    Today, we are faced with far more serious challenges than those of the last century: exhaustion of resources, climate change, demographic explosion, social inequalities, etc. Herein we argue that the thorough scientific approach can contribute to finding concrete solutions to the considerable problems now facing us. The scientific advances to be explored and put to good use are still considerable and herald significant progress. This has led to concrete emerging as a high-tech bulk material at the forefront of materials science and engineering that can be finetailored, not only to meet the classical strength, stiffness and durability requirements, but also to contribute to reducing the environmental footprint of our societies’ legitimate needs for housing, shelter, infrastructure, and so on.

    2.1. Introduction

    The advent of ultra high performance fiber-reinforced concrete (UHPFRC) stemmed from a scientific approach that is worth examining today, as it clearly shows the greater potential of scientific approaches over empirical or purely experimental approaches, opening the way to prospective advances that will meet the challenges of the 21st Century. After reviewing the history of concrete technology, we will show how the scientific approach has led to unprecedented advances in performance, which, as we will explain, could not be achieved using traditional approaches based just on experimentation, on the principle of trial-anderror, and heuristic exploration. We will then show that there is still a rich potential vein of progress to be made and that research on solutions to today’s problems requires this approach to be continued – the aim is to understand the material, which necessarily includes modeling its behavior.

    In the case of a formulated material, such as concrete (entailing an unlimited number of degrees of freedom for the formulator), understanding means knowing how to model (in terms of equations) the various aspects of its behavior (mechanical, physical, esthetic, thermal properties etc.) and how these vary over time. Such modeling is necessary in order to validate/invalidate explanatory hypotheses. Finally, we mention two disciplines that will have to be assimilated by engineers in order to succeed here, namely:

    – the probabilistic approach, essential for taking into account the unpredictable factors of service conditions for full analysis of life cycles; and

    – the optimization tools that will be necessary for identifying the best solutions within the complex, multi-criteria and variable context of sustainable development.

    2.2. Historical overview of concrete technology

    Concrete technology has undergone a technological revolution that can be placed very precisely at the end of the 1980s, as Figure 2.1 clearly shows.

    Figure 2.1 shows the evolution of the main characteristic quantity used by engineers when designing concrete structures: compressive strength. The values displayed here are not laboratory values or those obtained on experimental structures, but rather the maximum values actually used in the construction of major structures. For example, the last two points of this figure correspond respectively to:

    – the l’Ile de Ré bridge, constructed by Bouygues between 1987 and 1988 with 80 MPa HPC [RIC 99], which opened the way to the use of HPCs in large constructions;

    – the internals of the forced-draft air cooling tower of the Cattenom nuclear power station, with more than 2,000 prefabricated beams made between 1997 and 1999 of 200 MPa UHPFRC [VER 03], this being the first significant industrial application of a UHPFRC.

    Figure 2.1. Evolution during the last century in the compressive strength of concrete available on an industrial scale and used on real worksites

    It should be pointed out that, until then, the only property actually specified in construction contracts was, without exception, the compressive strength. This was at precisely the same time that new specified properties appeared, firstly the lifetime. Historically, the first significant structure for which this property was specified was the Channel tunnel, in which the required lifetime was 120 years. The structure was not to reach, somewhat qualitatively, a certain level of quality (as was sometimes the case previously) but the lifetime was specified because the exceptional level of financing that had to be put together by an international consortium of banks demanded it [ULM 99]. It should be noted that the procedure of formulating a concrete quickly showed that this property – lifetime – was critical and resulted in superseding the compressive strength as the sole specified material design reference. It also gave rise to quite a debate within the profession, a number of engineers still doubting at that time the possibility of guaranteeing a certain lifetime with the same reliability as, for example, fracture resistance. The example of earthquake resistance, solved during the 1980s, showed that if the probabilistic approach is used there is no difference in calculation level between engineering for lifetime, earthquakes, wind resistance, etc. [AFG 04].

    Other properties are now being increasingly specified:

    – processing properties, which now go beyond just the slump value (handling time, viscosity, segregation resistance);

    – surface properties (color, absence of staining, bubbling, etc.); it should be noted that the approach regarding surface properties amounts here to a scientific approach and that these properties can therefore be specified, with a duration over time, using the same engineering tools (measurement techniques, physical models for controlling key parameters, standards) as for other properties.

