Doctoral Thesis in Civil and Architectural Engineering
Multiphase models for freeze-thaw actions and mass transport in
concrete hydraulic structures
DANIEL ERIKSSON
Stockholm, Sweden 2021
kth royal institute
of technology
Multiphase models for freeze-thaw actions and mass transport in
concrete hydraulic structures
DANIEL ERIKSSON
Doctoral Thesis in Civil and Architectural Engineering KTH Royal Institute of Technology
Stockholm, Sweden 2021
Academic Dissertation which, with due permission of the KTH Royal Institute of Technology, is submitted for public defence for the Degree of Doctor of Philosophy on Thursday the 3rd of June 2021, at 10:00 a.m. in F3, Lindstedtsvägen 26, Stockholm.
© Daniel Eriksson ISBN 978-91-7873-832-8 TRITA-ABE-DLT-2111
Printed by: Universitetsservice US-AB, Sweden 2021
Abstract
A crucial task for civil engineers is to make appropriate designs of new concrete structures and assessments of existing structures to ensure a long service life and sustainable use of the infrastructure. This doctoral thesis aims to increase the understanding of how advanced mathematical models can be used to describe phenomena and processes governing concrete degradation and thereby ultimately contribute to improving tools for design and assessments. The focus is on degrada- tion processes that cause commonly observed concrete damage types in hydraulic structures exposed to cold climates and soft water. During a structure’s service life, it is subjected to various deteriorating actions, but for the typical exposure conditions considered in this work, degradation due to freeze-thaw exposure and calcium leaching is of particular concern for the durability. Hence, the work re- lated to improved modelling has been focused on phenomena related to these two degradation processes of concrete and how they may interact to produce damaging synergy effects.
All developed models in this doctoral project treat concrete as a multiphase porous medium and use poromechanics to describe the coupled hygro-thermo-mechanical behaviour of the material. Moreover, since the overall aim concerns degradation in hydraulic structures, the model development has focused on obtaining formula- tions applicable for structural-scale simulations. The models presented in this thesis describe long-term water absorption into air-entrained concrete and the response of partially saturated air-entrained concrete exposed to freeze-thaw conditions. In the latter models, the phase changes and the freeze-thaw hysteresis are explicitly considered in the formulations. The presented simulation examples are performed using the Finite Element Method (FEM), and the capabilities of the models are verified with experimental data from the literature. Additionally, accelerated leach- ing experiments on air-entrained concrete are presented, where the influence of leaching on the formation and melting of ice inside the pore space due to pore structure alternations are investigated.
The main research contribution of this work is the development and evaluation of advanced models applicable for structural-scale simulations that describe essential processes and phenomena related to freeze-thaw exposure of air-entrained con- crete. The experimental work shows the significant influence of calcium leaching on the freeze-thaw processes, and the results can also facilitate future development of models considering some of the interactions causing damaging synergy effects.
Adopting a multiphase modelling approach has been found suitable for describing
the coupled processes and including interactions between different deterioration
mechanisms. The theoretical models can also help gain further insights and im-
prove the understanding of the phenomena, and thus, e.g. aid in developing more
simplified models suited for daily engineering applications.
Sammanfattning
En viktig uppgift för anläggningsingenjörer är att utforma nya ändamålsenliga betongkonstruktioner och göra korrekta tillståndsbedömningar av befintliga kon- struktioner för att säkerställa en lång livslängd och därmed hållbart nyttjande av vår infrastruktur. Syftet med denna doktorsavhandling är att förbättra kunskaps- läget kring hur avancerade matematiska modeller kan användas för att beskriva de fenomen och processer som styr betongens nedbrytning och därigenom bidra till förbättrade verktyg som kan användas vid dimensionering och tillståndsbe- dömningar. Arbetet fokuserar på de nedbrytningsprocesser som leder till vanligt förekommande skador i vattenbyggandskonstruktioner som är exponerade för kal- la klimat och mjukt vatten. Under en konstruktions livslängd utsätts den för ett flertal olika nedbrytningsprocesser, där frysning och tining samt kalkurlakning är av särskilt intresse för beständigheten givet de typiska exponeringsförhållanden som beaktas i detta arbete. Arbetet avseende förbättrad modellering har därför fokuserat på fenomen som är relaterade till dessa två nedbrytningsprocesser av betong och hur de samverkar för att skapa skadliga synergieffekter.
Samtliga modeller som utvecklats inom detta doktorandprojekt baseras på en multifasbeskrivning av betong som ett poröst material samt poromekanik för att beskriva det kopplade hydro-termo-mekaniska materialbeteendet. Eftersom det övergripande målet avser nedbrytning i vattenbyggnadskonstruktioner har modell- utvecklingen fokuserat på modellformuleringar som kan användas för simulering på strukturskala. De modeller som presenteras i den här avhandlingen beskriver långtidsabsorption av vatten i lufttillsatt betong samt responsen hos delvis vatten- mättad lufttillsatt betong exponerad för frysning och tining. I de senare modellerna inkluderas fasomvandlingar samt hysteresen vid frysning och tining explicit i mo- dellformuleringarna. De presenterade simuleringsexemplen är genomförda med finita elementmetoden och modellernas beteende har verifierats med experimen- tella resultat från litteraturen. Dessutom presenteras accelererade urlakningsexpe- riment på lufttillsatt betong där urlakningens inverkan på isbildning och smältning i porsystemet på grund av förändringar i porstrukturen studerades.
Avhandlingens huvudsakliga forskningsbidrag är utveckling samt utvärdering av
avancerade modeller avsedda för simulering på strukturskala och som beskriver
viktiga processer och fenomen relaterade till frysning och tining av lufttillsatt be-
tong. Det experimentella arbetet visar på den betydande inverkan av kalkurlakning
på frysning- och tiningsprocesserna, där resultaten även kan underlätta fortsatt
modellutveckling där några av de samverkansmekanismer som orsakar skadliga sy-
nergieffekter beaktas. Multifasmodellering har visats vara lämpligt för att beskriva
de kopplade processerna samt för att inkludera samverkan mellan olika nedbryt-
ningsmekanismer. De teoretiska modellerna kan också bidra till ökad insikt och
förståelse av dessa fenomen. Därigenom kan de till exempel bidra till utvecklingen
av mer förenklade modeller som är anpassade för vanliga ingenjörstillämpningar.
Preface
The research presented in this doctoral thesis was carried out during 2015 to 2021 at the Division of Concrete Structures, Department of Civil and Architectural Engineering, KTH Royal Institute of Technology in Stockholm, Sweden. The re- search was made possible through the financial support provided by the Swedish Hydropower Centre (SVC) and their support is gratefully acknowledged.
