DURABILITY OF CONCRETE

LECTURE NOTES

PROF. DR. KAMBZ RAMYAR













Ege University, Civil Engineering Department, IZMIR

1. THE STRUCTURE OF CONCRETE
Structure property relationship is at the heart of modern material science. Concrete
has a highly heterogeneous and complex structure. Therefore, it is very difficult to
constitute exact models of the concrete structure from which the behavior of the material
can be reliably predicted. However, knowledge of the structure and properties of the
individual components of concrete and their relationship to each other is useful for
exercising some control on the properties of the material.
In this chapter, three components of concrete structure - the hydrated cement paste
(hcp), the aggregate, and the transition zone between the cement paste and the aggregate
- will be described. The structure-property relationships are discussed from the standpoint
of durability.
The composition, amount, size and distribution of phases in a solid constitute its
structure. The gross elements of the structure of a material can readily be seen by unaided
human eye. The term macrostructure is generally used for the gross structure, visible to
the human eye (limit of resolution of the unaided human eye: 1/5 mm or 200 m). The finer
elements are usually resolved with the help of microscope. The term microstructure is
used for the microscopically magnified portion of a macrostructure. The magnification
capability of the modern electron optical microscopes is of the order of 10
5
times. Progress
in the field of materials has resulted primarily from recognition of the principle that the
properties of a material originate from its internal structure; in the other words the
properties can be modified by making suitable changes in the structure of material. The
structure of concrete is heterogeneous and highly complex. The structure property
relationships in concrete are not yet well developed. However, an understanding of some
of the elements of the concrete structure is essential for discussing the engineering
properties of concrete such as strength, elasticity, shrinkage, creep and cracking as well
as durability.
Durability is the resistance of concrete to environmental (weathering) conditions,
chemical effects, abrasion and other harmful facts during its service life. Some of the
aspects involved in concrete durability are;
- Freezing thawing resistance
- Wetting drying resistance
- Heating cooling resistance
- Abrasion resistance

1-2


- Fire resistance
- Acid resistance
- Resistance to chemical reactions causing volume expansion such as alkali-
aggregate reaction, sulfate attack, etc
1.1. Complexities of Concrete Structure
At the macroscopic level, concrete may be considered to be a two-phase material,
consisting of aggregate particles of varying sizes and shapes dispersed in a matrix of the
hydrated cement paste.
At the microscopic level, the complexities of the concrete structure becomes visible;
i.e. the two phases of the structure are neither homogeneously distributed with respect to
each other, nor are they themselves homogeneous. For instance, in some areas the hcp
mass appears to be as dense as the aggregate while in others it is highly porous. It is also
known that the volume of capillary voids in the hcp would decrease with decreasing W/C
ratio or with increasing degree of hydration.
For a well-hydrated cement paste, the inhomogeneous distribution of solids and
voids alone can perhaps be ignored when modeling the behavior of the material. However,
micro-structural studies have shown that this cannot be done for the hcp present in
concrete. In the presence of aggregate particles is usually very different from the structure
of the bulk paste or mortar in the system. In fact, many aspects of concrete behavior under
stress can be explained only when the cement paste-aggregate interface is treated as a
third phase of the concrete structure. Thus, the structure of concrete can be summarized
as follows:
1. There is a third phase, the transition zone, which represents the interfacial region
between coarse aggregate particles and the hcp. Existing as a thin shell typically 10 to 50
m thick around large aggregate, the transition zone is generally weaker than the other
two phases of concrete. Therefore, it has a great influence on the mechanical properties of
concrete than is reflected by its size.
2. Each of the three phases is itself multiphase in nature. For instance, each
aggregate particle may contain several minerals, in addition to micro-cracks and voids.
Similarly, both the bulk hcp and the transition zone generally contain a heterogeneous
distribution of various types and amounts of solid phases, pores and micro-cracks.
3. Unlike other engineering materials, the structure of concrete does not remain

1-3


stable; i.e., it is not an intrinsic characteristic of the material. This is due to the fact that hcp
and the transition zone are subject to change with time, humidity and temperature.
The highly heterogeneous and dynamic nature of the structure of concrete are the
primary reasons why the theoretical structure property relationship models, generally so
helpful for predicting the behavior of engineering materials, are of little use in the case of
concrete. A broad knowledge of the important feature of the structure of individual
components of concrete is nevertheless essential for understanding and controlling the
properties of the composite materials.
1.2. Structure of the Aggregate Phase
The aggregate phase is essentially responsible for the unit weight, elastic modulus
and dimensional stability of concrete. These properties of concrete depend to a large
extend on the bulk density and strength of the aggregate, which, in turn, are determined by
the physical rather than chemical characteristics of the aggregate structure. In other
words, the chemical or mineralogical composition of the solid phase in aggregate is usually
less important than the physical characteristics such as the volume, size and distribution of
pores.
Obviously, the chemical and mineralogical character of aggregate in some cases, i.e.
alkali aggregate reaction, become of great importance affecting the durability of
concrete.
In addition to porosity, the shape and surface texture of the coarse aggregate also
affect the concrete properties. Generally, natural gravel has a rounded shape and a
smooth surface texture. Crushed rocks have a rough texture and usually angular shape;
however, depending on the rock type and choice of crushing equipment, the crushed
aggregate may contain a considerable proportion of flat and elongated particles, which
adversely affect many properties of concrete.
Lightweight aggregate particles from pumice, which are highly cellular, are also
angular and have a rough surface texture, but those from expanded clay or shale are
generally rounded and smooth.
Being generally stronger and more durable than the other 2 phases of concrete, the
aggregate phase has no direct influence on the properties of concrete except in the case
of some highly porous, or weak or alkali-reactive aggregates.
The size and shape of coarse aggregate may, however, affect the strength of

1-4


concrete in an indirect way. It is known that increasing the maximum size of the aggregate
increases the proportion of elongated and flat particles. Besides, in such a case (higher
Dmax) the tendency for water films to accumulate next to the aggregate surface (internal
bleeding) increases. Internal bleeding, in turn, weakens the cement paste aggregate
zone.
1.3. Structure of Hydrated Cement Paste
The details of the composition and properties of the hydrated cement paste are out of
the scope of this course. However, a summary of the composition will be beneficial in
understanding some aspects of concrete durability.
Anhydrous Portland cement is a gray or brownish gray powder that consists of
angular particles typically in the size range of 1 to 50 m. It is produced by pulverizing a
clinker with ~2-5% gypsum. The clinker is a heterogeneous mixture of various minerals
(mainly calcium silicate and calcium aluminates) produced by high-temperature reactions
between calcium oxide (CaO), Silica (SiO
2
), alumina (Al
2
O
3
) and iron oxide (Fe
2
O
3
). The
chemical composition of the principal clinker minerals (major compounds) corresponds
approximately to C
3
S, C
2
S, C
3
A and C
4
AF.
In ordinary portland cement the amounts of the major compounds varies in the
following ranges;
C
3
S 45 to 60%
C
2
S 15 to 30%
C
3
A 6 to 12%
C
4
AF 6 to 8%
When portland cement is dispersed in water, the calcium sulfate and the high-
temperature compounds of calcium tend to go into solution, and the liquid phase gets
rapidly saturated with various ionic species. As a result of combinations between calcium,
sulfate, aluminate and hydroxyl ions within a few minutes of cement hydration, first the
needle-shaped crystals of a calcium sulfoaluminate hydrate called ettrengite make their
appearance; a few hours later large prismatic crystals of calcium silicate hydrates begin to
fill the empty space formerly occupied by water and the dissolving cement particles. After
some days, depending on the alumina-to-sulfate ratio of the Portland cement, ettringite
may become unstable and decompose to form the monosulfate hydrate, which has
hexagonal-plate morphology. Hexagonal-plate morphology is also characteristics of

1-5


calcium aluminate hydrates, which are formed in the hydrated pastes of either
undersulfated or high-C
3
A portland cements. The relevant chemical reactions may be
expressed as:

| | | | | | 32 3 6 4 4 H S A C ag. Ca 6 SO 3 AlO + + +
+ 2 2
..Ettringite

| | | | | | 18 6 4 4 H S A C ag. Ca 4 SO AlO + + +
+ 2 2
..Monosulfate

CH H - S - C H S C or S C 3 2 + +

After depletion of sulfate ions in the solution, when the aluminate concentration goes
up again due to renewed hydration of C
3
A (and C
4
AF), ettringite becomes unstable and is
gradually converted into monosulfate, which is the final product of hydration of Portland
cements containing more than 5% C
3
A:

18 4 3 32 3 6 H S A 3C A 2C H S A C +

A model of the essential phases present in the microstructure of a well hydrated
Portland cement paste is shown in Fig 1.1.

Fig 1.1 Model of a well-hydrated Portland cement paste.
A represents aggregation of poorly crystalline C-S-H particles which have at least
one colloidal dimension (1 to 100 nm). Inter particle spacing within an aggregation is 0.5 to
3 nm (av. 1.5 nm).
H presents hexagonal crystalline products such as CH, 18 4 H S A C and C
4
AH
19
.

1-6


They form large crystals, typically 1 m wide.
C represents capillary voids originally occupied with water and is not filled
completely with the hydration products. Size of capillary voids ranges from 10 nm to 1 qm.
In low W/C ratio and well hcp they are < 100 nm.
From the microstructure model of the hcp shown in Fig. 1.1, it is obvious that the
various phases are neither uniformly distributed nor they are uniform in size and
morphology. In solids micro-structural heterogeneities may lead to serious affects on
mechanical properties because these properties are controlled by the micro-structural
extremes, not by the average microstructure. Thus, in addition to the generation of the
microstructure as a result of cement hydration, attention has to be paid to certain
rheological properties of freshly mixed cement paste which are also influential in
determining the microstructure of the hcp. For example, the unhydrated cement particles
have a tendency to attract each other and form flocks, which entrap large quantities of
mixing water. Obviously, local variations in W/C ratio would be the primary source of
evolution of the heterogeneous pore structure. With highly flocculated cement paste
systems not only the size and shape of pores but also the crystalline products of hydration
are known to be different when compared to well dispersed systems. The types, amounts
and characteristics of the four principle solid phases generally present in a hcp, that can
be resolved by an electron microscope, are as follows:
1.3.1. Calcium silicate hydrate:
C-S-H phase makes up 50 to 60% of the volume of solids in a completely hydrated
Portland cement paste and is, therefore, the most important phase in determining the
properties of the paste. C-S-H is not a well defined compound; the C/S ratio varies
between 1.5 and 2.0 and the structural water content varies even more. The morphology
of C-S-H also varies from poorly crystalline fibers to reticular network. Due to their colloidal
dimensions and tendency to cluster, C-S-H crystals could only be resolved by electron
optical microscopy.
The internal crystals structure of C-S-H also remains unresolved. Previously it was
assumed to resemble the natural mineral tobermorite, this is why C-S-H was sometimes
called tobermorite gel.
Although the exact structure of C-S-H is not known, several models have been
proposed to explain its properties. According to Powers Brunauer model, the material has
a layer structure with a very high surface area. Depending on the measurement technique,

1-7


surface areas on the order of 100 to 700 m
2
/g have been proposed for C-S-H. The
strength of the material is attributed mainly to the Van der Waals forces, the size of gel
pores or solid to solid distance being about 18 A. According to Feldman-Sereda model
the C-S-H structure is composed of an irregular or kinked array of layers which are
randomly arranged to creat interlayer spaces of different shapes and sizes (5 to 25 A).
Models of C-S-H structure will be discussed more in detail later. Two forms of C-S-H can
be identified in the microstructure.

Early Product C-S-H: During early hydration, C-S-H grows out from the particle
surface into the surrounding water-filled space in the form of a low-density arrangement of
thin sheets. This form of outer product C-S-H (early product) has a higher microporosity
and on drying rearranges to a variety of morphological forms and coarser porosity. This C-
S-H also contains a high level of impurities (aluminum, sulfate, alkalis) and probably
admixed with monosulfate at the nanometer level. This rather variable component of C-S-
H also has been called both the groundmass and undesignated product (UDP).

Late Product C-S-H: Once hydration has become diffusion controlled, C-S-H forms
primarily as a denser coating around the hydrating cement grains, referred to as either late
or inner product. These coatings form the diffusion barrier during later hydration and
thicken with time, growing inwards as well as outwards. The coatings maintain the shape
of the original grains by surrounding unhydrated residues. These prominent features have
been called phenograins although the term has been used to describe any significant
feature that is distinct from the groundmass regardless of their composition. This late
product C-S-H is denser, has less impurities, and is more resistant to physical change on
drying. The proportion of late product C-S-H increases as hydration increases or the W/C
ratio decreases.
Model of C-S-H Structure
Because of
- its amorphous character,
- compositional variation,
- and poorly resolved morphology,
C-S-H is a difficult material to study. Various conceptual models have been proposed that
emphasize different aspects of the structure to explain observed experimental results. No
one model can be considered to be correct description in any absolute sense, but a good

1-8


model will provide additional insights into the behavior of a material and predict hitherto
unrecognized properties.
Models must be modified as new data are obtained. The following simplified description is
based on several models.
C-S-H can be considered to have a degenerated clay structure by which it means
that it is based on a layer structure. A well x-tallized clay mineral has the structure shown
in Fig.a. it can be thought of as being composed of layers of bread and filling to make a
sandwich. The bread is composed of silico aluminate sheets that are stacked in a
specific orientation the filling is made up of metal ions that hold the sheets together with
comparatively weak electrostatic attractions between positive changes on the metal ions &
residual negative charges on the sheets. Water also present between the layers. In some
clays, the layers can be expanded to accommodate additional water, thereby expanding
the x-tal. Loss of interlayer water on drying allows the layers to collapse again and the x-tal
to contract. Thus, clays show large volume changes on wetting and drying, & C-S-H
behaves similarly.