    It thus appears that a holistic approach that integrates these surface properties into the materials design is important. The social acceptance of buildings, or more generally social progress, should be included in the engineering design equations. In other words, engineers cannot ignore the fact that bad building façades or rapidly degraded concrete surfaces may contribute to deteriorating the social climate within a community. In European cities, graffiti on concrete walls cannot be separated from young people’s rejection of architecture and engineering that ha- ignored them.

    2.3. The scientific concepts that have changed concrete technology

    The scientific advances that started this technological revolution are well known today. These include:

    – The theory of granular packing [DEL 99], using highly accurate mathematical models for optimizing, in the case of spheres, the compactness of a polydisperse mix. For actual aggregates, the model must be corrected by calibration on experimental data; but since the starting model captures the fundamental physics of the interactions between grain sizes, it remains, despite its experimental calibration, a much more powerful tool than the classical empirical model.

    – The theory of composite homogenization [DOR 05], with all the tools, both analytical and numerical, that constitute micromechanics today and provide the fundamental laws for linking compositional parameters (concentrations and intrinsic characteristics) of the constituents to macroscopic, mechanical and physical properties.

    – Physical understanding of capillary force mechanisms and their mechanical effects (shrinkage, creep) coupled with water exchange processes [ACK 04].

    – The theory of poromechanics [COU 04] which, together with the laws of thermodynamics, can deal with coupled processes (all couplings between mechanical, thermal, physical and chemical stresses).

    These tools are just about to step into day-to-day engineering practice. There is still an enormous amount of work ahead to translate them into engineering tools. If they were available to engineers, these tools would amount to true materials engineering – systematic performance optimization – whereas the current approach, based on the use of standards founded on essentially empirical approaches, entail unjustified and large safety margins that are not very reliable, incurring considerable cost.

    2.4. Current, economic, social and environmental challenges

    The challenges facing the construction sector today are well known. Construction accounts for between 35 and 40% energy expenditure and between 35 and 40%of greenhouse gases emissions. This is certainly the sector in which the most significant progress must be achieved over the next decade to meet the environmental challenges in the coming age of global warming. There is currently no other material on the horizon than cement-based concrete to meet the basic needs of humanity for building houses, schools, hospitals, transport infrastructures, etc., which already amounts to more than 20 billion tonnes per year. Concrete is and will be the back-bone material for the sustainable development of our society. No other building material (metal, wood, brick, etc.) is available in significant enough quantity on our planet – none of them can reach a billion tonnes per year for construction purposes. On the other hand, concrete has proven a unique versatility, not only in the level of each performance values:

    – compressive strength can vary from 5 to 200 MPa;

    – fluidity can be completely self-leveling as well as very rigid, as for extrusion processes;

    – density from less than 1 to 5;

    – thermal insulation can now be divided by three for an industrial product fitting the specification of concrete building projects;

    – porosity can be controlled for completely waterproof materials to a very porous one, as used for managing flood risks in dense urban regions;

    – thermal inertia, color, surface aspect, etc.

    The properties of concrete are also rendered very flexible by the very easy combination of all these properties.

    2.5. Conclusion

    Engineering science has been the driving force of a revolution in concrete technology, which is just about to enter the construction sector but which, if used at a massive scale, will be transformational, for instance in terms of considerable materials, energy, and thus CO2 emissions, savings. This trend already involves the development, validation and popularization of engineering tools – as seen in the development and implementation of UHPFRC; but it must also involve, without doubt the most difficult and lengthiest aspect, changes to standards, since current standards put a brake on exploiting these means of progress.

    The environmental challenges with which we are now faced – namely energy, natural resources, climate change, etc. – thus present many opportunities for the scientists and engineers to make a lasting impact on the benefit of our profession and society in large. As such, the development and implementation of UHPFRC may one day be seen as just the beginning of a new era of the truely sustainable development of concrete as a high-tech bulk material at the forefront of materials science and engineering that can be fine-tuned, not only to meet the classical strength, stiffness and durability requirements but also to contribute to reducing the environmental footprint of our concrete consumption. It will thus add a new green concrete value to the development of our society in terms of both economic growth and social progress. These three components – economic growth, social development and environmental protection – are the definition of sustainable development as set forth in the Rio Declaration on the Environment.

    2.6. Bibliography

    [ACK 04] ACKER P., Swelling, shrinkage and creep: a mechanical approach to cement hydration, Concrete Science & Engineering, vol. 37, p.237-243, 2004.