I would like to express my gratitude and sincere thankfulness to my main supervisor Prof Anders Ansell for his support and guidance during the work with this thesis. I also wish to express my gratitude towards my co-supervisor Dr Richard Malm for initiating the project and encouraging me to start my doctoral studies as well as for his support and advice along the way. Moreover, I would like to thank my second co-supervisor adjunct Prof Erik Nordström for his advice. In addition, I owe Dr Katja Fridh my gratitude for helping me performing experiments, our discussions as well as for her support and encouragements. A special thanks also goes to my former colleagues, adjunct Prof Manouchehr Hassanzadeh at Sweco, for providing me with pictures and for his advice, and Dr Tobias Gasch, for our discussions and collaboration. I would also like to thank Dr Romain Balieu for taking the time to review the thesis. Furthermore, a thank goes to my colleagues at the Department of Civil and Architectural Engineering.
Lastly, I would like to thank my family and friends, and especially my beloved partner in life, and soon to be wife, Karin for enduring all my late nights at the office and her endless resource of positive support.
Stockholm, April 2021
Daniel Eriksson
Funding Acknowledgement
The research presented in this thesis was carried out as a part of "Swedish Hy- dropower Centre - SVC". SVC has been established by the Swedish Energy Agency, Energiforsk and Svenska Kraftnät together with Luleå University of Technology, KTH Royal Institute of Technology, Chalmers University of Technology and Uppsala University.
Participating companies and industry associations are: Andritz Hydro, Boliden, Fortum Generation, Holmen Energi, Jönköping Energi, Jämtkraft, Karlstads En- ergi, LKAB, Mälarenergi, Norconsult, Rainpower, Skellefteå Kraft, Sollefteåforsens, Statkraft Sverige, Sweco Energuide, Sweco Infrastructure, Tekniska verken i Linköp- ing, Uniper, Vattenfall R&D, Vattenfall Vattenkraft, Voith Hydro, WSP Sverige, Zink- gruvan and ÅF Industry.
vi
List of Publications
This doctoral thesis consists of four journal papers and one peer-reviewed confer- ence paper. Throughout the thesis, these are referred to by their Roman numerals:
Paper I: Eriksson, D., Gasch, T. and Ansell, A. (2019). A hygro-thermo-mechanical multiphase model for long-term water absorption into air-entrained con- crete. Transport in Porous Media, 127, pp. 113-141.
Paper II: Eriksson, D. and Gasch, T. (2018). Influence of air voids in multiphase modelling for service life prediction of partially saturated concrete. In:
Proceedings of the Conference on Computational Modelling of Concrete and Concrete Structures (EURO-C 2018). Bad Hofgastein, Austria, February 26th to March 1st 2018, pp. 317-326.
Paper III: Eriksson, D., Gasch, T., Malm, R. and Ansell, A. (2018). Freezing of partially saturated air-entrained concrete: A multiphase description of the hygro-thermo-mechanical behaviour. International Journal of Solids and Structures, 152-153, pp. 294-304.
Paper IV: Eriksson, D., Wahlbom, D., Malm, R. and Fridh, K. (2021). Hygro-thermo- mechanical modeling of partially saturated air-entrained concrete con- taining dissolved salt and exposed to freeze-thaw cycles. Cement and Concrete Research, 141, 106314(1-15).
Paper V: Eriksson, D., Fridh, K. and Malm, R. (2021). Influence of calcium leach- ing on ice formation in air-entrained concrete: Accelerated experiments and hygro-thermo-mechanical modelling. Submitted for review, March 2021.
The author of the thesis (DE) wrote all appended papers, whereas the co-authors,
if otherwise not mentioned below, contributed by planning the work, discussing
the results and assisted with comments on the writing. DE collaborated with Gasch
to develop and implement a set of general governing equations used as a basis
for the model development in Papers I to IV. DE made the specific adaption and
necessary modifications of the governing equations for the models developed in
the papers as well as the development and implementation of the constitutive
relationships. All simulations in the papers were performed by DE, except in Paper II
where Gasch performed the simulations. In Paper V, the experiments were managed
and performed by Fridh. In addition to writing the paper, DE contributed to the planning of the work, processed and analysed the obtained data, and made the presentations of the experimental results.
Other relevant contributions by DE within the scope of this research project, but not included in the thesis, are:
– Gasch, T., Eriksson, D. and Ansell, A. (2019). On the behaviour of concrete at early-ages: A multiphase description of hygro-thermo-chemo-mechanical prop- erties. Cement and Concrete Research, 116, pp. 202-216.
– Eriksson, D., Malm, R. and Hellgren, R. (2019). Assessment of frost damage in hydraulic structures using a hygro-thermo-mechanical multiphase model. In:
Sustainable and Safe Dams Around the World - Proceedings of the ICOLD 2019 Symposium at the 87th ICOLD Annual Meeting, Ottawa, Canada, June 9th to 14th 2019, pp. 332-346. *Reference [63]
– Hellgren, R., Malm, R. and Eriksson, D. (2019). Modelling of the ice load on a Swedish concrete dam using semi-empirical models based on Canadian ice load measurements. In: Sustainable and Safe Dams Around the World - Proceedings of the ICOLD 2019 Symposium at the 87th ICOLD Annual Meeting, Ottawa, Canada, June 9th to 14th 2019, pp. 3068-3080.
– Nordström, E. and Eriksson, D. (2019). Inventering av inre vattenvägsbesikt- ningar [Review of performed inspections in waterways of tunnel-type]. Report 2019:566, Energiforsk, Stockholm, Sweden. *Reference [180]
– Eriksson, D. and Malm, R.(2018). Simulering av frostsprängning i betong [Mod- elling of frost actions in concrete]. Bygg och Teknik, 7, pp. 19-22.
– Eriksson, D. (2018). Numerical models for degradation of concrete in hydraulic structures due to long-term contact with water. Licentiate thesis, KTH Royal Institute of Technology, Stockholm, Sweden. *Reference [62]
– Eriksson, D., Gasch, T. (2017). Comparison of mechanistic and phenomenologi- cal approaches to model drying shrinkage of concrete. In: Proceedings of the XXIII Nordic Concrete Research Symposium. Aalborg, Denmark, August 21st to 23rd 2017, pp. 287-290.