Fig a) Well x-tallized clay mineral Fig b) poorly x-tallized C-S-H
In C-S-H the bread is calcium silicate sheet and the filling is additional calcium
ions and water molecules. Unlike a well x-tallized clay mineral, however, the sheets are
distorted and randomly arranged. Thus, they do not fit together neatly (Fig.b).
As a result, the spaces between the sheets are irregular and vary considerably in

1-9


their dimensions. (Clay minerals may be visualized as a stack of sheets of copy paper. If
these sheets are crumpled up one by one, smoothed out, and restacked, they will not lie
perfectly flat and will be more randomly arranged with respect to each other. This is the C-
S-H structure).
The space between the calcium silicate sheets is the intrinsic porosity of C-S-H.
Three kinds of pores may be distinguished.
- Interlayer pores (I),
- Micro-pores (M),
- Isolated capillary pores (P).
Capillary pores are spaces in which water can behave as bulk water and menisci are
created as the pores are filled or empitied. In microscopes, the adjoining surfaces are so
close together that water can not form menisci and consequently has a different behavior
from bulk water. Water in micro-pores acts to keep the layers apart by exerting a disjoining
pressure. The disjoining pressure depends on the relative humidity and disappears below
50%RH. When the sheets forming the micro-pores approach closely in a specific
orientation, they may form clay like interlayer spaces (I) that bond the sheets together at
this point. Interlayer bonding can be regarded as a special case of Van Der Waals
bonding. In addition, sheets will from time to time be bonded directly by strong ionic-
covalent bonds, which do not involve the weaker interlayer bonding.
Calcium-hydroxide Portlandite or calcium hydroxide crystals constitute 20 to 25% of
the volume of solids in the hydrated paste. In contrast to C-S-H, calcium hydroxide is a
crystalline material with a fixed composition. It tends to form large crystals with a distinctive
hexagonal prism morphology. The morphology is affected by the available space
temperature of hydration and impurities present in the system. Due to its considerably
lower surface area, the strength contributing potential of CH is very low because of limited
van der waals forces. Besides, the presence of a considerable amount of CH in hcp has
an adverse effect on chemical durability to acid solutions as well as sulfates. The former is
because of higher solubility and the latter is due to higher reactivity of CH compared to that
of C-S-H.
1.3.2. Calcium sulfa-aluminates
These occupy 15 to 20% of the solids volume in the hydrated paste and therefore,
play only a minor role in the structure-property relationships. It was already mentioned that
during the early stages of hydration the sulfate/alumina ionic ratio of the solution phase

1-10


generally favors the formation of ettringite ( 32 3 6 H S A C ), which forms needle-shaped
prismatic crystals. In pastes of ordinary Portland cement, ettringite eventually transforms to
he monosulfate hydrate, 18 4 H S A C , which forms hexagonal-plate crystals. The presence of
monosulfate in Portland cement concrete makes the concrete vulnerable to sulfate attack
upon secondary ettringite formation. It is known that both ettringite and monosulfate
contain small amounts of ironoxide, substituting for some alumina in the crystal structure.
1.3.3. Unhydrated Cement :
Depending on the particle size distribution and degree of hydration of the Portland
cement, some of its particles may remain unhydrated in the microstructure of hcp , even
long after hydration. As stated earlier, the particles of modern cements range in size from
1 to 50 qm. with the progress of hydration reactions, the smaller particles get dissolved
and than the larger particles appear to grow smaller. Since the available space between
particles is limited, the hydration products tend to crystallize in vicinity or even on the
surface of unhydrated clinker particles. This gives the appearance of a coating formation
around them. At later ages, due to lack of available space, the hydration of unhydrated
particles results in the formation of a very dense hydration product, which at times
resembles the original clinker particle in morphology.
1.4. Voids in Hydrated Cement Paste
In addition to solid phase, hcp contains several types of voids which have an
important effect on its properties. The typical types of voids in hcp as well as the size of
solid phases are shown in Fig. 1.2.

1-11



Fig.1.2. Typical size ranges of solids and pores
Interlayer spacing in C-S-H Powers suggested that the width of the interlayer space
within the C-S-H structure to be 18 A and determined that it accounts for 28 percent
porosity in solid C-S-H. However, Feldman and Sereda found that the space may vary
from 5 to 25 A. This pore size is too small to have an adverse effect on strength and
permeability of the hcp. However, water in these small voids can be held by hydrogen
bonding, and its removal under certain conditions may contribute to drying shrinkage and
creep.
Capillary pores:
Capillary pores are originally water filled spaces which are not filled by the solid
hydration products of cement. During the hydration process the total volume of cement-
water mixture remains essentially unchanged. Since the interlayer space within the C-S-H
phase is considered as a part of the solids in hcp, the average bulk density of the
hydration products is considerably lower than the density of unhydrated cement. It is
estimated that 1 cm
3
of cement, upon complete hydration, occupies roughly 2.05 cm
3
.
Thus, upon progress of hydration process the space originally filled by cement and water
is gradually replaced by hydration products. The space not taken up by the unhydrated
cement or hydration products consists of capillary voids, the volume and size of which
depends on W/C ratio (determining the original distance between the anhydrous cement
particles in the freshly mixed cement paste) and the degree of hydration.
In well-hydrated, low W/C ratio pastes, the capillary pores may range from 10 to 50
nm; in high W/C ratio, at early stages of hydration the capillary voids may be as large as 3

1-12


to 5 m. It is shown that the pore size distribution controls the strength, permeability and
volume changes in a hcp rather than the total porosity capillary pores larger than 50 nm,
referred to as macropores, are assumed to be detrimental to strength and impermeability,
while voids smaller than 50 nm, referred to as micropores, are assumed to be more
effective upon drying shrinkage and creep of the hcp.
Air Voids;
In spite of capillary pores which are irregular in shape, the air voids are generally
spherical. In order to improve the freezing-thawing resistance or workability or for
economical considerations, air entraining admixtures may be added to concrete to entrain
very small air voids (in the range of 50 to 200 m) in the cement paste. Besides, during
mixing and placing operations, entrapped air voids as large as 3 mm (3000 m) may
usually be formed in the fresh cement paste. Thus, both the entrapped and entrained air
voids in the hcp are much larger than the capillary voids, and are capable to affect the
strength and impermeability adversely.
1.5. Water in Hydrated Cement Paste
Electron microscopic examination reveals that the voids in the hcp are empty. This is
because the specimen preparation techniques require drying the specimen under high
vacuum. Actually, depending on the relative humidity and porosity of the paste, the
untreated cement paste is capable of holding a large amount of water. Water can exist in
the hcp in many forms. The classification of water into several types is based on the
degree of difficulty or ease with which it can be removed from the hcp. Since there is a
continuous loss of water from a saturated cement paste as the relative humidity is
reduced, the dividing line between various states of water is not rigid. In spite of this,
classification is useful for understanding the properties of hcp. In addition to vapor in
empty or partially water-filled pores, water exists in hcp in four different states.
1.5.1. Capillary Water;
This is the bulk water free from the influence of the attractive forces exerted by the
solid surface and existing in pores greater than 50 A. Actually depending on the behavior
of capillary water in the hcp, it may be divided into two categories. (1 nm = 10 A = 10
-10

m)
a. The water held by capillary tension in small capillaries (5 to 50 nm) which or

1-13


removal may cause shrinkage.
b. The free water held in large voids of the order of > 50 nm, the removal of which
does not cause any volume change.
1.5.2. Absorbed Water;
This is the water close to the solid surface due to the influence of attractive forces
causing water molecules physically absorbed on to the solid surface in the hcp. It has
been suggested that up to six moleculer layers of water (~15A) can be physically held by
hydrogen bonding. Since the bond energies of the individual water molecules decrease
with distance from the solid surface, a major portion of the absorbed water can be lost by
drying the hcp to 30% relative humidity. The loss of absorbed water is the main cause of
the shrinkage of the hcp on drying.
1.5.3. Interlayer Water;
This is the water associated with the C-S-H structure. It has been suggested that a
monomolecular water layer between the layers of C-S-H is strongly held by hydrogen
bonding. This water is lost only on strong drying (i.e. below ~10% relative humidity). The
C-S-H structure shrinks considerably upon interlayer water loss.
1.5.4. Chemically Combined Water;
This water is an integral part of the structure of cement hydration products. It is not
lost by drying, but it is evolved when the hydrates decompose on heating. The various
form of water classified according to the ease with which it may be removed from the hcp
are shown in Fig. 1.3.

Fig. 1.3. Types of water in hcp

1-14


HOMEWORK-1:
Calculate the volume of hydration products and capillary porosity and show them on bar
charts for
a) Hydration of 100 cm
3
of cement with a water/cement ratio of 0.63 for 0, 50, 75 and
100% hydration degrees.
b) 100% hydration of 100 cm
3
of cement for water/cement ratios of 0.3, 0.4, 0.5, 0.6
and 0.7
c) Calculate the compressive strength of the paste for each case

Assumptions:
- Specific gravity of cement : 3.14,
- 1 cm
3
of cement produces ~2cm
3
of the hydration products upon full hydration,
- The original volume of the paste remains unchanged during hydration,
- Volume of entrapped air is negligible,
- The strength solid to space (gel/space) ratio relationship is given as
S=23,5x
3
(MPa) . x = gel/space ratio

HOMEWORK-2:
Using 315 g cement, a paste with W/C ratio of 0.63 is prepared. The specific gravity
of cement is 3.15. Assuming that 1 ml of cement upon hydration produces ~2ml hydration
products,
a) Determine the volume of hydration products and capillary pores for 50, 75 and
100% degrees of hydration.
b) Repeat the same problem for W/C ratios of 0.7, 0.6, 0.5, 0.4 and 0.3 for 100%
degree of hydration,
c) Show the results of the both cases on bar chart diagrams.







1-15


1.6. Structure Property Relationships in hcp
The required engineering properties of hardened concrete, i.e., strength, dimensional
stability and durability, are influenced not only by the proportion but also by the properties
of the hcp, which, in turn, depend on the micro-structural features such as the type,
amount and distribution of solids and voids. The structure property relationships of the hcp
are discussed briefly below.
Strength: The major source of strength in the solid products of hcp is the existence of
Van der Waals forces of attraction. Adhesion between two solid surfaces can be attributed
to these physical forces, the degree of adhesive action being dependent on the extent and
nature of the surfaces involved. The small crystals of C-S-H, H - A - S - C and hexagonal
C-A-H posses large surface areas and adhesive capability. These hydration products tend
to adhere strongly to each other as well as to low- surface- area solids such as CH,
anhydrous clinker grains and fine or coarse aggregates particles.
Since strength resides in the solid part of the material, voids are detrimental to
strength; resultantly there is an inverse relationship between porosity and strength.
In the hcp, the interlayer space within the C-S-H structure and the small voids which
are within the influence of the Van der Waals forces of attraction can not be considered
detrimental to strength, because stress concentration and subsequent rupture on
application of load begin at large capillary pores and microcracks. It was mentioned earlier
that capillary porosity of the paste depends on W/C ratio of the mix and degree of
hydration of cement.
It is shown that there is a process of progressive reduction in the capillary porosity
either with increasing degree of hydration or with decreasing water/cement ratio. The
degree of hydration of the cement depends on the curing conditions (duration of hydration
or age, temperature and humidity).
For normally hydrated Portland cement mortar, Powers proposed the following
exponential equation;

S = k.x
3

Where
S: compressive strength (MPa)
x: Solids to space ratio (gel/space ratio)
k: a constant equal to 235 MPa

1-16


The effect of solid/space ratio (opposite of porosity) on the strength and permeability
and the combined effect of W/C ratio and degree of hydration on the porosity are shown in
Fig. 1.4.



Shaded area shows
Typical capillary
porosity range in
hcps.
Fig 1.4 a,b . Effect of W/C ratio and degree of hydration on porosity, strength and permeability of
hcp.

Dimensional Stability: Saturated hcp is stable at a relative humidity of 100%.
However, when exposed to environmental humidity (< 100%), the material begins to lose
water and shrink. The relationships between loss of water and relative humidity as well as
shrinkage and loss of water are shown in Fig. 1.5.