    [AFG 04] AFGC, Conception des bétons pour une durée de vie donnée des ouvrages (Design of the Concrete Mix for a given Service Life of the Construction), 2004.

    [COU 04] COUSSY O., Poromechanics, Wiley, West Sussex, England, 2004.

    [DEL 99] DE LARRARD F., Concrete Mixture Proportioning, A Scientific Approach, E&FN Spon, London, 421 p, 1999.

    [DOR 05] DORMIEUX L., ULM, F.-J. (Eds.), Applied Micromechanics of Porous Material, CISM Courses and Lectures, vol. 480. Springer, Wiley, New York, 2005.

    [RIC 99] RICHARD P. A new generation of UHP concrete: Ductal®. Damage resistance and micromechanical analysis, Proceedings of the 3d International RILEM Workshop (HPFRCC 3), Mainz, pp.101-111, 1999.

    [ULM 99] ULM F.J., ACKER P., LEVY M., The Chunnel-Fire: Thermal spalling due to chemoplastic softening in rapidly heated concrete. Part II: Analysis of concrete damage, Journal of Engineering Mechanics, vol. 125, no. 3, pp.283-289, 1999.

    [VER 03] VERNET C., UHPC microstructure and related durability performances – laboratory assessment and field experience examples, PCI/FHWA 3rd Symposium on HPC, Orlando, USA, 2003.

    Chapter written by Paul ACKER and Franz-Josef ULM.

    Chapter 3

    UHPFRC Development: Review of a Determining Application

    For almost 20 years, EDF (the French Electricity Company), the main electricity producer in France, had to face the extensive deterioration of the heat exchange packing elements of its nuclear power plant cooling towers. This has serious repercussions on efficiency losses and results in substantial expenses. Due to the aggressive and harsh environmental conditions within the towers and the restricting time constraints for renovation work due to outage schedules, some alternative materials had to be developed. This lead to ultra high performance fiber-reinforced concrete (UHPFRC) to pass from laboratory to industrial applications. The experience gained with the achievement of this first UHPFRC structural application in France is at the origin of the AFGC recommendations (French Civil Engineering Association).

    3.1. Introduction

    The 59 reactors of EDF’s thermal nuclear power stations (see Figure 3.1), most of them being of the pressurized water reactor (PWR) type, make use of oncethrough ocean water when in coastal areas, while operating with a closed system when using cooling water from nearby rivers. Riverside nuclear plant pollution controls, where discharge of radioactive materials is concerned, are drastic, which explains these preferred closed systems. Depending on the period and the power station, three types of cooling towers are found: counter-flow cooling towers, cross-flow cooling towers and forced air cooling towers. Cooling towers are devices operating according to the principle of evaporation, in which the water from the condenser is cooled by spraying on an air-to-water heat exchanger suspended by a reinforced concrete frame (see Figure 3.2).

    Figure 3.1. Locations of French nuclear power plants

    In 1991, the cross-flow cooling tower packing suspension frames of all the 900 MW and 1,300 MW PWR power plants begin to show premature wear marks. The extent of the damage could cause their collapse and, consequently, a drop in performance of the cold source and a loss of power of the stages concerned. The economical issue is considerable: a variation of one-tenth of a Kelvin degree during stage service life costs €3,000,000 (US$4,500,000). EDF, with an intervention first in the Saint-Laurent nuclear power plant (2x 900-MW PWR by the Loire River) and then in Chinon nuclear power plant (4x 900-MW PWR by the Loire River downstream), implemented a new approach to replace the faulty elements.

    Figure 3.2. Principle of cross-flow cooling. Pressurized water reactors except Creys-Malville. Blue: air flow (cold); and red: water

    Figure 3.3. Intermediate concrete portal frame that allows the thermal exchange device to be split. Stainless steel frame suspended form the hot water tank, on which upper laths are fixed. Middle in dark grey: RPC crossed beams structure simply supported on C35 reinforced concrete columns (in blue). Lower thermal exchange packing suspended from the RPC girders

    The retrofitting method presented in Figure 3.3 consists of strengthening the truss structure, then modifying the clamping system of the plastic laths and finally splitting the 20 m high packing into two superposed ones. Like in the initial configuration, the upper part is suspended from the hot water tank and the lower part is a portal frame resting on the foundations.