viii
Contents
1 Introduction 1
1.1 Background . . . . 2
1.2 Aims of the thesis . . . . 4
1.3 Methodology . . . . 5
1.4 Limitations . . . . 6
1.5 Outline of the thesis . . . . 7
2 Degradation of concrete in hydraulic structures 9 2.1 Temperature conditions typical for cold climates . . . . 9
2.2 Deterioration mechanisms . . . . 11
2.2.1 Temperature cracks . . . . 12
2.2.2 Frost damage . . . . 14
2.2.3 Erosion . . . . 16
2.2.4 Calcium leaching . . . . 18
2.2.5 Alkali-aggregate reactions . . . . 19
2.3 Observed damage in tunnel-type waterways . . . . 21
2.4 Synergy effects . . . . 24
2.4.1 Surface damage . . . . 25
2.4.2 Internal damage . . . . 26
2.5 Summary and relation to performed work . . . . 27
3 Moisture in concrete 31 3.1 Absorption and desorption . . . . 31
3.1.1 Types of pores . . . . 31
3.1.2 Moisture fixation . . . . 32
3.1.3 Absorption of free water . . . . 34
3.1.4 Hysteresis . . . . 36
3.2 Transport of water . . . . 37
3.3 Long-term water absorption . . . . 38
3.3.1 Long-term water filling mechanism . . . . 38
3.3.2 Models for long-term water absorption . . . . 39
4 Freezing and thawing of concrete 43 4.1 Ice formation and melting in porous networks . . . . 43
4.1.1 Equilibrium between ice and liquid water inside pores . . . 43
4.1.2 Volume of formed ice . . . . 47
4.1.3 Freeze-thaw hysteresis . . . . 49
4.1.4 Modelling freeze-thaw induced phase changes . . . . 52
4.1.5 Pressure exerted on pore walls . . . . 58
4.2 Frost deterioration theories . . . . 59
4.2.1 Hydraulic pressure theory . . . . 59
4.2.2 Microscopic ice lens growth and cryo-suction . . . . 61
4.2.3 Macroscopic ice lens growth . . . . 63
5 Leaching of concrete 65 5.1 Process of calcium leaching in concrete . . . . 65
5.2 Alternation of microstructure and material properties . . . . 69
5.2.1 Pore size distribution . . . . 69
5.2.2 Mechanical properties . . . . 71
5.2.3 Mass transport properties . . . . 73
5.2.4 Coupled degradation . . . . 76
5.3 Accelerated testing of leaching . . . . 78
5.3.1 Aggressive solution . . . . 78
5.3.2 Electrochemical methods . . . . 80
6 Multiphase models of concrete 83 6.1 Background and definitions . . . . 83
6.2 Macroscopic governing conservation equations . . . . 86
6.2.1 Mass balance . . . . 87
6.2.2 Momentum balance . . . . 88
6.2.3 Energy balance . . . . 88
6.2.4 State variables . . . . 89
6.2.5 Boundary conditions . . . . 91
6.3 Constitutive relationships . . . . 92
6.3.1 Equation of state . . . . 93
6.3.2 Sorption equilibrium and isotherms . . . . 94
6.3.3 Mass and heat flux . . . . 95
6.3.4 Stress tensors . . . . 96
6.4 Example of couplings in multiphase models . . . . 98
6.5 Finite element discretization of the PDEs . . . 101
7 Summary of appended papers 107 7.1 Paper I . . . 107
7.2 Paper II . . . 109
7.3 Paper III . . . 110
7.4 Paper IV . . . 112
7.5 Paper V . . . 113
8 Discussion 117 8.1 Varying exposure conditions . . . 117
8.2 Multiphase models and their applicability . . . 119
8.3 Modelling synergy effects . . . 121
x
9 Conclusions and suggestions for future work 125 9.1 Conclusions . . . 125 9.2 Suggestions for future work . . . 127
Bibliography 131
List of Notations 155
Chapter 1 Introduction
Concrete structures of various types are essential for the infrastructure in all mod- ern societies around the world. Many of these structures were designed and built decades ago and are, thus, approaching their designed service life, but are often still functioning rather well. Hence, it may often be suitable to repair and rehabili- tate these structures instead of demolishing them and building new ones to meet the requirements of sustainability. However, to extend the service life, some of the structures may need substantial repair work, whereas others might function well beyond the designed service life with relatively small efforts. Being able to assess the remaining service life of existing structures is thus of significant importance to make an information-based decision on what measures to take in each specific case.
Moreover, making such assessments is, of course, also essential to design and build new durable concrete structures that meet the societal demand for sustainability.
During the service life of concrete structures, they are usually exposed to different
potentially deteriorating actions such as mechanical loads, varying environmen-
tal conditions and other physical and chemical phenomena. It is typically a com-
bined effect of the exposure conditions and the type and quality of the concrete
that determines the rate of deterioration in a specific structure. When designing
new structures or performing an assessment of the residual service life of an ex-
isting structure, it is important to consider all processes and aspects that might
influence the degradation. Hence, to obtain an adequate and accurate assessment
of the service life, the analysis tools should be capable of considering the cou-
pled hygro-thermo-chemo-mechanical behaviour of the material. During recent
decades, the use of multiphysics to describe the coupled physical and chemical
behaviour of porous materials has made significant advancements and is becoming
an increasingly adopted approach to model different processes in concrete, see
e.g. [85, 90, 91, 95, 105, 143, 155, 186, 218]. The porous structure of the con-
crete is often considered by adopting a multiphase description, where the different
phases constituting the material are treated separately in the multiphysical model
formulation. Concrete is typically divided into a solid phase, corresponding to its
solid skeleton that constitutes the interconnected pore space, and one or several
interstitial fluid phases that occupy the porous network. Each phase may also be
divided into several species depending on the processes and phenomena one aims
CHAPTER 1. INTRODUCTION
Water-retaining wall
Water-retaining dam
Headrace tunnel
Penstock
Powerhouse
Upstream reservoir
Downstream outlet Draft tube
Figure 1.1: Schematic illustration of a cross-section through a hydropower plant.
to describe with the multiphase model. Through adopting this modelling approach, different deterioration mechanisms can be studied theoretically and also aid to establish a better understanding of the interactions between different processes.
Nevertheless, it should be pointed out that experimental studies still are required and essential to verify and validate these models. The use of multiphase models may not always be suitable for practical engineering work. However, the knowl- edge gained by adopting more advanced models to study the different deterioration processes can serve as a basis for the development of improved but more simplified analysis tools.
1.1 Background
Hydraulic structures are built to regulate the natural flow of water in, e.g. rivers, lakes or seas. A hydropower plant consists of several hydraulic structures, including dams and waterways, which are used to control and redirect the flow of water to power turbines to produce electrical energy. Figure 1.1 shows a schematic illus- tration of a typical cross-section through a hydropower facility and includes some common types of hydraulic structures. Hydropower facilities are still being built all over the world, but in Europe and North America, the majority of the larger hydropower dams were commissioned during the early and mid-20th century. For example, in Sweden, most of the 2000 hydropower plants currently in operation were commissioned during the early and mid-20th century, where the larger plants mainly were built between 1945 and 1975 [154, 200]. In the United States, the average age of the dams was 57 years as of 2021, and by year 2030 approximately 70 % of the dams will be over 50 years old [3]. Furthermore, according to Pardo- Bosch and Aguado [183], more than 40 % of the large hydropower dams in the world are older than 40 years. Hence, many hydraulic structures at hydropower facilities all around the world are starting to approach their designed service life.
Some of these, but primarily the oldest structures, have also started to show signif-
2
1.1. BACKGROUND
icant age-related wear and degradation [127].