1-17



Fig.1.5 a) Loss of water as a function of relative humidity
b) Shrinkage of cement mortar as a function of the water loss

As soon as the relative humidity drops below 100%, the free water held in large
cavities (e.g.,> 50 nm) begins to escape to the environment. Since the free water is not
attached to the structure of the hydration products by any physico-chemical bonds, its loss
would not be accompanied by shrinkage (Curve A-B in Fig.1.5). Thus, a saturated hcp
exposed to relative humidities close to 100% may lose a considerable amount of total
evaporable water before undergoing any shrinkage.
When most of the free water has been lost, on continued drying it is found that further
loss of water begins to result in considerable shrinkage. This phenomenon shown by curve
B-C in Fig. 1.5, is attributed mainly to the loss of absorbed water and the water held in
small capillaries. It has been suggested that when confined to narrow spaces between two
solid surfaces, he absorbed water causes a disjoining pressure. The removal of the
absorbed water reduces the disjoining pressure causing shrinkage of the system. The
interlayer water, present as a mono-molecular water film within C-S-H layer structure, can
also be removed by severe drying conditions. This is because the closer contact of the
interlayer water with the solid surface and the tortuosity of the transport path through the
capillary network, call for a stronger driving force. Since the water in small capillaries (5 to

1-18


50nm) exerts hydrostatic tension, its removal tends to induce a compressive stress on the
solid walls of the capillary pore, thus also causing contraction of the system.
It is known that the mechanisms which are responsible for drying shrinkage are also
responsible for creep of the hcp. In the case of creep, a sustained external stress moves
the physically absorbed water and the water held in small capillaries. Thus, creep strain
can occur even at 100% relative humidity.
Durability: The service life of concrete may be markedly reduced by the disintegrating
effects of:
1- Weathering, including the disruptive action of freezing and thawing; the differential
length changes due to temperature variations, and alternative wetting and drying,
2- Reactive aggregates,
3- Aggressive waters in alkali regions,
4- Leaching in hydraulic structures,
5- Chemical corrosion, and
6- Mechanical wear or abrasion.
The term durability of a material relates to its service life under given environmental
conditions. Exposure to acidic solutions is detrimental to hcp due to its alkaline character.
Under these conditions, impermeability or water tightness becomes a primary factor in
determining durability. It is known that the methods of preparing and subsequent treatment
(curing) of concrete are among the major factors influencing the water tightness. The
impermeability of concrete is also of great significance because it is assumed that an
impermeable hcp would result in an impermeable concrete (regarding the aggregate being
impermeable). Permeability is defined as the ease with which a fluid can flow through a
solid. Obviously, the pore structure (size and continuity of the pores) determines the
permeability of the material. Strength and permeability are interrelated in the sense that
both are closely related to the capillary porosity or the solid/space (gel/space) ratio as
shown in Fig. 1.4. (a).
The exponential relationship between permeability and porosity are shown in Fig 1.4.
can be understood from the influence that various pore types exert on permeability. As
hydration proceeds, the void space between the originally discrete cement particles
gradually is filled up with the hydration products. It was mentioned that the total capillary
porosity decreases with the reducing W/C ratio and/or increasing degree of hydration.
Mercury-intrusion porosimetric studies on the cement pastes hydrated with different
W/C ratios (Fig 1.6) and to various ages have shown that the total capillary porosity was

1-19


associated with reduction of large pores in the hcp.

Fig 1.6. Pore size distribution of pastes having different W/C ratios.
When the data of Fig. 1.6 are plotted after omitting the large pores
(i.e.>1320 A), it is found that a single curve could fit the pore size distributions in the 28-
day-old pastes made with four different W/C ratios (Fig.1.7). This indicates that in hcps,
the increase in total porosity resulting from increasing W/C ratio manifests itself in the form
of large pores only. This observation has great significance from the stand point of the
effect of W/C ratio on strength and permeability, which are controlled by large pores.


Fig. 1.7. Distribution plots of small pores in cement pastes of varying W/C ratios


1-20


From the data in Fig.1.4 it is obvious that the coefficient of permeability shows an
exponential drop when the fractional volume of capillary porosity reduces from 0.4 to 0.3.
This range of capillary porosity, therefore, seems to correspond to the point when both the
volume and size of capillary pores in the hcp are so reduced that the interconnections
between them have become difficult. As a result, upon the progress of the hydration of a
young paste, its permeability may show reductions in the order of 10
6
times. It is shown
that a cement paste with even a W/C ratio of 0.6 upon full hydration is as impermeable as
a dense rock such as basalt or marble.
It should be noted that the porosities due to the C-S-H interlayer space and small
capillaries do not contribute to permeability of hcp. On the contrary, with increasing degree
of hydration, there is a considerable increase in the volume of these pores, but the
permeability is greatly reduced. Mehta and Manmohan noted a direct relationship between
the permeability of hcp and the volume of the pores larger than about 100nm. This is
probably because the pore systems, comprised mainly of small pores, tend to become
discontinuous.
1.7. Transition Zone in Concrete
The following features of concrete are interesting,
- It is brittle in tension but relatively tough in compression.
- The paste matrix and aggregate (two main components of concrete) when tested
separately in a uniaxial compression remain elastic until fracture, whereas
concrete itself shows inelastic behavior.
- The compressive strength of concrete is higher than its tensile strength by an
order of magnitude.
- At a given cement content, W/C ratio and age of hydration cement mortar will
always be stronger than the corresponding concrete. Also the strength of
concrete goes down as the coarse aggregate size is increased (at least from a
definite size on).
- The permeability of concrete made from a very dense aggregate will be higher by
an order of magnitude than the permeability of the corresponding cement paste.
- On exposure to fine, the elastic modulus of a concrete drops more rapidly than its
compressive strength
- Regardless of the strength of concrete, at later ages the elastic modulus

1-21


increases at a faster rate than the compressive strength
The reasons for the above behavior of concrete lie in the transition zone existing
between large aggregate particles and the hcp. Although composed of the same elements
as the hcp, the structure and properties of the transition zone are different from the bulk
hcp. Therefore, it is desirable to treat it as a separate phase of the concrete structure.
1.7.1. Structure of the Transition Zone
Due to experimental difficulties, relatively loss information is available about the
transition zone of concrete. However, some understanding of its structural characteristics
may be obtained by following the sequence of its development from the time concrete is
placed.
First, in freshly placed and compacted concrete, water films accumulate around the
large aggregate particles due to the internal bleeding. Thus, there will be a higher W/C
ratio in areas closer to the larger aggregate particles than in the bulk mortar.
Next, similar to the bulk paste, Ca
2+
, SO4
2-
, OH
-
and aluminate ions, produced by the
dissolution of calcium sulfate and calcium aluminate compounds, combine to form
ettrengite and calcium hydroxide. Owing to the high W/C ratio, these crystalline products in
the vicinity of the coarse aggregate consist of relatively larger crystals, and therefore, form
a more porous framework than in the bulk cement paste or mortar matrix. The plate like
CH crystals tends to form in oriented layers perpendicular to the aggregate surface.
Finally, with the progress of hydration, poorly crystalline C-S-H and a second
generation of smaller crystals of ettrengite and CH start filling the empty space existing
between the framework created by the large ettrengite and CH crystals. This improves the
density and hence the strength of the transition zone.
A diagrammatic representation of the transition zone in concrete is shown in Fig 1.8.

1-22




Fig.1.8. Diagrammatic representation of the transition zone and bulk cement paste in
concrete.

The volume and size of voids in transition zone are larger than in the bulk cement
paste or mortar. The size and concentration of 32 3 6 H S A C and CH crystalline compounds
are also larger in the transition zone. The cracks are formed easily in the direction to the C
axis, which account for lower strength of the transition zone.

1.7.2. Strength of the Transition Zone (Bond Strength)
As in the case of the hcp, the cause of bonding hydration products and aggregate
particles is the Van der Waals forces of attraction; therefore, the strength of the transition
zone at any point depends on the volume and size of the voids presents. Even for low W/C
ratio concrete, at early ages the volume and the size of voids in the transition zone will be
larger than in the bulk mortar; consequently, the former is weaker in strength. However,
with increasing age the strength of the transition zone may become equal to or even
greater than the strength of the bulk mortar. This could happen as a result of crystallization
of new products in the voids of the transition zone by slow chemical reactions between the
cement paste constituents and aggregate (probably formation of calcium silicate hydrates
bin the case of siliceous aggregates, or formation of carboaluminate hydrates in the case
of limestone.
Such interactions are strength contributing because they also tend to reduce the
concentration of CH in the transition zone. The large CH crystals posses less adhesion
capacity, not only because of lower surface area and correspondingly weak Van der Waals

1-23


forces of attraction, but also because they surve as preferred cleavage (kayma sayfalar)
sites owing to their oriented structure.
In addition to the large volume of capillary pores and oriented CH crystals, a major
factor responsible for the lower strength of the transition zone in concrete is the presence
of microcracks. The amount of microcracks depends on numerous parameters, including
aggregate size and grading, cement content, W/C ratio, degree of compaction of fresh mix,
curing conditions, environmental humidity and thermal history of concrete. For example, a
concrete mixture containing poorly graded aggregate is more prone to segregation in
compacting; thus thick water films can form beneath the coarse aggregate particles. Under
identical conditions, the larger the aggregate size the thicker would be the water film. The
transition zone formed under these conditions will be susceptible to cracking when
subjected to the tensile stresses induced by differential movements between the
aggregate and the hcp.
Such differential movements commonly arise either on drying or on cooling of
concrete. Thus, concrete has microcracks in the transition zone even before loading.
Obviously, short-term impact loads, drying shrinkage, and sustained loads at high stress
levels will have the effect of increasing the size and number of microcracks.

1.7.3. Influence of Transition Zone on Properties of Concrete
The transition zone being the weakest link of the chain is considered the strength -
controlling phase in concrete. The presence of the transition zone causes the concrete to
fail at a considerably lower stress level than the strength of either of the two main
components of concrete (i.e., aggregate and hcp or mortar). Since it does not take very
high energy levels to extend the cracks already existing in the transition zone, even at 40
to 70% of the ultimate strength, higher incremental strains are obtained per unit of applied
stress. This explains the phenomenon that the concrete components usually remain elastic
until fracture in a uniaxial compression test, whereas concrete itself shows inelastic
behavior.
At stress levels higher than 70% of the u, the stress concentrations at large voids in
the mortar matrix become sufficient to initiate cracking there. With increasing stress, the
matrix cracks gradually spread until they join the transition zone cracks. The crack system
then becomes continuous and the material ruptures. Considerable energy is needed for
the formation and extension of matrix cracks under a compressive load. However, under

1-24


tensile loading cracks propagate rapidly and at a much lower stress level. This is why
concrete fails in a brittle manner in tension but is relatively tough in compression. This is
also the reason why the tensile strength is much lower than the compressive strength of a
concrete.
The structure of the transition zone, especially the volume of voids and microcracks
present, have a great influence on the elastic modulus of concrete. In the composite
material, the transition zone serves as a bridge between the mortar matrix and the coarse
aggregate particles. Even when the individual components are of high stiffness, the
stiffness of the composite may be low due to the broken bridges (i.e., voids and
microcracks in the transition zone ) which prevents stress transfer. Thus, due to the
microcracking on exposure to fire, the elastic modulus of concrete drops faster than the
compressive strength.
The characteristics of the transition zone also influence the durability of concrete.
Prestressed and reinforced concrete elements often fail due to corrosion of the
reinforcement. The rate of corrosion of steel is greatly affected by the permeability of
concrete. The presence of microcracks at the steel coarse aggregate interface is the
primary reason that concrete is more permeable than the corresponding hcp or mortar. It
should be noted that the existence of air and water is a necessary prerequisite to corrosion
of the steel in concrete.
The effect of W/C ratio on permeability and strength of concrete is generally
attributed to its influence on the porosity of the hcp in concrete. However, regarding the
effect of structure and properties of the transition zone on concrete, it is more appropriate
to consider the effect of W/C ratio on the concrete mixture as a whole. This is because
depending on aggregate characteristics, such as Dmax and grading, it is possible to have
large differences in the W/C ratio between the mortar matrix and the transition zone. In
general, everything else remaining the same, the larger the aggregate the higher will be
the local W/C ratio in the transition zone and, consequently, the weaker and more
permeable will be the concrete.







1-25


2. DURABILITY OF CONCRETE
A properly designed, produced and cured concrete is inherently durable to the
environments it will be exposed. Besides, a carefully produced concrete with good quality
control is capable of maintenance-free performance for decades without the need for
protective coatings, except in highly corrosive environments. However, concrete is
potentially susceptible to attack in variety of different exposures unless certain precautions
are taken. Deterioration of concrete can be caused by the adverse performance of any one
of the three major components: aggregate, paste or reinforcement, and can be due to
either chemical or physical causes (Table 2.1). n most of the cases, an individual
environment factor initiates distress, then other factors may contribute and aggravate the
situation.
A major difficulty in studying durability is predicting concrete behavior several
decades in the future on the basis of short-term tests. Most of the knowledge of the
durability has been accumulated through a direct study of actual field problems. The
prediction of concrete durability under a variety of service conditions, is still a major
problem.

Table 2.1. Durability of Concrete
Chemical Attack Physical Attack
Leaching and efflorescence (P) Freezing and Thawing (P, A)
Sulfate Attack (P) Wetting and Drying (P)
Alkali Aggregate Reaction (A) Temperature Changes (P, A)
Acids and Alkalis (P) Wear and Abrasion (P, A)
Re-bar Corrosion (R)

Letters in parenthesis indicates the concrete component most affected, in order of
importance: AAggregate ; P Paste ; R Reinforcement
Water is generally involved in every form of concrete deterioration, and in porous
solids permeability of the material to water usually determines the rate of deterioration.
Therefore, in this chapter the structure and properties of water are described with special
reference to its destructive effect on porous materials; then the permeability of cement
paste, aggregates and concrete as well as the factors controlling their permeability are
discussed.