    The conditions within the packing elements are particularly severe. The presence of ice during winter can cause spalling of the concrete members (freeze-thaw cycles). Moreover, the chemical treatment applied to the condenser cooling water is aggressive for concrete and, in the case of Cattenom plant where raw water is naturally brackish, the aggression is particularly intense. The repair work therefore requires the development of specific materials.

    3.2. Context of repair works at Cattenom

    3.2.1. Cooling tower operating characteristics and environmental constraints

    Exposure of the inner systems of the cooling towers can be characterized as being relative to the waters surrounding the structural elements. The PH of the water needs to be below 6.5 (to avoid growth of bacteria like Legionella, amoebae, etc.). The cold water temperature range is from 0 to 4°C and the hot water temperature ranges from 5 to 55°C. Finally, the presence of Javel water (sodium hypochlorite) in the proportion of 50 g/m³ (as an antifungal) is noticeable. The initial water temperature ranges from 0 to 27°C and the addition of hydrochloric acid (as a flocculant) is to be noted. The condensate water is composed of clean water (acidic compared to concrete), and industrial pollutants: SOx, NOx and carbon dioxide.

    3.2.2. Soil characteristic conditions

    The size of the cooling tower foundations is determined for a number of loading cases. The maximum overload rating admitted by the soil characteristics is 8 tonnes per pile. The technical repair solution, therefore, must take this condition into account to avoid additional ground-work (micropiling and drilling of the cold water collecting basin), for which the time necessary for work completion would exceed the outage deadline.

    3.2.3. Renovation work conditions

    Cooling tower maintenance and renovation work, like other maintenance works in the plant, is carried out during outages. This planned unavailability time ranges from a few weeks for short outages to three months for longer ones. However, because of the money at stake, a one-year outage or even a delay in starting over cannot be considered. Consequently, in most cases, the repair alteration work cannot be completed within the outage and must be spread over several periods. Therefore, the consideration of outage times is a prerequisite condition for work planning.

    3.3. Possible solutions

    Considering these environmental, soil and time conditions, the recommendations given by the design phases for main beams are detailed in following subsections, either for a metal or concrete-based intermediate supporting structure.

    3.3.1. Metal frame

    The intermediate portal frame erected to sustain half of the metal packing must be made of URANUS stainless steel because standard stainless steel materials cannot withstand the oxidability of the river Moselle water. However, the cost of this solution is prohibitive and only a few parts of the Cattenom cooling tower are actually made of URANUS stainless steel.

    3.3.2. Concrete structures

    There are four concrete-based solutions, detailed in Table 3.1. These are graphically compared in Figure 3.4.

    Table 3.1. Concrete-based repair solutions

    Figure 3.4. Comparison between the different concrete solutions. Left vertical scale – vertical overload in tonnes (limit value: 8 t). Right vertical scale – beam height in m

    Pre-tensioned prestressed concrete could not be used here because of the aggressiveness of the ambient conditions. Figure 3.4 summarizes concrete solution features and proves that only the one using UHPFRC could meet the requirements. Several drawbacks, however, at first hindered the application of UHPFRC:

    – the mixes were still considered laboratory materials;

    – the structures were too recent to provide long-term feedback information. The Sherbrooke pedestrian bridge, for instance, dates back to 1997 [ADE 98];

    – there was no design code regarding UHPFRC yet, only the AFREM (French Branch of RILEM) recommendations for fiber-reinforced concrete dating back 1995 [ROS 95] were available;

    – their implementation a priori generated high-cost estimates.

    3.4. Chinon prototype

    For more than 30 years, research work on high performance concrete (HPC), then, ultra high performance concrete (UHPC) has continuously progressed with the erection of an ever greater number of large structures demonstrating their benefit.

    In France, some national research and development projects like Voies Nouvelles du Béton (VNB 1986-1990) and then BHP 2000 (1995-2001) have contributed to make these new materials known [BRA 05, MAL 92]. The use of HPC for engineering structures is becoming widespread, and finds particular applications in exceptional EDF structures like containment walls for nuclear power plant reactor buildings (60 MPa-HPC for Civaux power plant in 1988 [DEL 90] and 72 MPa-HPC for # 3 and # 4 units of Penly power plant in 1992). Despite this, it still struggles to find a place on the construction market for other applications. Neither the direct material gain, nor the other recognized benefits (life expectancy, resistance to corrosive agents) are compensating for the delicate phase of mix design perfecting at the start of the job or for the unchanging length of the production cycle (forming – reinforcement – casting).