The moisture state inside concrete often has a significant influence on the type and rate of degradation, and thus also for the durability of the structures. Since hydraulic structures typically are fully or partially submerged in water, the degree of moisture saturation in the structures is generally relatively high. The suscep- tibility to frost damage in concrete is, e.g., closely linked to the moisture state and is almost fully mitigated if the degree of saturation is below a certain critical level. Moreover, sufficiently moist conditions are also required by other degradation mechanisms to cause a deteriorating effect inside structures, such as the swelling effect caused by the alkali aggregate reactions (AAR). The exposure to an external reservoir containing soft water also leads to leaching of calcium ions from the solid compounds in the concrete, which results in a weakening of the material and also makes it more susceptible to other deterioration mechanisms. Furthermore, the flowing water usually carries different amounts of sediments and other debris that have an abrasive effect and may, therefore, erode the concrete surfaces. Substantial cracking can also occur due to e.g. extreme loads, seasonal temperature variations or other environmental effects.
Hydraulic concrete structures are, though, usually exposed to several different de- terioration mechanisms either simultaneously or sequentially as the exposure con- ditions change between the seasons. Furthermore, it is well-known that commonly observed types of damage in hydraulic structures typically are a consequence of syn- ergy effects between these degradation processes [57, 80, 119, 125, 127, 147, 203].
In this context, synergy is referring to the interactions between two or more mech- anisms that together yield a larger degrading effect than the sum of their individ- ual effects. The combination of degradation processes varies depending on the exposure conditions, the type of hydraulic structure and also between different regions in the same structure. In a constructed waterway of channel-type or a water-retaining dam located in regions with cold climates, exposure to seasonal temperature variations, frost actions, leaching and erosion is rather common. How- ever, in a waterway of tunnel-type, e.g. a penstock, the common exposure might instead consist of high water pressures, leaching and erosion due to high water velocities. Both physical and chemical phenomena cause the interactions between the different mechanisms, and it is challenging to establish exactly how they inter- act and thus also to model their combined effect. Furthermore, the interactions are usually not one-way couplings, which in the case of two active deterioration mecha- nisms means the degradation rate of one mechanism is accelerated by the other one and vice versa. For hydraulic structures exposed to cold climates and in long-term contact with soft water, which e.g. corresponds to typical Swedish exposure con- ditions, the combination of calcium leaching and frost actions has been shown to be particularly essential for the durability [60, 120, 127, 163, 200, 202, 203]. The leaching process leads to a weakened material, and an increased porosity, which means the concrete becomes more susceptible to frost damage. During subsequent freezing conditions, the frost actions further weaken the material and increase both the permeability and porosity of the concrete, which promotes the leaching process.
In this way, the two processes accelerate each other and progressively deteriorates
CHAPTER 1. INTRODUCTION
the concrete material though their combined deteriorating effect.
Being able to perform adequate assessments of the remaining service life of dam- aged hydraulic structures is important to ensure the dam safety and continuing operation of the hydropower plants. More specifically, the results from such as- sessments should form the basis for deciding if, when, and how repair works are necessary to extend a hydraulic structure’s service life or if a new structure is re- quired to ensure future safe operation of the power plant. Most of the methods and manuals used today for this type of assessments require that a dominating deterioration mechanism can be identified, see e.g. the manuals developed in the European innovation project CONTECVET [72, 74, 199, 229]. However, to perform more accurate assessments, it is as already indicated in many cases also necessary to consider the interactions between the degradation processes. The development of advanced mathematical models describing the different degradation processes can help improve the understanding of their role and how they may interact to produce synergy effects. The results from such theoretical studies can also be used as a basis to establish new and more accurate assessment tools for the long-term performance of concrete in hydraulic structures exposed to various loading and environmental conditions.
1.2 Aims of the thesis
The overall aim of this project is to improve the knowledge of how advanced mathe- matical models can be used to describe the physical and chemical processes leading to degradation of concrete in hydraulic structures exposed to cold climates and soft water. The work also includes compiling and reviewing commonly observed concrete damage types in hydraulic structures at hydropower facilities located in cold regions, and describe the typical deterioration mechanisms causing the damage and how these may interact. However, as already indicated, freeze-thaw actions and calcium leaching have been found to be particularly essential for the durability of concrete in hydraulic structures subjected to cold temperatures and soft water. Hence, the more specific aim of the work related to modelling is to improve the knowledge of how to consider phenomena related to these two degra- dation processes. To this end, the work includes to review, develop, implement and validate mathematical models based on state-of-the-art modelling techniques and using findings from both theoretical and experimental studies. Since the overall aim concerns concrete degradation in hydraulic structures, it is also essential that the developed models are applicable for structural-scale simulations. Moreover, the aim is to adopt a modelling framework that is suitable for also considering interactions between different deterioration mechanisms. The research questions addressed in this thesis can be summarised as:
1. What concrete damage types and deterioration mechanisms are commonly observed in hydraulic structures located in cold regions, and where and in what type of structures are the different damage types typically observed?
4
1.3. METHODOLOGY
2. How do the moisture state inside the concrete and different mass transport processes in the pore space influence the degradation?
3. Which phenomena related to the degradation processes are essential to con- sider and how can they be modelled on the structural scale?
4. What mathematical framework is suitable for modelling the coupled physical and chemical processes leading to concrete degradation?
5. How do the deterioration mechanisms interact, and how can the major inter- actions be considered in a model?
1.3 Methodology
The adopted methodology to address and answer the research questions consists of three main activities: literature studies, model development and experimental work. This section outlines and briefly summarises the performed steps of the work to answer the questions.
Initially, a literature study was conducted of commonly reported concrete damage types in hydraulic structures exposed to cold climates and soft water, including where they usually can be expected to occur in the structures. The typical deterio- ration mechanisms causing the damage and how they may interact to produce syn- ergy effects are also addressed and described. Based on this review and to limit the scope of considered degradation types, the following modelling-related activities were focused on the processes identified as particularly essential for the durability of concrete in hydraulic structures subjected to these exposure conditions. More specifically, and as indicated above, processes related to freeze-thaw exposure and leaching have been the primary focus in this work. Hence, more in-depth literature reviews related to these two degradation processes were conducted to identify the essential phenomena that need to be considered in the model development.
This work also included reviewing earlier published models to identify the needs for further development of models intended for structural-scale simulations of the important phenomena related to the degradation processes.
The development of models intended for structural-scale simulations in this work has been focused on processes identified as essential for freeze-thaw exposure. In this project, models with increasing complexity have gradually been developed and validated to establish a suitable modelling framework for the considered phe- nomena. As a first step, a model was developed and validated for simulation of the long-term moisture state inside air-entrained concrete to improve the capa- bility of assessing the moisture state inside water-retaining structures accurately.
In a second step, using the same mathematical framework, a model describing
the mechanical response of partially saturated air-entrained concrete exposed to
freezing conditions was developed. This model was subsequently further extended
to include the freeze-thaw hysteresis for the formation and melting of ice inside
CHAPTER 1. INTRODUCTION
the pore space and the influence of dissolved salt in the pore liquid. The models were primarily validated using experimental data from the literature.