2-26


Physical effects that adversely influence the durability of concrete include surface
wear, cracking due to crystallization pressure of salts in pores, and exposure to extreme
temperatures. Deleterious chemical effects include leaching of the cement paste by acidic
solutions, and expansive reactions involving sulfate attack, alkali-aggregate reaction and
rebar corrosion in concrete. The significance, physical manifestations, mechanism, and
control of various causes of concrete deterioration are discussed in detail.
Definition: durability is generally considered synonymous with long service life.
Since durability under one set of conditions does not necessarily mean durability under
another, it is customary to include a general reference to the environment when defining
durability. According to ACI Committee 210, durability of Portland cement concrete is
defined as its ability to resist weathering action, chemical attack, abrasion, or any other
process of deterioration that is, durable concrete will retain its original form, quality and
serviceability when exposed to its environment.
Generally, as a result of environmental interactions the microstructure and
consequently, the properties of materials change with time. A material is assumed to reach
the end of service life when its properties under given conditions of use have deteriorated
to an extent that the continuing use of the material is ruled either unsafe or uneconomical.
Significance: it is generally accepted now that in designing structures the durability
characteristics of the materials under consideration should be evaluated as carefully as
other aspects such as mechanical properties and initial cost.
Mostly a substantial portion of the total construction budget is used for the repair and
replacement of existing structures arising from material failures. For example, it is
estimated that in industrially developed countries, over 40% of the total resources of the
building industry are applied to repair and maintenance of existing structures, and less
than 60% to new installations. The escalation in replacement cost of structures and the
growing demand on life-cycle cost rather than first cost are forcing engineers to become
durability conscious. Furthermore, a close relationship exists between durability of
materials and ecology. Conservation of natural resources by making materials having
longer service life is, after all, an ecological step. Besides, the uses of concrete are being
extended to new applications, such as offshore platforms, containers for handling liquefied
gases at cryogenic temperatures and high pressure reaction vessels in the nuclear
industry.
General Observations: before a discussion of important aspects of durability of
concrete, a few general remarks on the subject will be helpful.

2-27


1. Water, the primary cause of both creation and destruction of many natural
materials, is also control to most important durability problems in concrete. In porous
solids, water is the case of many types of physical process of degradation. As a carrier of
aggressive ions, water can also be a source of chemical process of degradation.
2. The physico-chemical phenomena associated with water movement in porous
solids are controlled by the permeability of the solid. For example, the rate of chemical
deterioration would depend on whether the chemical attack is limited to the surface of
concrete or whether it is also work inside the material.
3. The rate of deterioration is also affected by the concentration level of ions in water
and by the composition of solids. Due to the presence of alkali calcium compounds in
hydration products of Portland cement, unlike many natural rocks and minerals, concrete
is a basic material. Therefore, acidic waters are expected to be harmful to it.
Most of our knowledge of physico-chemical processes responsible for concrete
deterioration comes from case histories of structures in the field, because it is difficult in
the laboratory to simulate the combination of long-term conditions normally present in real
life. However, in practice, deterioration of concrete is seldom due to a single cause;
usually, at advanced stages of material degradation more than one deleterious
phenomena are found at work. In general, various causes of deterioration are so closely
intertwined and an interacting so that even separation of the cause from the effect often
becomes impossible. Therefore, a classification of concrete deterioration processes into
neat categories should be treated with some care. Since the purpose of such
classifications is to explain, systematically and individually, the various phenomena
involved, there is a tendency to overlook the interactions when several phenomena are
present simultaneously.








2-28


2.1. Water as an Agent of Deterioration
Water is the most aboundant fluid in nature in the form of seawater, groundwater,
rivers, lakes, rain, snow and vapor. Being small, water molecules are capable of
penetrating extremely fine pores or cavities. As a solvent, water is able to dissolve more
substances than any other liquid. This is due to the presence of many ions and gases in
some waters, which in turn, become instrumental in causing chemical decomposition of
solid materials.
It may also be noted that eater has the highest heat of vaporization among the
common liquids, therefore, at ordinary temperatures it has the tendency to remain in a
material in the liquid state, rather than to vaporize and the material dry.
In porous solids, internal movements and changes of structure of water are known to
cause disruptive volume changes. For example, freezing water into ice, formation of
ordered structure of water inside fine pores, development of osmotic pressure due to
differences in ionic concentration, and hydrostatic pressure build up by differential vapor
pressures can lead to high internal stresses within a moist solid. A brief review of the water
structure will be useful for understanding these phenomena.
2.1.1. Structure of Water
The H-O-H molecule is covalently bonded. Due to asymmetric character of water
molecule, the charge centers of hydrogen and oxygen are different. Thus the porosity
charged proton of the hydrogen ion belonging to a water molecule attracts the negatively
charged electrons of the neighboring water molecules. This relatively weak force of
attraction, called the hydrogen bond is responsible for the ordered structure of water.
The highest manifestation of the long-range order in the structure of water due to
hydrogen bonding is seen in ice (Fig 2.1.a). Each molecule of water in ice is surrounded by
other four molecules, one molecule at the center and four molecules at the corners of
tetrahedron. In all three directions the molecule and groups of molecules are held together
by hydrogen bonds. When ice melts at 0C ~15% of the hydrogen bonds breakdown in
directionality of the tetrahedral bond, thus, each water molecule can acquire more than
four nearest neighbors, which causes the density to rise from 0.917 to 1. Upon
solidification of liquid water, reverse process occurs, thus expansion forms rather than
contraction.


2-29












Fig. 2.1. a) Structure of ice ; b) structure of oriented water molecules in micro pore. (The structure
and properties of water affected by temperature and by the size of pores in a solid).
Compared to the structure of ice, water at room temperature has about 50% of the
hydrogen bonds broken. Materials in the broken-bond state have unsatisfied surface
charges, which give rise to surface energy. The surface energy in liquids causes surface
tension, which accounts for the tendency of a large number of molecules to adhere
together. It is the high surface tension of water (defined as the force required to pull the
water molecules apart) which prevents it from acting as an efficient plasticizing agent in
concrete mixes until suitable admixtures are added.
Formation of oriented structure of water by hydration bonding in micropores causes
expansion. In solids the surface energy is more when numerous fine pores are present. If
water is able to permeat such micropores, and if the forces of attraction at the surface of
pores are strong enough to break down the surface tension of bulk water and orient the
molecules to an ordered structure (analogous to the structure of ice), this oriented or
ordered water, being less dense than the bulk water, will require more space and will
therefore tend to cause expansion (Fig 2.1-b)
2.2. Permeability
Water as a necessary ingredient for the cement hydration and as a plasticizing agent
for concrete components is found in the structure of concrete from beginning. Gradually,
depending on the ambient conditions and the thickness of a concrete element, most of the
evaporable water in concrete, will be lost leaving the pores unsaturated or empty. Since it
is the evaporable water which is freezable and also free for internal movement, the

(a) (b)

2-30


concrete will not be vulnerable to water-related destructive phenomena, provided that
there is a little or no evaporable water left drying, and provided that the subsequent
exposure of the concrete to environment does not lead to resaturation of the pores. The
resaturation of the pores to a large extent depends on the hydraulic conductivity, formed
as (coefficient of) permeability (K).
Many attempts have been made to relate the microstructural parameters of cement
hydration products with either diffusivity (the rate of diffusion of ions through water-filled
pores) or permeability. (The rate of viscous flow of fluids through the pore structure).
According to Garboezi as cited by Mehta, for a variety of reasons the diffusivity
predictions need more development and validation before their practical usefulness can be
proven therefore, in this course only permeability is discussed, implying that this property
covers the overall fluid transport characteristic of the material.
Permeability is defined as the property that governs the rate of flow of a fluid into a
porous solid. For steady-state flow, the coefficient of permeability (K) is determined from
Darcys expression:

L
HA
K
dt
dq
=
Where
dt
dq
: rate of fluid flow
: viscosity of the fluid
H : pressure gradient
A : the surface area
L : thickness of the solid
The coefficient of permeability of a concrete to gas or water vapor is much lower than
the coefficient for liquid water, therefore, tests for measurements of permeability are
generally carried out using water that has no dissolved air. Besides, due to their
interactions with cement paste the permeabilities of solutions containing ions would be
different from the water (pure water) permeability.
2.2.1. Permeability of Cement Paste
In a hcp the size and continuity of the pores at any stage during the hydration
process would control the coefficient of permeability. The mixing water is indirectly
responsible for permeability of the hcp, because its content determines first the total space

2-31


available for hydration products and subsequently the unfilled space after the water is
consumed by either cement hydration reactions or evaporation to the environment.
Generally the permeability of hcp is controlled by W/C ratio and degree of hydration as
shown below:
***Lets consider the hydration process of 1cc cement, accepting that 1 cc cement
upon full hydration produces 2.1 cc hydration products.
For W/C = 0,40 by weight
Weight of cement Wc = 1 x 3,15 = 3,15 g
Weight of water Ww = 0,40 x 3,15 = 1,26 g
Volume of water Vw = 1,26 cc


For 50% hydration
Volume of hydrated cement V
hc
= 0.5 cc
V
hp
= 0.5 x 2.1 = 1.05 ml volume of hydration products
V
cp
= Total volume - Volume of hydration products - Volume of unhydrated cement
V
cp
= 2.26 - 0.50 - 1.05 = 0.71 cc



For 100% hydration
Volume of hydrated cement V
hc
= 1 cc
V
hp
= 1 x 2.1 = 2.1 ml volume of hydration products
V
cp
= Total volume - Volume of hydration products - Volume of unhydrated cement
V
cp
= 2.26 - 2.1 = 0.16 cc









1,26 cc 0.16cc
,Vcp Vcp

1,26 cc

Water

1.26 cc 2.1 cc
Vhp Vhp


1 cc

Cement

0.5 cc
Cement

2-32


0% hydration 50% hydration 100% hydration
*** Lets repeat the same problem for W/C = 0.80
Vc = 1 cc
Wc = 1 x 3,15 = 3,15 g
Ww = 0,80 x 3,15 = 2.52 g
Vw = 2.52 cc

For 50 % hydration V
hp
= 0,5.(2,1)=1,05 ;
V
cp
= 3,52 - 0,5 - 1,05 = 1,97 cc

For 100 % hydration V
hp
= 1,0 . (2,1) = 2,10 ;
V
cp
= 3,52 - 2,10 = 1,42 ml










1,97 cc 1.42cc
Vcp Vcp

2,52 cc

Water

1 cc 2.1 cc
Vhp Vhp

1 cc

Cement

0.5 cc
0% hydration 50% hydration 100% hydration
C
The coefficient of permeability of freshly mixed cement paste is of the order of 10
-4
to
10
-5
cm/s with the progress of hydration both the capillary porosity and coefficient of
permeability decrease. However there is no direct proportionality between the two. This is
because, in the beginning, as the cement hydration process progresses even a small
decrease in the total capillary porosity is associated with considerable segmentation of
large pores, thus greatly reducing the size and number of channels of flow in the cement
paste. Typically, 30% capillary porosity represents a point when the interconnections
between the pores have already become so tortuous that a further decrease in porosity of
the paste is not accompanied by a substantial decrease in the permeability coefficient.
In general, when W/C ratio is high and the degree of hydration is low, the cement
paste will have high capillary porosity; it will contain a relatively large number of big and

2-33


well-connected pores and, therefore, its coefficient of permeability will be high. As
hydration progresses, most of the pores will be reduced to small size (100 nm or less) and
will also lose their interconnections; thus the permeability drops. The coefficient of
permeability of hcp when most of the capillary pores are small and disconnected is of the
order of 10
-12
cm/s. It is observed that in normal cement pastes the discontinuity in the
capillary network is generally reached when the capillary porosity is about 30%. With 0,40 ,
0,50 , 0,60 and 0,70 W/C ratio pastes this generally happens in 3, 14, 180 and 365 days of
moist curing, respectively. Since the W/C ratio in most concrete mixtures seldom exceeds
0,70, it should be obvious that in well-cured concrete the cement paste is not the principal
contributing factor to the coefficient of permeability.
2.2.2. Permeability of Aggregates
Compared to 30 to 40% capillary porosity of typical hcps, the volume of pores in most
natural aggregates is usually under 8% and rarely exceeds 10%. Thus, it is expected that
permeability of aggregate would be much lower than that of typical cement paste. This
may not necessarily be the case. It is shown that the coefficient of permeability of
aggregates are as variable as those of hcp of W/C ratios in the range of ~0,4 to 0,7. The
reason some aggregates, with as low as 10% porosity, may have much higher
permeability than cement paste is because the size of capillary pores in aggregate is much
larger than cement paste. (Most of capillary porosity in nature hcp is 10-100nm; average
pores in aggregate > 10 m ; 100 1000 times greater than pores in hcp).

2.2.3. Permeability of Concrete
Theoretically, the interaction of aggregate particles of low permeability into a cement
paste is expected to reduce the permeability of the system, because the aggregate
particles should intercept the channels of flow within the cement paste matrix. Therefore,
compared to neat cement paste, mortar or concrete with the same W/C ratio and degree of
maturity, should give a lower coefficient of permeability. However, this is not the case, and
the addition of aggregates to cement paste or mortar increases the permeability
considerably. The permeability of concrete depends mainly on the W/C ratio and degree of
hydration (Which controls the size, volume and continuity of capillary pores) as well as
maximum aggregate size (which determines the microcracking in the transition zone
between coarse aggregate and cement paste); in fact, the larger the max aggregate size,

2-34


the greater the coefficient of permeability.
Owing to the significance of the permeability to physical and chemical processes of
deterioration of concrete, a brief review of the factors controlling permeability of concrete
should be useful. Since strength and permeability are related to each other through the
capillary porosity, as a first approximation the factors that influence the strength also affect
the permeability (Fig.2.2).