    From the mid-90s, the French rules contained in the BPEL (Béton Précontraint aux Etats Limites) design code (design code for prestressed concrete structures based on the concept of limit states) were evolving toward 80- and even 100-MPa tensile strength classes [DEL 96,TOU 03]. As the EDF nuclear construction program was slowing down, new concrete was studied as a possible solution for the next generation of power plants.

    Before any decision regarding both the technical solutions and the materials could be taken, EDF asked the scientific department of Bouygues to develop a chemically-treated RPC 200® beam prototype. An appreciable number of laboratory studies conducted to examine this material, especially in terms of durability, were made available, their results forming the subject of several presentations at international congresses. However, the first in situ experiment was conducted during the renewal of part of thermal exchange devices in the Chinon plant in 1996.

    The first approach adopted for the design codes was presented during the International Symposium on High Performance and Reactive Powder Concretes in Sherbrooke in 1998 [CHA 98]. A bending test had been carried out on a 350 mm-high and 6 m-long beam in order to adjust the computation model using actual results. A 1 m-long I-shaped beam mock-up had also been cast for sampling cores within the web, upper and bottom flanges. The direct tensile tests, carried out on the samples according to the procedure defined by the AFREM (Association Française de Recherche et d’Études des Matériaux) [ROS 95], reveal scattered results. In fact, counting the fibers in an in-situ concrete specimen demonstrated the significant influence of the direction of flow during casting and of the mould wall effect. This observation can also account for the scattering of the direct tensile test results performed on the samples drilled from the mock-up. Consequently, the EDF’s safety coefficients were taken as being higher for beam designing.

    Finally, thanks to these results, the first experimental renewal work was completed within outage schedule time and, in 1996, EDF decided to call for tenders for the cooling tower renovation of the four Cattenom units on a UHPFRC structure basis.

    3.5. Material specifications

    The application design criteria and requirements used by EDF can be found in two internal documents ([CCT 92] and [CST 91]). They include:

    – minimum compressive strength before and after heat treatment (at 28 days: fc 28 j = 180 MPa);

    – minimum direct tensile strength Ft 28 = 8 MPa and post-peak tensile strength σfu = 5 MPa;

    – resistance to freeze/thaw: a loss of compressive strength smaller than 10% after 100 thermal cycles within the range -20 + 20°C);

    – void ratio smaller than 8% (for the UHPFRC mix including the fibers);

    – limited water permeability;

    – conservative minimum concrete cover thicker than 30 mm;

    – gacing quality in accordance with picture number 3 of EDF’s concrete inspection book in the prestressing cable area, including bubble depth lower than 2 mm.

    3.6. Codes

    In the absence of rules, EDF worked out additional technical specifications drawn from the results of the Chinon nuclear power plant experiment as an adaptation of the BPEL rules [REG 99] to up to 80 MPa prestressed concrete. The main adaptations concern: the constitutive law including delayed deformations (shrinkage, creep); UHPFRC Young’s modulus; the tensile strength value; and the shear force verifications.

    Design, qualification and quality control tests were also adapted. For suitability tests, the acceptance criterion based on the arithmetical means fcE, of three 28-day compression tests was fixed as follows:

    fcE ≥ fc28 +CE - (Cmean - 3S)

    fcE ≥ 1.15 fc28

    fc28 specified compressive strength (required characteristic value at 28 days)

    where:

    CE: mean compressive strength of the cement used for the test;

    Cmean: mean value of the compressive strength observed at the cement manufacture;

    S: standard deviation.

    For quality control of the application to an industrial facility, the three following conditions had to be satisfied for routine checks of conformity:

    S ≤ 10 MPa (instead of 3 MPa as usually admitted for traditional concrete);

    fc ≥ fc28 + 1.3 S (mean value of the compressive strength);

    fcl ≥ fc28 – 3 MPa (individual strength value).

    3.7. The Cattenom nuclear power plant renovation project

    The call for tender for the renewal work of the packing of the four units of Cattenom power plant was notified in September 1996. Two solutions were proposed: one with the heat-treated Bouygues RPC 180® (now known as DUCTAL®-FM) made using a concrete mixing plant equipped with a vertical mixer specially designed for high turbulent mixing; the other with an alternative UHPFRC material without heat treatment obtained using the Eiffage BSI 150® pre-mix.