From the knowledge gained by developing and validating the models, essential data needed to develop models considering one of the major couplings between freeze-thaw actions and calcium leaching could be identified. Hence, to facilitate future development of models considering this coupling effect, experiments were performed to investigate and measure the influence of leaching on the formation and melting of ice inside the pore space of air-entrained concrete. The work also examines whether the adopted experimental method is suitable to establish the rel- evant input data by utilising the results in simulations where one of the previously developed freeze-thaw models is used.
1.4 Limitations
The degradation processes covered in this thesis are focused on those causing common types of concrete damage in hydraulic structures at hydropower facilities located in cold regions and in contact with soft water. More specifically, the work has been primarily directed towards studying processes and phenomena related to freeze-thaw exposure and calcium leaching of concrete, which usually are the dominating processes in, e.g. typical Swedish conditions. Hence, other degradation processes that in some cases might also be important for the durability are only briefly covered, such as erosion and alkali-silica reactions (ASR). Although the covered aspects are applicable for hydraulic structures in all regions with cold climates, it should be noted that local conditions will influence which degradation processes are most essential for the overall durability.
The mathematical models developed and presented in this work are limited to essential processes and phenomena related to freeze-thaw exposure. As mentioned above, the model development has aimed to obtain model formulations applicable for structural-scale simulations. Although the focus herein is on hydraulic struc- tures, the models are general and can be applied to other types of concrete struc- tures subjected to similar exposure conditions. Many of the presented models in the literature for calcium leaching are already applicable and validated for structural- scale simulations. This is the reason why the modelling work was focused on processes related to freeze-thaw. Furthermore, the development of models that explicitly includes interactions between freeze-thaw actions and calcium leaching is outside the scope of this thesis. However, to facilitate future model development, it is still essential to improve the understanding of how single deterioration mecha- nisms can be modelled on the structural scale in a mathematical framework that is suitable and allows for the implementation of these interactions. In this work, all models have been developed based on poromechanics and a multiphase description of concrete as a porous medium. The performed experimental work is limited to investigating the influence of calcium leaching on the formation and melting of ice inside the pore space due to the pore structure alternations. The tested materials
6
1.5. OUTLINE OF THE THESIS
were limited to an air-entrained and a non-air-entrained concrete, where the only difference between the mixtures was the addition of an air-entrainment agent.
1.5 Outline of the thesis
This compilation thesis consists of five appended scientific papers, denoted I-V, together with an introductory part. The latter puts the research presented in the scientific papers into a wider context, summarises and discusses the results of the work, and presents its overall conclusions. In addition, it also provides some further theoretical background to the models and methods that are developed and used in the papers.
Chapter 2 of the introductory part presents a literature study of commonly ob- served damage types in hydraulic structures at hydropower facilities exposed to cold climates and in contact with soft water. The chapter also briefly describes the typical deterioration mechanisms causing the damage and addresses how they may interact to cause damaging synergy effects. Additionally, the results of a sur- vey of reported damage types in waterways of tunnel-type are summarised. In Chapter 3, the fundamentals of moisture fixation and water transport in concrete are described with particular emphasis on the long-term absorption of water into air voids. Chapter 4 describes and discusses the freezing and thawing processes of liquid solutions inside porous media and the role of air voids in air-entrained concrete during freezing conditions. Furthermore, three fundamental frost dete- rioration theories of concrete are briefly described. Calcium leaching in concrete is addressed in Chapter 5 and emphasises the alterations of the microstructure and material properties caused by the leaching as well as accelerated experimental techniques. Chapter 6 presents the set of general balance equations for a generic porous medium which was used as a basis when deriving the multiphase models presented in the appended papers. The chapter also discusses the terms in the general equations that are usually neglected when modelling concrete, some con- stitutive relationships and the numerical solution of the models using the finite element method (FEM). This is followed by Chapter 7, which presents a summary of the five appended papers:
– Paper I is a published, peer-reviewed journal paper. The paper presents a mul- tiphase model describing the long-term absorption of water in air-entrained concrete and is adopted in two examples to verify the model and to show its capabilities.
– Paper II is a published, peer-reviewed conference paper. The paper further
investigates the capabilities of the developed model in Paper I for simulations
on the structural scale. The model is adopted in an example studying the
long-term moisture conditions in a typical water-retaining wall built with
air-entrained concrete and located in a waterway of channel-type.
CHAPTER 1. INTRODUCTION
– Paper III is a published, peer-reviewed journal paper. The paper presents a coupled hygro-thermo-mechanical multiphase model for simulation of freezing- induced deformations in partially saturated air-entrained concrete, and is ap- plicable for structural scale simulations. Two examples are solved using the model to verify its behaviour and show its capabilities.
– Paper IV is a published, peer-reviewed journal paper. This paper further de- velops the model proposed in Paper III by extending the formulation also to consider the ice formation hysteresis in freeze-thaw cycles and the influence of dissolved salt in the pore liquid. The developed model is verified using experimental data from a previous study by one of the co-authors, whereas some additional capabilities of the proposed formulation are investigated in a second example.
– Paper V is submitted to a scientific journal and currently under review. The paper presents an experimental study investigating the influence of calcium leaching on the formation and melting of ice inside the pore space of air- entrained concrete caused by the pore structure alternations. The experi- mental results are also used in a simulation example aiming to illustrate the influence of leaching on the hygro-thermo-mechanical response of the tested concrete during freeze-thaw exposure.
The main results and findings of the appended papers are discussed in Chapter 8, whereas Chapter 9 presents the conclusions of the work together with suggestions for future research.
As customary at the School of Architecture and the Built Environment at KTH Royal Institute of Technology, parts of this doctoral thesis were previously published as a licentiate thesis [62].
8
Chapter 2
Degradation of concrete in hydraulic structures
Several different deterioration mechanisms can cause concrete degradation in hy- draulic structures. The degradation rate depends on various factors, such as the material quality, the quality of the construction works and the exposure conditions at the site of the structure. Examples of the latter are the ambient temperature conditions, the chemical composition of the water, the amount and type of sedi- ments or other debris in the water, and the water flow velocity. Additionally, several studies have shown that synergy effects between two or more mechanisms may significantly increase the total deleterious effect compared to the sum of their in- dividual effects, see e.g. [80, 147, 163, 203]. This chapter presents a review of commonly observed concrete damage types in hydraulic structures at hydropower facilities, and briefly describes the most common deterioration mechanisms caus- ing the damage and how they may interact to produce damaging synergy effects.
The emphasis is on damage types and processes that are especially common in regions with cold climates and where the structures are in contact with soft water.
It should be noted that other degradation processes, in addition to those presented, may be important for concrete in hydraulic structures, but it is outside the scope of this work to cover all of them. Instead, the interested reader is referred to the more comprehensive summaries reported in e.g. [57, 120, 125–127, 129, 238], where also a number of case histories for hydropower dams around the world can be found. Nevertheless, the chapter begins with a short section covering typical temperature conditions in cold regions since these are of significant importance for some of the described processes.