Fig 2.2 Influence of capillary porosity on comp. strength and coefficient of permeability
A reduction in the volume of a large capillary pores (>100 nm pores) in the hcp
matrix would reduce the permeability. This would be possible by using low W/C ratio,
adequate cement content, proper compaction and curing. Similarly proper attention to
aggregate maximum size & grading, thermal & drying shrinkage strains, and avoiding
premature or excessive loading are necessary steps to limit the transition zone
microcracking which appear to be the major cause of high permeability in concrete. Finally,
the thickness of the concrete element is of importance since it determines the tortuosity of
the path of fluid flow, which in turn, determines permeability.







2-35


3. CLASSIFICATION OF CAUSES OF CONCRETE DETERIORATION
As it was mentioned earlier, the causes of concrete deterioration may be physical or
chemical. The typical causes of concrete deterioration may be grouped into two
categories; surface wear and cracking as follows:

Fig 3.1. Physical causes of concrete deterioration
Similarly, the chemical causes of deterioration may be grouped into three as follows;

Fig 3.2. Chemical causes of concrete deterioration
It should be emphasized that the distinction between the physical and chemical
causes of deterioration is purely arbitrary; in practice the two are frequently superimposed

3-36


on each other. For example, higher permeability of concrete increases the risk of rebar
corrosion and corrosion of reinforcement causes cracking and further increase in
permeability chemical causes of deterioration will be discussed later. Cracking of concrete
due to normal temperature and humidity gradients are out of scope of this course. Here,
the physical causes of concrete deterioration will be explained.
3.1. Deterioration by Surface Wear (Abrasion)
Under many conditions, such as abrasion, erosion and cavitation, progressives loss
of mass from concrete surface may occur.
Abrasion: refers to dry attrition (e.g. wear on pavements and industrial floors by
vehicular traffic)
Erosion: occurs mainly in hydraulic structures, refers to wear by the abrasive action
of fluids containing solid particles in suspension. Erosion takes place in spillways, pipes,
sewage system.
Cavitation : refers to loss of mass by formation of vapor or bubbles and their
subsequent collapse due to sudden charge of direction in rapidly flowing water.
Hardened cement paste has a low resistance to attrition. Service life of concrete can
be greatly reduced under repeated cycles of attrition, especially when paste matrix of
concrete is of high porosity or low strength, and is inadequately protected by an aggregate
which lacks wear resistance. There is a good correlation between the W/C ratio and
abrasion resistance of concrete. Thus, ACI Committee 201 recommends a minimum 28
day compressive strength of ~30 MPa. Suitable strengths may be attained by a low W/C
ratio, proper grading of fine and coarse aggregate (limiting Dmax to ~25mm), lowest
consistency practicable for proper placing and compacting (max. slump 75 mm; for
toppings 25mm), and minimum air content consistent with exposure conditions.
When a fluid containing suspended and rolling particles is in contact with concrete,
the impinging, sliding and rolling action of particles may cause surface wear. The rate of
erosion depends on the quantity, shape, size and hardness of the particles being
transported, on the velocity of the moving particles as well as on the porosity or strength of
concrete.
For silt-size particles (2-150 m), erosion will be negligible at bottom velocities up to
1.8 m/s (min. velocity to transport particles).
When severe erosion or abrasion conditions exist, it is recommended that in addition

3-37


to the use of hard aggregates, the min. 28 day compressive strength of concrete should be
~40 MPa, and before exposure to aggressive conditions concrete should be adequately
cured. European Standard ENV206 (1992) recommends a period of curing twice as long
as normal in order to achieve good resistance to surface wear.
For a proper attrition resistance, at least the surface of concrete should be of high
quality. The properties of concrete surface zone are strongly affected by the finishing
operations; vacuum dewatering is beneficial. The presence of laitance should be avoided
by delaying floating and troweling until the concrete has lost its surface bleed water.
Moreover, the bleeding capacity should be reduced by taking suitable measures such as
using mineral admixtures. Industrial floors or pavements may be designed to have a 25 to
75 mm thick topping, consisting of a low W/C ratio concrete containing hard aggregate of
~12 mm Dmax. Due to very low W/C ratio, concrete toppings containing Latex admixtures
are becoming increasingly popular for abrasion resistance.
Resistance to abrasion can also be achieved by application of surface hardening
solutions to well-cured new floors or abraded old floors. Solutions most commonly used
are magnesium or zinc fluosilicates or sodium silicate which react with CH present in hcp
to form insoluble reaction products, thus, sealing the capillary pores at or near the surface
and hence to improve the resistance to ingress of fluids or dusting due to abrasions.
As far as aggregate is concerned, strong and hard aggregate is preferred with
inclusion of some crushed sand. However, the abrasion resistance of aggregate, as
determined by the Los Angeles test does not seem to be a good indicator of the abrasion
resistance of concrete made with a given aggregate.
From cement content point of view, rich mixes are undesirable, a max. cement
content of 350 kg/m
3
is necessary to allow coarse aggregate particles present just below
the surface of concrete.
Shrinkage compensating concrete has a significantly increased abrasion resistance
probably due to the absence of fine cracks which would encourage the progress of
abrasion.
The non linear flow at velocities exceeding 12 m/s (7 m/s in closed conduits) may
cause severe erosion of concrete through cavitation. In flowing water, vapor bubble form
when the local absolute pressure at a given point in the water is reduced to ambient vapor
pressure of water corresponding to the ambient temperature. As the vapor bubbles flowing
downstream with water enter a region of higher pressure, they implode with great impact
because of the entry of high-velocity water in to the previously vapor occupied space, thus

3-38


causing severe local pitting. Therefore, the concrete surface affected by cavitation is
irregular or pitted, in contrast to the smoothly worn surface by erosion from suspended
solids. The cavitation damage does not progress steadily: usually, after a period of initial
small damage, rapid deterioration occurs, followed by damage at a slower rate.
Best resistance to cavitation damage is obtained by the use of high strength
concrete. The D
max
near the surface should not exceed 20 mm, because cavitation tends
to remove large particles. Hardness of aggregate is not important (unlike the case of
erosion resistance) but a strong bond between aggregate and mortar is vital.
Use of steel fibers may improve the cavitation resistance. However, in contrast to
erosion or abrasion, a suitable and strong concrete may not necessarily be effective in
preventing damage due to cavitation for an indefinite time. The best solution lies in
removal of the causes of cavitation, such as surface misalignments, or abrupt changes of
slope or curvature that tend to pull the flow away from the surface. If possible, local
increase in velocity of water should be avoided as damage is proportional to the sixth or
seventh power of velocity.
Test methods for the evaluation of wear resistance of concrete are not always
satisfactory, because the damaging action varies depending on the exact cause of wear,
and none of the test procedures may satisfactorily simulate the field conditions of wear.
Therefore, laboratory methods are not intended to provide a quantitative measurement of
the length of service that may be expected from a given concrete surface; they can be
used to evaluate the effects of concrete ingredients and curing or finishing procedures on
the abrasion resistance of concrete.
ASTM C 418-90 prescribes the procedure for wear determination by sandblasting;
the loss of volume of concrete serves as a basis for judgment. ASTM C 779-89 describes
three optional methods for testing the relative abrasion resistance of horizontal concrete
surfaces. In the steel-ball abrasion test, load is applied to a rotating head containing steel
balls while the abraded material is removed by water circulation; in the dressing wheel
test, load is applied through rotating dressing wheels of steel; and in the revolving-disk
test, revolving disks of steel are used in conjunction with a silicon carbide abrasive. In
each of the tests, the degree of wear can be measured in terms of weight loss after a
definite time.
Proneness to erosion by solids in flowing water can be measured by means of a
shot-blast test. Here, 2000 pieces of broken steel shot (850 m [#20 ASTM sieve size]) are
ejected under air pressure of 0.62 MPa against a concrete specimen 102 mm away.

3-39


Besides, due to direct relationship between the abrasion and erosion resistance, the
abrasion resistance data can be used as a guide for erosion resistance.
3.2. Cracking by Crystallization of Salts in Pores
A purely physical action of crystallization of the (sulfate) salts in the pores may
account for considerable damage in concrete. For instance, when one side of a slab or
retaining wall of a permeable concrete is in contact with a salt solution and the other sides
are subjected to evaporation, the material can deteriorate by stresses resulting from the
pressure of salt crystallization in pores.
In many porous materials, the crystallization of salts from super saturated salt
solutions can occur only when the concentration of the solute (C) exceeds the saturation
concentration (Cs) at a given temperature. As a rule the higher the (C/Cs) ratio (or degree
of super saturation) the greater the crystallization pressure. The crystallization pressures
for salts that are commonly found in pores of concrete, calculated from the density,
molecular weight, and molecular volume for a C/Cs ratio of 2 are shown in Table 3.1. At a
degree of super position of 10, the calculated crystallization pressure of NaCl is 1835 atm
at 0C and 2190 atm at 50C.


Table 3.1. Crystallization Pressures for Some Salts
Pressure (atm)
C/Cs = 2
Salt
Chemical
Formula
Density
(g/cm
3
)
Molecula
r Weight
(g/mol)
Molar
Volume
(cm
3
/g) 0C 50C
Anhydrite CaSO
4
2,96 136 46 335 398
Epsomite MgSO
4
.7H
2
O 1,68 246 147 105 125
Gypsum CaSO
4
.2H
2
O 2,32 127 55 282 334
Halite NaCl 2,17 59 28 554 654
Hepta hydrite Na
2
CO
3
.7H
2
O 1,51 232 154 100 119
Hexa hydrite MgSO
4
.6H
2
O 1,75 228 130 118 141
Kieserite MgSO
4
.H
2
O 2,45 138 57 272 324
Mirabilite Na
2
SO
4
.10H
2
O 1,46 322 220 72 83
Natron Na
2
CO
3
.10H
2
0 1,44 286 199 78 92
Tach hydrite
2MgCl
2
.CaCl
2
.
12H
2
O
1,66 514 310 50 59
Thenardite Na
2
SO
4
2,68 142 53 292 345
Thermonatrite Na
2
CO
3
.H
2
O 2,25 124 55 280 333

3-40


4. DETERIORATION BY FROST ACTION
The frost action is one of the major problems in concrete structures, such as
pavements, retaining walls, bridge decks and railings, requiring heavy expenditures for
repair and replacement. The deterioration of hardened concrete by frost action is attributed
to the complex microstructure of the material as well as to the specific environmental
conditions. Thus, a concrete that is frost resistance under a given freeze-thaw condition
may be damaged under a different condition. The frost action in concrete can take several
forms. The most common is cracking and spalling of concrete causing by progressive
expansion of cement paste matrix from repeated freeze-thaw cycles. Concrete slabs
exposed to freezing and thawing in the presence of moisture and deicing chemicals are
susceptible to scaling. (i.e. finished surface or skin flake or peels off). The following
theories, each may contribute partially to an explanation of the ultimate failure of concrete
surfaces, are proposed to explain scaling:
- Pressure developed by expulsion of water from saturated aggregate particles.
- Hydraulic pressure developed in capillaries just below the concrete surface
- Acceleration of moisture to ice crystals in capillaries below the surface
- Osmotic pressure caused by concentration of salt in capillaries immediately
beneath the concrete surface
- Finishing operation which may create a surface condition where the finished
surface is dissimilar to the underlying concrete. Improper finishing may also
density the surface and destroy the effectiveness of air entrainment.
- Additional freezing of subsurface ice crystals caused by melting snow and ice with
deicing salts,
- Replenishment of surface moisture during freezing by melting snow and ice with
deicing salts.

Certain coarse aggregates, which contain a relatively high pore volume confined to a
narrow pore size range (0.1 to 1 m), in slabs are known to cause cracking, usually
parallel to joints and edges, which eventually acquires a pattern resembling a large capital
letter D (cracks curving is described as D-line (deterioration line) cracking or simply D-
cracking.

4-41



Fig. 4.1. D-cracks run approximately parallel to joints or edges of concrete surface. As deterioration
progresses these parallel cracks occur further away from the joint. The deterioration is further
advanced to cause disintegration and spalling of concrete near the joint.
4.1. Frost Action on Hardened Cement Paste
According to Powers a saturated cement paste containing no entrained air expands
on freezing due to generation of hydraulic pressure. With increasing air entrainment, the
tendency to expand decreases because the entrained air voids provide escape boundaries
for the hydraulic pressure. A diagrammatic representation of Powers hypothesis is shown
below.