    With the aim of supporting the development of new materials for the construction of future nuclear power plants in view, EDF decided that RPC180® would be used for three of the four plant units and that BSI 150® would be used for the last unit if the design and qualification tests were satisfactory. If not, then RPC 180® would be used instead.

    The RPC 180® qualification was obtained after all the tests had been performed on the nominal mix design and on the derivative ones to simulate inevitable component variations occurring when using a concrete mixing plant [BIR 98a]. Bending and shear tests were conducted on prototype I-shaped beams (Figure 3.5).

    Figure 3.5. RPC beam tests, 1996 at LCPC (Laboratoire Central des Ponts et Chausées): a) overview of the 6 m-long beam testing setup; b) bending crack development at mid-span; c) shear failure of the 2 m-long beam

    BSI 150® concrete was developed between September 1996 and March 1998. Validation was carried out on a 350 mm beam similar to that previously used for the validation of RPC180® in 1996 [BRA 00, DEL 90]. Bending and shear tests (see Figure 3.6) were also carried out on real-size 6 m-long I-shaped beams. These small beams are reinforced with prestressing strands only and have no transverse reinforcement. They are resting on 14 m-long 750 mm-high main beams (see Figure 3.7), thus constituting the intermediate truss supporting structure within the renewed cooling tower as represented in Figure 3.8.

    During the design and qualification phases, a 4 m-long scale model was fabricated under representative jobsite conditions to study the material’s in-situ characteristics. Six saw cuts were performed at an angle of 30° (Figure 3.9) and for each slice six 74 mm diameter samples were cored to be used for the AFREM direct tensile tests [ROS 95]. The results, like those obtained in 1996, were very heterogeneous. This can be attributed to the fiber distribution, which is affected by the placing mode of UHPFRC, and the wall effect; and to the AFREM test procedure, according to which the 74 mm diameter test samples are notched with a necked section of 15.2 mm², which is small and induces uncontrolled stress concentrations.

    Figure 3.6. BSI® beam tests, 1998: a) Bending test at CSTB; b) shear test at LCPC [BIR 98b]

    Figure 3.7. Renovation of cooling towers at Cattenom power plant: a) cross-section of main beams (14.3 m-long); b) 6 m-long crossbeams

    Figure 3.8. UHPFRC beams location after renovation work

    These observations have been taken into account by the members of the AFGC Working Group in charge of the UHPFRC recommendations for the revision of their characterization procedure. They recommend carrying out the three-point bending test on notched samples rather than a delicate direct tensile test.

    Finally, at the conclusion of the testing phase, it was decided to carry out renewal work of the #4 unit using BSI 150® concrete, thus giving a first reference for industrial application in France for both UHPFRC materials.

    Figure 3.9. Mockup for characterization of BSI 150® tensile properties as placed in the structure

    3.8. Quality control of the production process

    In total, 270 14 m-long and 0.75 m high main beams (see Figure 3.10) and 2,376 6 m-long and 0.35 m-high small beams were manufactured. Thus a total volume of 1,650 m³ of UHPFRC was cast, with a daily production of 6.5 m³. As specified, quality controls were performed during the beam production process. At the same time, two control beams were cast and a preload was applied up to the theoretical cracking limit of the tensioned part. These beams have been used for further durability monitoring [TOU 10].

    The results of the compressive-strength tests, as illustrated Figure 3.11, were more scattered than those obtained with usual concrete or even with HPC. This result may be attributed to the quality of both the preparation of the test samples’ end surface and the centering of the samples on the testing machine plates. Consequently, the acceptance criteria of the control results had to be revised and the value of 10 MPa corresponding to the accepted standard deviation initially considered in the specifications had to be replaced, on AFGC Working Group’s recommendation, by a value equal to 10% of the characteristic value. Young’s modulus results also exhibited a high scatter (see Figure 3.12).

    Production of the beams and erection inside the cooling towers of Cattenom nuclear power plant is illustrated Figure 3.13.

    Figure 3.10. Some of the BSI 14 m-long main beams (Hürks-Eiffage document)

    Figure 3.11. Distribution of the results of the 155 quality control compressive strength tests

    Figure 3.12. Distribution of the results of the 150 control tests for the determination of Young’s modulus

    Figure 3.13a. Renovation work of the cooling towers of Cattenom nuclear power plant – BPR 200® beam precasting site

    Figure 3.13b. Renovation work of the cooling towers of Cattenom nuclear power plant – erection of the main beams

    Enjoying the preview?
    Page 1 of 1