2.1 Temperature conditions typical for cold climates
In cold regions of the world, significant seasonal temperature variations are com-
mon. As mentioned by Malm et al. [172], temperature differences between sum-
mer and winter can be up to about 70
◦C during a year in countries located in
the northern hemisphere, but the magnitude also varies between different parts
CHAPTER 2. DEGRADATION OF CONCRETE IN HYDRAULIC STRUCTURES
(a) (b)
Figure 2.1: Monthly averaged temperatures for a normal day in Sweden during the
months (a) January and (b) July. The contour plots are from the climate database provided by the Swedish Meteorological and Hydrological Institute SMHI [226].
of the countries, see also e.g. [127, 156, 157, 169, 203]. For example, in Sweden and the other nordic countries in Europe, the seasonal temperature variations are significant, and there are also often substantial differences when comparing the northern and southern parts of the countries. In the northern part of Sweden, the average monthly temperature is approximately -14
◦C during the winter months and roughly 13
◦C during summer. The corresponding temperatures in the south- ern parts are approximately -1
◦C and 16
◦C, respectively [226]. The two contour plots in Fig. 2.1 shows the monthly averaged temperatures for a typical day in Sweden during January and July, respectively. However, comparing the daily max- imum and minimum temperatures, the difference between summer and winter becomes even more noticeable. During summer in the northern parts, it is rather common with temperatures well above 20
◦C, whereas temperatures lower than -30
◦C can prevail for an extended time during winter. The same magnitude of temperature variations is also frequently observed in, e.g. Canada, where Léger and Seydou [156] have reported variations that range up to -35
◦C to 35
◦C for dams in Quebec. Furthermore, during spring and autumn, the temperature com-
10
2.2. DETERIORATION MECHANISMS
C D 1
2
3 Highest water level
Average water level Lowest water level B
A
Figure 2.2: Schematic illustration of different structures (1-3) in a hydropower dam
together with four regions (A-D) subjected to different exposure conditions.
Based on [210].
monly oscillates around 0
◦C in these northern regions, which thus may result in a large number of freeze-thaw cycles [127, 157, 203]. The air temperature, of course, directly affects the temperature in the concrete structures, but another important aspect is the radiation of heat. Depending on the orientation of the concrete sur- faces, a structure can be significantly heated by solar radiation during the days and reach temperatures well above the air temperature. During clear and cold nights, heat can instead radiate to the atmosphere and cool the concrete to temperatures below the air temperature [128, 238]. Studies have shown that these significant temperature variations are important for observed damage in hydraulic structures located in cold regions, see e.g [127, 156, 169–171, 200, 238]. For example, they may result in temperature-induced displacements of the structures because of ther- mal volume changes in the concrete material, or, under certain conditions, due to freezing and thawing of the pore liquid. These two deterioration mechanisms are further addressed below, see Sections 2.2.1 and 2.2.2.
2.2 Deterioration mechanisms
This section presents an overview of some typical concrete damage types com- monly observed in hydraulic structures at hydropower facilities located in cold climates, such as in Sweden. It also briefly describes the deterioration mechanisms causing the concrete degradation. The dominating mechanisms vary depending on the exposure conditions the structure is subjected to and may, consequently, also vary between different parts of the same structure. Figure 2.2 shows a schematic illustration of different structural parts in a hydropower dam, together with four regions, denoted A-D, subjected to different exposure conditions. The illustration and the definition of the different structures and regions were inspired by the work presented by Sandström et al. [210]. The three different structures indicated in the figure, and denoted with the numbers 1 to 3, correspond to:
1. A water-retaining structure where some parts always are in contact with
CHAPTER 2. DEGRADATION OF CONCRETE IN HYDRAULIC STRUCTURES
either air or water, whereas the part in between is intermittently subjected to air or water.
2. A structure always exposed to unfrozen water.
3. A structure that is seldom in contact with air on one side, whereas the oppo- site side is exposed to air and/or soil, e.g., a concrete wall in a channel-type waterway.
The four regions A to D are located on the surfaces of the defined structures and correspond to regions with the following exposure conditions:
A. Regions located above the highest water level and, thus, only exposed to air.
B. Regions located between the highest and lowest water level in a dam, also called the splash zone. These regions are intermittently subjected to air or water.
C. Regions located below the lowest water level.
D. Regions located on or near the bottom and, thus, always in contact with unfrozen water.
These defined structures and regions will be used in the following to indicate where and under which exposure conditions one can expect to find damage caused by the described deterioration mechanisms.
2.2.1 Temperature cracks
It is a well-known phenomenon that most materials, including concrete, expand when heated and contracts during cooling. As long as there are no restraints to these movements, no stresses arise in the material. However, most concrete structures are usually subjected to some degree of external restraint either from the foundation on which it is cast or from other connecting structures. Because of the restrained displacements, stresses arise, and the concrete is damaged and cracks when these reach the tensile strength of the material. The same type of phenomenon may also be induced by wetting and drying of the material, but ac- cording to Hassanzadeh and Westberg Wilde [120], cracking caused only by drying shrinkage is not very commonly observed in dam structures, see also [127].
The heat of hydration from the cement reaction in hardening concrete is often an important factor for observed temperature-induced cracks in concrete structures.
As the cold fresh concrete is cast into the formwork and the cement hydration starts, the material heats up, expands and hardens while still being warm. In massive con- crete structures [240], the subsequent cooling results in a temperature gradient because it takes time for the heat in the interior parts to dissipate, whereas the
12
2.2. DETERIORATION MECHANISMS
Figure 2.3: Observed crack pattern in a concrete buttress dam caused by seasonal tem-
perature variations and a photo of an in-situ temperature crack. Photo provided by Richard Malm.
surface cools rapidly to ambient temperature. The surface contraction is thus in- ternally restrained by the warmer interior concrete, which creates tensile stresses that may fracture the surface. Furthermore, the contraction during the cooling phase can also be restrained by other connecting structures or the foundation on which a structure is cast. Therefore, thin structures may also crack because of the hydration heat. This type of temperature cracks are usually detected a short time after casting. However, various measures can be taken to minimize or avoid these issues, e.g. by using a low heat cement, reducing the amount of cement to lower the heat development, use of adequately designed expansion joints, or embedding a cooling system into the concrete.
Temperature cracks can also appear in an initially crack-free structure due to vary- ing seasonal temperatures. Malm and Ansell [170] showed that the seasonal tem- perature variations in Sweden are sufficiently large to cause substantial cracking in a 40 m high concrete buttress dam. The observed crack pattern in the studied dam is shown in Fig. 2.3 together with a picture of one of the major cracks. The study concluded that the seasonal temperatures cause the dam to move in the upstream- downstream direction, which together with the restraints of the structure, results in the observed crack pattern. Similar conclusions have also been reported as an explanation for observed cracking in a Swedish concrete arch dam [171, 172] and in a multiple arch dam located in Quebec, Canada [127, 169]. Moreover, it should be noted that this type of temperature cracking usually appears at a later age and can thus be seen as a long-term issue.