Fig.4.2. Response of saturated cement paste to freezing and thawing both without entrained air.
Powers also proposed that, in addition to hydraulic pressure caused by water

4-42


freezing in large cavities, the osmotic pressure resulting from partial freezing of solutions in
capillaries may be another source of destructive expansion in hcp. Water in capillaries is
not pure and solutions freeze at lower temperatures than pure water. (The higher the
concentration of a salt solution, the lower the freezing point). The existence of local salt
concentration gradients between capillaries is envisaged as the source of osmotic
pressure.
The hydraulic pressure due to an increase in the specific volume of water on freezing
in large cavities and osmotic pressure due to salt concentration differences in the pore
fluid are not the only causes of expansion of hcp exposed to frost action. This became
clear by observing the expansion of cement even when benzene was used as pore fluid
instead of water (Benzene contracts on freezing).
A capillary effect involving large-scale migration of water from small pores to large
cavities is believed to be the main cause of expansion in porous hcp body. Accordingly,
the rigidly held water by C-S-H (both interlayer and absorbed in gel pores) can not
rearrange itself to from ice at normal freezing point of water because the mobility of water
existing in an ordered state is rather limited.
It is estimated that water in gel pores does not freeze above -78 C. Therefore, in a
saturated cement paste subjected to freezing the water in large capillaries turns into ice,
while the gel water continuous to exist as liquid water in a super-cooled state. This creates
a thermodynamic disequilibrium between the frozen water in capillaries, which acquire a
low-energy state, and the super cooled gel water which is in a high-energy state. This
causes the gel water to migrate to lower-sites (large cavities where it can freeze. This
fresh supply of water from the gel pores to the capillary pores increases the volume of ice
in the capillary pores steadily until there is no room to accommodate more ice. Any
subsequent tendency for the super-cooled water to flow toward the ice bearing regions
would obviously cause internal pressure and expansion of the system.
It may be noted that during frost action on cement paste, the tendency for certain
regions to expand is balanced by other regions that undergo contraction (e.g., loss of
absorbed water from C-S-H). The net effect on the specimen is obviously the result of two
opposite tendencies. This satisfactorily explains why cement paste containing air
entrainment shows contraction during freezing.



4-43


4.2. Frost Action on Aggregate
Depending on the response of aggregate on frost action, an air entrained concrete
may still be damaged. The mechanism involved in the development of internal pressure on
freezing a saturated cement paste is also applicable to porous aggregates. However, not
all porous aggregates are susceptible to frost damage; the behavior of an aggregate
particle to freeze-thaw cycles primarily depends on its pore size distribution (size, number
and continuity of pores) and its permeability.
From the stand point of lack of concrete durability to frost action attributed to the
aggregate, three classes of aggregate are proposed.

1- The low permeability and high strength aggregate: so that upon freezing of water
elastic strain is accommodated without causing fracture,

2- The aggregates of intermediate permeability: i.e., those having a significant
proportion of total porosity represented by small pores of the order of 500 nm and smaller.
Capillary forces in those small pores cause the aggregate to get easily saturated and to
hold water. On freezing the magnitude of pressure developed depends on the rate of
temperature drop and the distance that water under pressure must travel to find an escape
boundary to relieve the pressure. Pressure relief may be available either as an empty pore
inside the aggregate or at the aggregate surface. The critical distance for pressure relief in
hcp is of the order of 0.2 mm.
These considerations have given rise to the concept of critical aggregate size with
respect to frost action damage. For a given pore size distribution, permeability, degree of
saturation and freezing rate, the larger the aggregate, the larger is the frost action.
There is no single critical size for an aggregate type because this depends on freezing
rate, degree of saturation and permeability of aggregate. Generally, when aggregates
larger than the critical size are present in concrete, freezing is accompanied by pop-outs,
that is failure of the aggregate in which a part of the aggregate piece remains in the
concrete and other part comes out with mortar flake.

3- Aggregates of high permeability; which generally contains a large number of
relatively big pores. Although they permit easy entry and expelling of water, they may
cause durability problems. This is because the transition zone may be damaged by

4-44


expelled water from an aggregate particle. In such cases, the aggregate particles
themselves are not damaged as a result of frost action. This shows why the results of
freeze-thaw and soundness tests on aggregate alone are not always reliable in predicting
its behavior in concrete.
4.3. Factors controlling Frost Resistance of Concrete
The ability of concrete to resist damage due to frost action depends on the
characteristics of both the cement paste and the aggregate. In each case, however, the
outcome is controlled actually by the interaction of several factors, such as;
- The location of escape boundaries (the distance by which water has to travel for
pressure relief),
- The pore structure of the system (size, number and continuity of pores),
- The degree of saturation (amount of freezable water present),
- The rate of cooling and the tensile strength of the material that must be exceed to
cause rupture.
The provisions of escape boundaries in the cement paste matrix and modification of
its pore structure are the two parameters that are relatively easy to control; the former can
be controlled by means of air entrainment in concrete and the latter by the use of proper
mix proportions and curing.
Air entrainment:
It is not the total air, but void spacing of the order of 0.1 to 0.2 mm within every point
in the hardened cement, which is necessary for protection of concrete against frost
damage. By adding small amounts of certain air-entraining agents to the cement paste
(e.g. 0.05 % by weight of the cement) it is possible to incorporate 0.05 to 0.1 mm bubbles.
Thus, for a given volume of air depending on the size of air bubbles, the number of voids,
void spacing and degree of protection against frost action can vary a great deal.
Although the volume of entrained air is not a sufficient measure of protection of
concrete against frost action, assuming that mostly small air bubbles are present, it is the
easiest criterion for the purpose of quality control of concrete mixtures. Since the cement
paste content is generally related to the maximum aggregate size, lean concretes with
large aggregates have less cement paste than rich concretes with small aggregates,
therefore, rich mixes would need more air entrainment for an equivalent degree of frost
resistance. The recommended air contents for frost resistance, by ACI are given below:

4-45


Air Content (%) Nominal D
max

(mm) Severe Exposure Moderate Exposure
9.00
7.5
6.0
12.5 7.0 5.5
19.0 6.0 5.0
25.0 6.0 4.5
37.5 5.5 4.5
50.0 5.0 4.0
76.0 4.5 3.5

The aggregate grading affects the volume of entrained air, which is decreased by an
excess of very fine sand particles. Addition of mineral admixtures or the use of cements of
higher fineness has a similar effect. Generally, a more cohesive mix is able to hold more
air than either a very wet or very stiff concrete. Besides, insufficient mixing or over mixing,
excessive time of handling or transportation of fresh concrete, and over vibration tend to
reduce the air content. For these reasons, it is recommended that air content should be
determined on concrete as placed.
W/C ratio and Curing:
It was earlier explained that the pore structure of hcp is determined by the W/C ratio
and degree of hydration. In general, the higher the W/C ratio for a given degree of
hydration or the lower the degree of hydration for a given W/C ratio, the higher will be the
volume of large pores in the hcp. Since the readily freezable water resides in large pores,
it can therefore be accepted that at given temperature of freezing the amount of freezable
water will be more with higher W/C ratios and at earlier ages of curing. The relationships
between freezable water, concrete temperature and W/C ratio are shown in Fig. 4.3 (a).
The figure shows that the amount of water that can be frozen in concrete with a given W/C
ratio increases with decreasing temperature. It also shows that the amount of water that
will be freeze at a given temperature increases with W/C ratio (Higher W/C ratio results in
larger and greater number of capillaries in which more freezable water can be present).
Fig. 4.3 (b) shows that a combination of low W/C ratio and entrained air ensures a high
durability factor to frost action.



4-46



Fig 4.3. Effect of W/C ratio and air content on durability of concrete to frost action
The importance of W/C ratio on the frost resistance of concrete is recognized by ACI
318 building code, which requires that normal weight concrete subjected to frost action in a
moist condition should have a maximum W/C ratio of 0.45 for thin sections such as curb,
and 0.5 for other elements. These W/C ratio limits are based on assumption of adequate
cement hydration; therefore, at least 7 days of moist curing at normal temperature is
recommended prior to frost exposure.
Degree of saturation:
Dry or partially dry materials do not suffer frost damage. There is a critical degree of
saturation above which concrete is likely to damage when exposed to very low
temperatures. In fact, the difference between the critical and actual degree of saturation
determines the frost resistance of concrete. A concrete may fall below the critical degree of
saturation after adequate curing, but depending on the permeability it may again reach or
exceed the critical degree of saturation when exposed to a moist environment. Thus, the
permeability is vital in frost resistance of concrete because it controls not only the hydraulic
pressure associated with internal water movement on freezing but also the critical degree
of saturation prior to freezing. From the standpoint of frost damage the effect of increase in
permeability, as a result of cracking due to any physical or chemical causes, should be
apparent.
Strength:
The direct relationship between strength and durability does not hold in the case of
frost damage. Relatively low strength air entrained concrete may show higher frost
resistance than a non-air entrained concrete having higher strength, because of protection
against development of high hydraulic pressures. Generally, in medium and high strength

4-47


concretes, every 1% increase in air content reduces the strength by ~5%. Without any
change in W/C ratio, 5% air entrainment would, therefore, lower the concrete strength by
25%. Due to improved workability as a result of entrained air, it is possible to make up a
part of strength loss by reducing the W/C ratio a little while maintaining the desired level of
workability. Nevertheless, air entrained concrete is generally lower in strength than the
corresponding non air entrained one.




Homework-2 : What are the recommendation of Turkish Building
Codes for concrete subject to freezing & thawing
TS500, TS802
4.4. Concrete Scaling
The resistance of concrete against the combined influence of freezing and deicing
salts (chlorides of ammonia, calcium or sodium) which are commonly used to melt ice and
snow from pavements is generally lower than its resistance to frost alone. The maximum
damage to the concrete surface by scaling occurs at salt concentrations of ~4 to 5%.
It seems that the use of deicing salts has both negative and positive effects on frost
damage, and the most dangerous salt concentration is a consequence of both effects. The
super cooling effect of salt on water (i.e., the lowering the temperature of ice formation)
may be viewed as a positive effect. On the other hand, the negative effects are:
1- An increase in degree of saturation of concrete due to the hygroscopic character
of the salts,
2- An increase in the disruptive effect when the super-cooled water in pores
eventually freezes,
3- The development of differential stresses caused by layer-by-layer freezing of
concrete due to salt concentration gradients,
4- Temperature shock as a result of dry application of deicing salts on concrete
covered with snow and ice,
5- Crystal growth in supersaturated solutions in pores.

Overall, the negative effects of deicing salt application far outweigh the positive
effect; therefore, the frost resistance of concrete under combined influence of freezing and
deicing salts is significantly lowered.

4-48


4.5. Factors Affecting the Amount of Entrained Air
As it was mentioned earlier, the amount and spacing of entrained air bubbles with in
the paste are major factors controlling frost resistance of concrete. Entrained air content
may not be distinguished from entrapped air by routine tests. Thus, it is the total air
content can be measured in the fresh concrete.
In addition to the dosage of the air entraining agent, there are some other factors
influencing the air content and the characteristics of the air voids (even when the dosage
of air entraining agent is kept constant):

Cement Content : Increasing cement content reduces the amount of air entrained in
concrete (at a given dose of air entraining agent, lean mixes entrain more air than the rich
mixes).

Cement Composition: Alkali content of the cement is the major factor affecting
entrained air. For a given dosage of admixture an increase in alkali content, increases air
content.

Cement Fineness: Fine cements entrain less air than coarse ones.

D
max
: An increase in the D
max
reduces the mortar content, thus increasing air content
for a given dosage of the air entraining agent.

Fine Aggregate Quantity & Grading: Fine aggregate contributes to air entrainment
by retention of air bubbles within the interstitial voids between its particles. For a given
dosage of air entraining admixture, air content increases as the amount of fine sand
decreases and as the amount of middle sized sand particles (0.3 to 0.6 mm) increases.

Mineral Admixture: The finely divided mineral admixtures reduce air content. The
effect is more pronounced with fly ash having higher carbon content.

W/C Ratio: Increase in W/C ratio results in an increase in air content.

Initial Consistency of the Concrete: Within practical slump range (2.5 15 cm),

4-49


increase in initial slump is accompanied by an increase in air content. Each 2.5 cm
increase in slump (in the range of 2.5 to 12.5 cm) is accompanied by an increase of about
0.5% point in air content. Further increase in slump cause a rapid decrease in air content.
Above a slump of 15 to 18 cm, many mixtures become too fluid to retain entrained air.

Type & Condition of Mixer: The type of mixer (stationary or transit) and its condition
affect air content considerably. For a given air content decreases by worn blades &
impaired mixing action (due to accumulation of hardened mortar in the drum and on the
blades). Under loading the mixer increases the air content, overloading reduces it.

Duration of Mixing Time: Air content increases with increased time up to ~2
minutes in stationary and up to ~15 minutes in most transit mixers after which air content
may remain approximately constant for a considerable period before it begins to drop off.

Chemical Admixtures: Generally, the presence of chemical admixture increases the
air content to a certain extent, from a slight to a significant amount. In the case of using
chemical admixtures, trial mixes are definitely necessary to determine the required dosage
of the air entraining agent.

Temperature: For a given dosage of air entraining agent, air content varies inversely
with the temperature. (Increasing temperature from 10 to 38 C halves the air content).

Vibration: A proper application of vibration, just enough to compact the mix (~15
seconds) eliminates only the large accidentally formed entrapped air voids. Prolonged
vibration and pumping decrease the air content. Excessive finishing reduces the air
content at the surface zone.
4.6. Recommended Air Contents for Frost Resistance
Concrete may withstand frost action when the cement paste contains proper amount
of entrained air. There is an optimum air bubble size distribution and spacing of bubbles
above which further increase does not increase the durability of concrete.
Actually the entrained air in the concrete is the air content occurring within the paste
fraction. When the entrained air is expressed as a function of content, the optimum air

4-50


content is generally in the range of about 9% of the mortar volume. However, it is not
practical to express air content as a part of the mortar volume. It is easier to express it as a
% of the total volume of concrete.
The air content recommended for a satisfactory frost resistance varies in the range of
3.5 to 8% of the concrete volume. The actual amount of air necessary for frost resistance
depends on the mix proportions, D
max
, and exposure conditions to which the concrete will
be subjected. ACI 201.1-81, recommends the following levels of air content for different
exposure conditions and aggregate sizes. For concretes of compressive strength greater
than 40 MPa the air contents given below may be reduced by ~1%.

ACI Air Content Recommendations for Frost Resistance
Air Content *(%) D
max
(mm) Mild Exposure Moderate Exposure* Severe Exposure **
9.5 4.5 6.0 7.5
12.5 4.0 5.5 7.0
19.0 3.5 5.0 6.0
38.0 2.5 4.5 5.5
50.0 2.0 4.0 5.0
75.0 1.5 3.5 4.5

* Air content in the field may be tolerated by + 1.5%.
* Outdoor exposure in the cold climate, where concrete is occasionally exposed to
moisture prior to freezing, e.g. exterior walls, slabs not in direct contact with soil.
** Outdoor exposures in the cold climate, where concrete is in almost continuous
contact with moisture prior to freezing or where deicing salts are used, e.g.,
pavements, bridge decks, water tanks, side wall..







4-51


5. LEACHING AND EFFLORESCENCE
Efflorescence occurs quite frequently on the surface of concrete when water can flow
through the concrete either continuously or intermittently, or when exposed face is
alternatively wetted and dried. Efflorescence consists of white deposited salts that are
leached out of the concrete and are crystallized on subsequent evaporation of the water or
interaction with CO
2
in the atmosphere. Typical salts are sulfates and carbonates of Na, K,
Ca, the major constituent being CaCO
3
.
Efflorescence seems to be an aesthetic rather than durability problem. However, it
indicates a substantial leaching from concrete which causes an increase in porosity,
thereby lowering the strength and increasing the vulnerability of concrete to aggressive
chemicals. CH is the hydration product most readily leached from concrete. Thus, pastes
having higher content of CH are more prone to leaching and efflorescence.
The rate of leaching is dependent on the permeability of concrete as well as on the
amount of dissolved salts contained in the percolating water. Pure water and soft waters
are the most aggressive, while hard waters containing large amounts of Ca
2+
are less
dangerous. The temperature of water is also consideration, since CH is more soluble in
cold water than it is in warm water. In the case of flowing water or seepage under
pressure, leaching is of course most prevalent, however, concrete is not significantly
leached by water flowing over its surface.















5-52


6. CHEMICAL ASPECTS OF DURABILITY
The resistance of concrete to deleterious processes caused by chemical reactions
involves generally, but not necessarily, chemical interactions between aggressive agents
present in the external environment and the constituents of the cement paste. Among the
exceptions are alkali-aggregate reactions, delayed hydration of crystalline CaO and MgO if
present in excessive amounts in Portland cement, and electrochemical corrosion of
embedded steel in concrete.
In a well-hydrated cement paste, the solid phase, which is composed primarily of
relatively insoluble hydrates of calcium (such as C-S-H, CH, and C-A- S -H) exist in a state
of stable equilibrium with a high pH pore fluid. Large concentrations of Na
+
, K
+
, and OH
-

ions account for a high pH value, 12.5 to 13.5, of the pore fluid in portland cement pastes.
It is obvious that portland cement concrete would be in a state of chemical disequilibrium
when it comes in contact with an acidic environment.
Theoretically, any environment with less than 12.5 pH may be considered aggressive
because a reduction of the alkalinity of the pore fluid would, eventually, lead to
destabilization of the cementitious products of hydration. Thus, most industrial and natural
waters can be categorized as aggressive to concrete. However, the rate of chemical attack
on concrete will be a function of pH of the aggressive fluid and the permeability of
concrete. When the permeability of concrete is low and the pH of the aggressive fluid is
above 6, the rate of chemical attack is considered too slow to be taken seriously. Free CO
2

in soft waters or stagnant waters, acidic ions such as SO
4
2-
and Cl
-
in groundwater and sea
water, and H
+
in some industrial waters are frequently responsible for lowering the pH
below 6, which is considered detrimental to portland cement concrete.
It should be reemphasized that the chemical reactions manifest into detrimental
physical effects, such as increase in porosity and permeability, decrease in strength, and
cracking and spalling. In practice, several chemical and physical processes of deterioration
act at the same time and may reinforce each other. Types of chemical reactions
responsible for concrete deterioration are summarized in Fig 6.1.






6-53



F
i
g
.

6
.
1
.

T
y
p
e
s

o
f

C
h
e
m
i
c
a
l

r
e
a
c
t
i
o
n
s

r
e
s
p
o
n
s
i
b
l
e

f
o
r

c
o
n
c
r
e
t
e

d
e
t
e
r
i
o
r
a
t
i
o
n
.

A
,

s
o
f
t

w
a
t
e
r

a
t
t
a
c
k

o
n

C
H

a
n
d

C
-
S
-
H
;

B
(
I
)
,

a
c
i
d
i
c

s
o
l
u
t
i
o
n
s

f
o
r
m
i
n
g

s
o
l
u
b
l
e

C
a

c
o
m
p
o
u
n
d
s

s
u
c
h

a
s

C
a
C
l
2
,

C
a
S
o
4
,

C
a
(
C
O
3
H
)
2

;

B
(
I
I
)
,

s
o
l
u
t
i
o
n
s

o
f

o
x
a
l
i
c

a
c
i
d

a
n
d

i
t
s

s
a
l
t
s
,

f
o
r
m
i
n
g

C
a

O
x
a
l
a
t
e
;

B
(
I
I
I
)
,

l
o
n
g

t
e
r
m

s
e
a
w
a
t
e
r

a
t
t
a
c
k

w
e
a
k
e
n
i
n
g

C
-
S
-
H

b
y

s
u
b
s
t
i
t
u
t
i
n
g

o
f

M
g
+
2

f
o
r

C
a
+
2
;

C
,

1
)
S
u
l
f
a
t
e

a
t
t
a
c
k

f
o
r
m
i
n
g

e
t
t
r
i
n
g
i
t
e

a
n
d

g
y
p
s
u
m
,

2
)

A
A
R
,

3
)

C
o
r
r
o
s
i
o
n

o
f

s
t
e
e
l



































D
E
T
E
R
I
O
R
A
T
I
O
N

O
F

C
O
N
C
R
E
T
E

B
Y

C
H
E
M
I
C
A
L

R
E
A
C
T
I
O
N
S

R
e
a
c
t
i
o
n
s

i
n
v
o
l
v
i
n
g

f
o
r
m
a
t
i
o
n

o
f

e
x
p
a
n
s
i
v
e

p
r
o
d
u
c
t
s

R
e
a
c
t
i
o
n
s

i
n
v
o
l
v
i
n
g

h
y
d
r
o
l
y
s
i
s

a
n
d

l
e
a
c
h
i
n
g

o
f

t
h
e

c
o
m
p
o
n
e
n
t
s

o
f

t
h
e

h
a
r
d
e
n
e
d

c
e
m
e
n
t

p
a
s
t
e

E
x
c
h
a
n
g
e

r
e
a
c
t
i
o
n
s

b
e
t
w
e
e
n

a
g
g
r
e
s
s
i
v
e

f
l
u
i
d

a
n
d

c
o
m
p
o
n
e
n
t
s

o
f

h
a
r
d
e
n
e
d

c
e
m
e
n
t

p
a
s
t
e
.

I
n
c
r
e
a
s
e

i
n

i
n
t
e
r
n
a
l

s
t
r
e
s
s

I
n
c
r
e
a
s
e

i
n

p
o
r
o
s
i
t
y

a
n
d

p
e
r
m
e
a
b
i
l
i
t
y

S
u
b
s
t
i
t
u
t
i
o
n

r
e
a
c
t
i
o
n
s

r
e
p
l
a
c
i
n
g

C
a
+
+

i
n








C
-
S
-
H

R
e
m
o
v
a
l

o
f

C
a
+
+

i
o
n
s

a
s

n
o
n
-
e
x
p
a
n
s
i
v
e

i
n
s
o
l
u
b
l
e

p
r
o
d
u
c
t
s


R
e
m
o
v
a
l

o
f

C
a
+
+

i
o
n
s

a
s

s
o
l
u
b
l
e

p
r
o
d
u
c
t
s


D
e
f
o
r
m
a
t
i
o
n

C
r
a
c
k
i
n
g
,

S
p
o
l
l
i
g
,


P
o
p
-
o
u
t
s

L
o
s
s

o
f

s
t
r
e
n
g
t
h

a
n
d

r
i
g
i
d
i
t
y

I
n
c
r
e
a
s
e

i
n

d
e
t
e
r
i
o
r
a
t
i
o
n

p
r
o
c
e
s
s

L
o
s
s

o
f

m
a
s
s

L
o
s
s

o
f

a
l
k
a
l
i
n
i
t
y

Detrimental effects of
Chemical reaction
C

A

I
I
I

I
I

B

I


6-54


6.1. Hydrolysis of Cement Paste Components
Water from ground, lakes and rivers contains chlorides, sulfates and bicarbonates of
calcium and magnesium; it is generally hard water and it does not attack the constituents
of the portland cement paste. Pure water from condensation of fog or water vapor, and soft
water from rain or melting of snow and ice contain no or little calcium ions. When these
waters come in contact with portland cement paste, they tend to hydrolyze or dissolve the
calcium-containing products. Once the contact solution attains chemical equilibrium,
further hydrolysis of the cement paste would stop. However, in the case of flowing water or
seepage under pressure dilution of the contact solution will take place, thus providing the
condition for continuous hydrolysis.
In hcp, CH is the most susceptible constituent to hydrolysis. This is due to its
relatively high solubility in pure water (1230 mg/lit). Theoretically, the hydrolysis continues
until most of the CH has been leached away; this exposes the other cementitious
constituents to chemical decomposition. Eventually, the process leaves behind silica and
alumina gels with little or no strength. Results from several investigations show that on the
average, the compressive strength drops by about 2% when 1% CaO is removed from the
portland cement paste.
In addition to loss of strength, leaching of CH from concrete may considered
undesirable for aesthetic reasons. Frequently, the leachate interacts with CO
2
present in
air & results in the precipitation of white crusts of CaCO
3
on the surface (efflorescence).
6.2. Cation Exchange Reactions
Based on cation exchange reaction three types of deleterious reactions may occur:
1- Formation of soluble Ca salts,
2- Formation of insoluble & non-expansive Ca salts,
3- Chemical attack by solutions containing Mg salts.







6-55


6.2.1. Formation of Soluble Ca Salts
The cation exchange reaction between the acidic solutions (such as HCl, H
2
SO
4
or
HNO
3
present in effluents of the chemical industry, acetic, formic or lactic acid found in
many food products, H
2
CO
3
in soft drinks or high CO
2
concentrations found in natural
waters) and the constituents of portland cement paste give rise to soluble salts of calcium
such as calcium chloride, calcium acetate (Ca[CH
2
COO]
2
), and calcium bicarbonate
( Ca[HCO
3
]
2
), which may removed by leaching.
The cation exchange reaction involved in the solution of ammonium chloride and
ammonium sulfate, which are commonly found in the fertilizer and agriculture industry, are
able to transform cement paste components into highly soluble products, for example;

2NH
4
Cl + Ca(OH)
2
CaCl
2
+ 2NH
4
OH

The effect of NH
4
Cl is more severe than MgCl
2
since Mg(OH)
2
is insoluble and its
formation does not increase the porosity and permeability of the system.

MgCl
2
+ Ca(OH)
2
CaCl
2
+ Mg(OH)
2


The typical cation exchange reactions between carbonic acid and CH present in the
hcp can be shown as follows;

Ca(OH)
2
+ H
2
CO
3
CaCO
3
+ 2H
2
O

CaCO
3
+ CO
2
+ H
2
O Ca(HCO
3
)
2

After the precipitation of CaCO
3
which is insoluble, the first reaction would stop
unless some free CO
2
is present in the water. By transforming CaCO
3
into soluble
Ca(HCO
3
)
2
in accordance with the second reaction, the presence of free CO
2
aids the
hydrolysis of Ca(OH)
2
. Since the second reaction is reversible, a certain amount of free
CO
2
, referred to as balancing CO
2
, is needed to maintain the reaction equilibrium. Any free
CO
2
over and above the balancing CO
2
would be aggressive to the cement paste because
by driving the second reaction to the right it would accelerate the process of transformation
of Ca(OH)
2
present in the hydrated paste into the soluble calcium bicarbonate. The

6-56


balancing CO
2
content of the water depends on its hardness (i.e., the amount of Ca and
Mg present in solution).
It should be noted that the acidity of naturally occurring water is generally due to the
dissolved CO
2
which is found in significant concentrations in mineral waters, seawater,
and groundwater when decaying vegetable or animal wastes are in contact with the water.
Normal groundwaters contain 15 to 40 mg/liter CO
2
; however, concentrations of the order
of 150 mg/lit are not uncommon; seawater contains 35 to 60 mg/lit CO
2
. As a rule, when
pH of groundwater or seawater is 8 or above, free CO
2
concentration is generally
negligible; when the pH is below 7, harmful concentrations of free CO
2
may be present.
6.2.2. Formation of Insoluble and Non-expansive Ca Salts
The production of reaction between CH and contain anions (such as oxalic, tartaric,
tannic, humic, hydrofluoric, phosphoric acid and their salts) are non-expansive and
insoluble salts of Ca. the formation of these may not cause damage to concrete unless the
reaction product is removed by erosion, seepage or vehicular traffic.
6.2.3. Chemical Attack by Solutions Containing Mg Salts
The magnesium solutions (MgCl
2
, MgSO
4
, Mg(HCO
3
)
2
) found in frequently in
groundwater, seawater and some industrial effluents readily react with CH present in hcp
to form soluble Ca salts.
MgSO
4
solution is the most aggressive because it decomposes both the alumina-
bearing hydrates and C-S-H as follows;

C
4
AS H
12
+ 3MS (aq) 4CS H
2
+ 3MH + AH
3

3CaO.Al
2
O
3.
CaSO
4
.12H
2
O + 3MgSO
4
S (aq) 4CaSO
4
.2H
2
O + 3Mg(OH)
2
+3Al(OH)
3


C
3
S
2
H
3
+ 3MS (aq) 3CS H
2
+ 3MH + 2SH
x


Silica gel

The silica gel thus formed may react with MH slowly to form a x-talline magnesium
silicate having no cementitious value. The precipitation of MH also increases the rate of

6-57


cation exchange between CH and MS as follows;

CH + MS (aq) CS H
2
+ MH

This reaction is also expansive, but the formation of MH is such that often forms in
the existing pores without causing disruptive expansions. In this way it tends to seal the
concrete and hinder further penetration of sulfates.


