Temperature cracks due to seasonal temperature variations are most likely to be
observed in structures 1 and 3 of the schematic hydropower dam illustrated in
Fig. 2.2 since these are exposed to the ambient air. However, cracking caused
by the heat of hydration can potentially occur in all three structures. A more
comprehensive discussion of temperature-induced cracking in general and more
examples of observed temperature cracks in Swedish hydraulic structures can be
CHAPTER 2. DEGRADATION OF CONCRETE IN HYDRAULIC STRUCTURES
found in the report by Hassanzadeh and Westberg Wilde [120]. More examples and case histories of temperature-induced cracking in hydraulic structures from other parts of the world can be found in e.g. [127, 238].
2.2.2 Frost damage
As mentioned in the introduction of this thesis, one of the main focuses of this research project has been on modelling frost actions in concrete. Hence, Chapter 4 has been devoted to freezing and thawing of concrete. This brief introduction is meant to provide a conceptual understanding of the governing processes to understand why frost damage occurs, and where one can expect to observe frost damage in a hydraulic structure.
Frost damage is commonly observed in hydraulic structures located in regions with cold climates and is caused by freezing of the pore water inside the con- crete. Liquid water expands by approximately 9 % when it freezes and thereby phase changes into ice. If there is no space for this excess volume inside the con- crete, high pressures arise, and the material is damaged when the pressures reach the tensile strength of the concrete. The magnitude of the pressures can, though, be reduced by using air-entrained concrete, which contains a network of evenly distributed air-filled voids in which the excess water can enter upon freezing. How- ever, the distance from the freezing sites inside the material to these air-filled voids is essential for the magnitude of the pressures, where an increasing distance re- sults in higher pressures. The average distance depends on the quality of the air void system and the degree of water saturation in the material. As will be fur- ther described in Section 3.3, these air voids are slowly filled with water when the concrete is in contact with an external reservoir containing free water. This gradual long-term water filling of the air voids means that the average distance from the freezing sites increases with time and, hence, higher freezing pressures can be expected in the material as the moisture content gradually increases. Fur- thermore, several experimental studies have reported an increased absorption of water from external reservoirs when concrete is subjected to repeated freeze-thaw cycles compared to isothermal conditions [131, 165, 200, 209]. The observation has been explained by a pumping effect caused by the freezing and melting of the pore water [99, 130, 162, 166, 223]. This additional absorption, of course, affects the degree of saturation and, thus, also the average spacing between empty air voids and the freezing-induced pressures. Fagerlund [68] has shown that concrete is internally damaged by frost actions only when the degree of saturation reaches a critical level that is unique for each concrete mixture, commonly denoted the critical degree of saturation. It depends on the microstructure and, to some degree, also on the lowest temperature in the freeze-thaw cycle, since not all water in the pore network phase changes into ice immediately when the material reaches freez- ing temperatures [73]. Instead, the freezing temperature of a certain pore depends on its size, where larger pores freeze at higher temperatures than smaller pores.
Hence, at each freezing temperature, there is a specific content of freezable water in the material that affects the magnitude of the freezing-induced pressures. More-
14
2.2. DETERIORATION MECHANISMS
(a) (b)
Figure 2.4: (a) Surface damage caused by freezing at the level of the water surface in
a hydraulic structure. Photo provided by Manouchehr Hassanzadeh. (b) Spalling of concrete on the upstream side of a dam caused by frost action.
Photo provided by Uniper.
over, it has been shown both experimentally and theoretically that the freezing rate has a relatively limited effect on the critical degree of saturation [68, 69]. Several theories have been proposed in the literature to explain the mechanisms of frost deterioration in concrete, where the described mechanisms above basically corre- spond to Power’s hydraulic pressure theory [192]. This theory and two additional theories are more thoroughly addressed and described in Section 4.2.
Frost damage can occur both on the surface of a concrete structure and in its in- ternal parts. In the case of surface damage, the deteriorated section needs to be in contact with water for an extended time to reach the critical degree of saturation and also be regularly exposed to the ambient climate. In relation to the defined regions in Fig. 2.2, this type of damage is most likely to be observed at the points denoted B and C. However, studies have shown that observed surface damage in hydraulic structures is usually not only caused by frost actions, but rather through a combination of several mechanisms, see e.g. [200, 203]. In Section 2.4, these interactions and synergy effects are further discussed. Figure 2.4a shows a typical example of observed surface damage in a hydraulic structure at the level of the water surface, which at least partly was caused by frost actions. Additionally, Rosen- qvist et al. [201] have experimentally shown that significant spalling of concrete surfaces might occur in regions exposed to the conditions represented by point D.
However, this primarily applies to thin water-retaining structures since the freezing
temperatures on the downstream side must penetrate through the structure and
freeze the pore water located close to the upstream side. An example of this dam-
age type is shown in Fig. 2.4b. Internal frost damage also requires that the critical
degree of saturation is reached in the material, but because of the low permeability
in concrete, the absorption is slow, and it can take substantial time before reaching
the critical state. The damage is typically observed as irregular crack patterns on
the surface of the structure with several cracks oriented in different directions, see
Fig. 2.5. Internal frost damage is most commonly observed in structures of type 1
CHAPTER 2. DEGRADATION OF CONCRETE IN HYDRAULIC STRUCTURES
Figure 2.5: Examples of observed internal frost damage in hydraulic structures. Photos
provided by Manouchehr Hassanzadeh.
and 3 in Fig. 2.2 since these are directly exposed to the ambient climate, where the damage typically appears above the water surface level [120]. Furthermore, inter- nal damage might also occur under the water surface if the thickness and design of the structure allow freezing temperatures to penetrate to critical regions from the downstream side. More examples of observed damage in hydraulic structures caused by frost actions including some case histories from different cold regions in the world are presented in e.g. [57, 127, 128, 238].
2.2.3 Erosion
Erosion can be defined as the progressive loss of material from a solid surface caused by mechanical interactions between the surface and solid objects, particles or fluids [191]. In hydraulic structures, the flowing water normally carries sedi- ments, which at impact with a concrete surface may exert a force that exceeds the tensile strength of the material. It is primarily the cement matrix that is affected by these impacting particles. As the matrix continuously deteriorates, the embedded aggregates eventually also come loose. This type of erosion is commonly denoted hydro-abrasion, and the particles are usually transported through saltation, rolling, or sliding along the surface [7]. However, it has been shown that the impacting effect, or saltation of the particles, is the dominating cause of observed wear of surfaces due to hydro-abrasion, see e.g. [224, 225]. The rate of hydro-abrasion is mainly affected by the particle velocity, the hardness and shape of the particles, and the strength of the surface material. The amount of sediments in the waters do vary significantly between different countries and regions, where, e.g. Swedish rivers typically contain relatively small amounts. However, in the alpine regions of Europe, the waters often contain larger amounts of sediment and substantial dam- age caused by hydro-abrasion is more frequently observed [7]. Figure 2.6 shows two examples of concrete surfaces damaged by hydro-abrasion in a spillway and a stilling basin.
At the level of the water surface in a hydraulic structure, drifting ice in the water
16
2.2. DETERIORATION MECHANISMS
(a) (b)
Figure 2.6: Example of concrete surfaces damaged by hydro-abrasion in a spillway (a)
and a stilling basin (b). Photos provided by Manouchehr Hassanzadeh and Erik Nordström.
can cause abrasive wear of concrete surfaces. In the literature, this type of erosion is usually called ice abrasion, and according to Jacobsen et al. [132], the wear can be attributed to at least three different mechanisms:
1. Asperities of the sliding ice cause tensile stresses that are high enough to fracture the material.
2. Particles released from the damaged surface are dragged with the ice along the surface and increases the wear.
3. Water is forced into defects in the surface, which results in a pressure that can initiate and propagate cracks.
All of these mechanisms act simultaneously during ice abrasion and contributes to the total wear of the surface. Furthermore, in a study by Huovinen [124], it was concluded that the strength of the concrete at the surface is the most important parameter for the resistance to ice abrasion. Additionally, driftwood and other debris that flows with the water can also cause abrasive wear of the concrete at the water surface level.
Cavitation is another phenomenon that can cause severe surface erosion of hy-
draulic structures in sections with high water flow velocities, typically higher than
10 m/s according to Falvey [75]. The destructive mechanism is induced by implod-
ing vapour bubbles in the flowing water. These bubbles are created in sections
where the absolute pressure in the water is below the water vapour pressure. As
these are transported further downstream and enter a section where the pressure is
significantly higher, they can implode and cause a high local pressure that exceeds
the strength of the surface material.
CHAPTER 2. DEGRADATION OF CONCRETE IN HYDRAULIC STRUCTURES
Ice abrasion and wear caused by driftwood and other debris in the water are typi- cally limited to regions close to the water surface level. Thus, this type of abrasive wear can primarily be expected in the regions denoted A-C in Fig. 2.2. Damage caused by cavitation and hydro-abrasion can appear in all regions where there is flowing water, and hence one can expect to find damage caused by these mech- anisms in the regions denoted B-D. However, the damage is usually more severe in regions close to the bottom [210]. Moreover, regardless of which mechanism is causing erosion of a concrete surface, the removed material volume is closely linked to the material’s strength. Hence, other active deterioration mechanisms that weaken the concrete will affect the total amount of removed material. These synergy effects are further discussed in Section 2.4.
2.2.4 Calcium leaching
This section gives a brief introduction to calcium leaching of concrete and some important aspects connected to the leaching of hydraulic structures located in cold regions and in contact with soft water. However, it should be noted that Chapter 5 more thoroughly describes the leaching process and its influence on the microstructure and properties of cement-based materials, together with some experimental techniques used to accelerate the process.
Leaching of concrete generally refers to the process where different solid com- pounds, primarily in the cement matrix, are dissolved into the pore water and then transported to the exterior surface. The dissociated ions are transported ei- ther through diffusion, which is driven by the concentration gradient of ions, or by advection where the pressure-induced flow of the pore water through the porous network transports the dissolved ions [60, 94, 102, 146]. Because of the advective transport, the permeability of the material and cracks may have a significant effect on the rate of leaching in a hydraulic structure. The detrimental effect of leaching in hardened concrete is caused by the decalcification of the hydration products in the cement matrix, which mainly consists of different calcium-based compounds.
In ordinary Portland cement concretes, the major volume fraction of the matrix con- sists of calcium hydroxide (Ca(OH)
2) and calcium silicate hydrates (CSH) [239].
The solubilities of these two solid phases in water are different, where Ca(OH)
2is more soluble than CSH, see e.g. [60, 72, 102, 241, 245]. Calcium leaching is sometimes also called lime leaching, where the word lime refers to the dissolution of calcium-based compounds in the material. Furthermore, according to Halvorsen [114], observed leaching attacks of concrete in hydraulic structures can generally be divided into three types:
1. Leaching of very porous concrete.
2. Leaching through cracks in the concrete.
3. Leaching of concrete surfaces exposed to free water.
18
2.2. DETERIORATION MECHANISMS
Regardless of the type of leaching, the dissolution process weakens the material by increasing its porosity, which leads to lower stiffness and strength as well as higher diffusivity and permeability.
The chemical composition of the water in contact with a concrete surface or pen- etrating a structure has a significant effect on the rate of leaching. Soft water contains low concentrations of calcium and magnesium ions, and hence the expo- sure to such water can lead to substantial leaching of the material. The chemical composition of river waters depend on various factors and local conditions, such as the geology, and results in quite substantial variations in rivers around the world.
For example, Rosenqvist [200] compiled a table of the chemical composition of the water in 15 major Swedish rivers, which shows that a majority of these contain soft water. In the central and northern parts of Sweden, soft water is especially common since the bedrock primarily consists of gneiss and granite in these parts of the country. The mineral concentrations in these rivers are also generally low compared to other rivers in North America, Asia and other countries in Europe [203]. Moreover, in mountainous areas, the rivers usually contain very pure water because of the snowmelt, and can thus also cause significant calcium leaching in concrete structures [72, 125]. In addition, the pH of the water also affects the rate of leaching, where a lower value leads to an increasing rate, see e.g. the study performed by Rosenqvist et al. [203].
A clear visible indication of calcium leaching through cracks in hydraulic structures is the formation of curtains of precipitated calcium carbonates (CaCO
3) on the downstream side in the vicinity of the cracks, see Fig. 2.7. On the contrary, leach- ing of internal parts or at the surface of a structure is more difficult to observe visually since the process primarily weakens the material and increases the poros- ity. However, because of these changed properties, the deleterious effects of other deterioration mechanisms are accelerated and can, e.g., lead to increased damage caused by frost actions or ice abrasion. According to Rosenqvist et al. [203], these synergy effects are a major contributing factor to why surface scaling is frequently detected at the water surface level in hydraulic structures exposed to cold climates and in contact with soft water such as in Sweden, see Fig. 2.4a. Access to water is obviously essential for leaching, and thus only the regions denoted B-D in Fig. 2.2 are susceptible to surface leaching. Furthermore, internal leaching and leaching through cracks can occur in all three types of structures indicated in Fig. 2.2, but the former leaching type usually requires that the concrete has a high porosity and permeability. More examples and case histories of observed damage caused by calcium leaching in hydraulic structures at hydropower facilities can be found in the ICOLD bulletin on concrete dams exposed to aggressive waters [125].
2.2.5 Alkali-aggregate reactions
The AAR were identified already in the early 1940s by Stanton [230]. Cement
contains the two alkali metals sodium (Na) and potassium (K) in the form of
oxides. During the process of hydration, these oxides react with water and form
CHAPTER 2. DEGRADATION OF CONCRETE IN HYDRAULIC STRUCTURES
Figure 2.7: Leaching of concrete through cracks in water-retaining structures with