6-58


7. REACTIONS INVOLVING FORMATION OF EXPANSIVE PRODUCTS
Some chemical reactions may result in volume expansion, deformations or
displacements in various parts of the structure, cracking, spalling and pop-outs. The most
frequently occurring five phenomena causing expansion are:
1- Sulfate attack
2- Alkali-Silica reaction (ASR)
3- Alkali-carbonate reaction
4- Delayed hydration of free CaO and MgO
5- Corrosion of reinforcement
7.1. Sulfate Attack
The alkali (Na, K, NH
4
) sulfates found mostly in alkali and agricultural soils and Mg
sulfate present in alkali soils and waters are known to cause deterioration in concrete.
Most soils contain some sulfate (0.01 to 0.05% expressed as SO
4
) in the form of
gypsum. This amount is harmless to concrete. The solubility of gypsum in water is rather
limited (~1400 mg/lit SO
4
). Higher concentrations of sulfates in groundwater are generally
due to the presence of magnesium or alkali sulfates which are of higher solubility. Effluent
from furnaces that use high-sulfur fuels and from chemical industry may contain sulfuric
acid.
Decay of organic matter in marshes, shallow lakes, mining pits and sewer pipes often
leads to the formation of H
2
S, which can be transformed into sulfuric acid by bacterial
action. The water used in concrete cooling towers can also be a source of sulfate attack
because of the gradual build up of sulfates from evaporation of the water. Thus, it is not
uncommon to find deleterious concentrations of sulfates in natural and industrial
environments.
Degradation of concrete due to sulfate attack may occur in two different forms. Which
one of the deterioration processes is predominant in a given case depends on the
concentration and source of the sulfate ions in the contact water and the composition of
cement paste in concrete. Sulfate attack may take place in the form of expansion of
concrete. When concrete cracks, its permeability increases and the aggressive ions
penetrate more easily into the interior, thus accelerating the process of deterioration.
Sulfate attack may also occur in the form of progressive loss of strength and loss of mass
due to deterioration in the cohesiveness of the cement hydration products.

7-59


7.1.1. Chemical Aspects of Sulfate Attack
Calcium hydroxide and alumina-bearing phases of hcp are more susceptible to attack
by sulfate ions. On hydration, portland cements with more than 5% potential C
3
A will
contain most of the alumina in the form of mono-sulfate hydrate, C
3
A.CS H
18
. If the C
3
A
content of the cement is more than 8%, the hydration products will also contain C
3
A.CH.H
18
.
In the presence of CH in portland cement pastes, when the paste comes in contact with
sulfate ions, both the alumina-bearing hydrates are converted to the high-sulfate form (ettringite):

C
3
A.CS H
18
+ 2CH +2 S + 12H C
3
A.3CS .H
32
C
3
A.CH.H
18
+ 2CH +3 S + 11H C
3
A.3CS .H
32

There is a general agreement that the sulfate attack is due to ettringite formation in concrete.
However, the mechanism by which the ettringite formation causes expansion is not clear yet. Two
hypotheses supported by most researchers are;
i) Exertion of pressure by growing ettringite crystals,
ii) Swelling due to the adsorption of water in alkaline environment by poorly crystalline
ettringite.
Gypsum formation as a result of cation-exchange is also capable of causing expansion.
However, the expansion is several times lower than ettringite formation.
CH + S CS + H
Depending on the cation type in sulfate solution (Na
+
or Mg
2+
) both CH and C-S-H
may be converted to gypsum by sulfate attack.

NS + CH + 2H CS .H
2
+ NH
MS + CH + 2H CS .H
2
+ MH
3MS + C
3
S
2
H
3
+ 8H 3 CS .H
2
+ 3MH + S
2
H

In sodium sulfate attack, formation of sodium hydroxide retains the high alkalinity of
the system, which is essential for stability of C-S-H. on the other hand, in magnesium
sulfate attack, conversion of CH to CS is accompanied by formation of the relatively insoluble
and poorly alkaline magnesium hydroxide, thus the stability of the C-S-H is reduced and it is also
attacked by the sulfate solution. Therefore, in high concentrations magnesium sulfate attack is
more severe on concrete.


7-60


7.1.2. Control of Sulfate Attack
The major factors affecting sulfate attack are;

1. the amount & nature of the sulfate present,
2. the level of the water table & its seasonal variation,
3. the flow of groundwater and soil porosity,
4. the type of construction,
5. the quality of concrete

If the sulfate water cannot be prevented from reaching the concrete, the only remedy
is increasing the quality of concrete. It is observed that the rate of sulfate attack on a
concrete structure with all faces exposed to sulfate water (e.g., foundations, piles, etc.) is
less than if moisture can be lost by evaporation from one or more surfaces (e.g., culverts,
retaining walls, slab on ground).
The quality of concrete, particularly a low permeability is the best protection against
sulfate attack. Adequate concrete thickness, high cement content, low W/C ratio, proper
compaction and curing are among the factors contribute to low permeability.
In the event of cracking due to drying shrinkage, frost action, reinforcement
corrosion, etc. additional safety can be provided by the use of sulfate-resisting cements.
Portland cements containing < 5% C
3
A are sufficiently sulfate resisting under
moderate conditions of sulfate attack (i.e. when ettringite formation is the only
consideration). However, at high sulfate concentrations (>1500 mg/lit which are commonly
associated with the presence of Mg& alkali cations) the Type V cement may not be
effective against the cation-exchange reactions involved in gypsum formation, particularly
if C3S content of the cement is high. Under these conditions those cements that upon
hydration contain less CH perform much better, such as high alumina cement, portland
blast furnace slag cement with more than 70% slag, portland pozzolan cement with at
least 25% pozzolan.
In ACI Building Code 318 the sulfate exposure is classified into four degrees of
severity as follows;


- Negligible Attack: When the sulfate content is less than 0.1% in soil or under

7-61


150 ppm (mg/lt) in water, there shall be no restriction on the cement type and W/C ratio.

- Moderate Attack : When the sulfate content is 0.1 to 0.2% in soil, or 150 to 1500
ppm in water, Type II or portland pozzolan or portland slag cement shall be used, with less
than an 0.5 W/C ratio for normal-weight concrete.

- Severe Attack: When the sulfate content is 0.2 to 2% in soil, or 1500 to 10.000
ppm in water, Type V portland cement, with less than an 0.45 W/C ratio, shall be used.

- Very Severe Attack: When the sulfate content is over 2% in soil, or over 10.000
ppm in water, Type V cement plus a pozzolanic admixture shall be used with less than an
0.50 W/C ratio.

For lightweight-aggregate concrete, the ACI Building Code specifies a minimum 28-
day compressive strength of (~30 MPa) 4250 psi for severe or very severe sulfate attack
conditions.
In short, to control sulfate attack, utilization of sulfate resisting cement with a
minimum 370 kg/m
3
cement content, a maximum 0.45 W/C ratio, use of proper mineral
admixtures, a sufficient concrete cover over reinforcement, an adequate curing and finally
a protective coating on concrete surface are recommended.
It should be noted that concrete coatings are not a substitute for high-quality or low-
permeability concrete because it is difficult to ensure that thin coatings will remain
unpunctured and that thick ones will not crack.
The critical sulfate levels as well as required precautions specified by Canadian
Standards Association (CSA Standart CAN3-A23, 1-M77) are given below (Table 7.1).
Under the provision of this standard, seawater would be classified as severe, but when
concrete is totally immersed, the protection given by the formation of MH allows it to be
classed as mild.







7-62


Table 7.1. The critical sulfate levels specified by Canadian Standards Association
Water Soluble
Potential
Degree of
Sulfate Attack
Total Sulfate
(CaSO4) in
Soil Sample
(%)
Sulfate In
Soil Sample
(%)
Sulfate In
Ground Water
(mg/lt)
Type of
Cement to
Be Used
Max. W/C
Ratio
Negligible less than 0.1 --- less than 150 all types ---
Mild 0.1 0.2 --- 150 1000 20 40 50 0.5
Considerable --- 0.2 0.5 1000 2000 50 0.5
Severe --- over 0.5 over 2000 50 0.45

* Canadian cement types 10, 20, 30, 40, 50 correspond to ASTM types I, II, III, IV
and V, respectively.
The use of mineral admixtures as well as Type V cement or use of Type VP or VS
cements give superior protection under very severe sulfate condition where even Type V
may not give satisfactory protection. Under such extreme conditions, super sulfated slag
cements
#
, if available, are alternative. In these cements the higher sulfate resistance is
caused by a low CH content and also by the fact that most of the alumina stays combined
as ettringite, which is stable in this cement and does not transform into monosulfate-
aluminate.
Another way of reducing sulfate attack that can be used for precast concrete
products is high-pressure steam curing. Curing at elevated temperatures below 100 C
may not improve sulfate resistance, and even may decrease it. Sulfate resistance
improves with an increase in curing temperature above 100 C (Table 7.2). Steam curing,
with the use of a silica addition, removes CH from the hydrated pastes, and calcium sulfo-
aluminate hydrates are no longer stable phases. The alumina is incorporated in to C-S-H,
and may form some C
3
AH
6
, which is much more stable to sulfate action. More stable
crystalline calcium silicate hydrates are formed at high temperatures.

Table 7.2. Sulfate Resistance of Autoclaved Mortars (Time to reach given expansions)
Temperature of Curing (24h)
Solution
Linear
Expansion
(%)
21C 100C 125C 150C 175C
2.1% Na
2
SO
4
0.02 3 days 10 days 1 year 1 year > 1 year
Saturated CaSO
4
.2H
2
O 0.02 7 days 18 days 1 year 1 year > 1 year
1.8 % MgSO
4
0.02 6 days 32 days 75 days 130 days > 1 year
#
Super-sulfated cement consists mainly of slag which is activated by anhydrite (CaSO
4
), together with a
small amount of lime or PC.

7-63


Although CH plays an important role in sulfate attack, the susceptibility of a hydrated
cement to sulfate attack depends on its monosulfoaluminate content when exposed to
sulfate attack. Expansive cements are generally susceptible to sulfate attack because the
ettringite that causes the early expansion converts to monosulfoaluminate during later
hydration. However, if ettringite can be maintained as the stable sulfoaluminate hydrate, as
can occur in Type K cement, which is made by blending C
4
AS (anhydrous calcium
sulfoaluminate) with a Type V cement, sulfate resistance will be good.
The form of alumina (called reactive alumina) in some pozzolans (fly ashes & slags)
can also react with sulfates to form ettringite causing undesirable expansion.
7.1.3. Tests for Sulfate Attack
There is no standards ASTM or TS rapid (or even long-term) test to asses the sulfate
resistance of concrete. ASTM C452 which is proposed for research purposes only, uses
an abnormally high gypsum content in the cement to test for sulfate resistance. It is a test
for cement only, not for concrete. This procedure has several drawbacks, the major
objections being that the test is too slow for use as a routine specification test, that the
behavior of concrete does not necessarily parallel the behavior of mortar bars, and that
realistic exposure conditions are not reproduced. Furthermore, tests that are used with
portland cements do not always adequately asses the behavior of cements blended with
pozzolans or slags. More recently ASTM 1012 is proposed for blended cements containing
mineral admixtures. Although this test provides more realistic results, the objections to the
duration of test, the differences between the behavior of concrete and mortar bars (used
also in this test method) and to the concentration of sulfate solution are still remained.












7-64


7.1.4. Selected Case Histories
The U.S. Bureau of Reclamation experienced several cases of sulfate attack on
hydraulic structures located in Wyoming, Montana, South Dakota, Colorado and California.
In some cases, the soluble sulfate content of soil was as high as 4.55%, and sulfate
concentration of water was up to 9900 mg/liter. Many cases of serious deterioration of
concrete structures, 5 to 30 years old, were reported. Research studies showed that
sulfate-resisting cements containing 1 to 3 percent potential C
3
A performed better than 0
percent C
3
A cements, which contained unusually large amounts of C
3
S (58 to 76%).
As a result of sulfate exposure for 20 years, strength loss was reported from concrete
structures of the Fort Peck Dam in Montana. The Dam contain about 1 million m
3
concrete
made with Type I PC (7 to 9%C
3
A), an 0.49 W/C ratio and 335 kg/m
3
cement content.
Compressive strength of cores on 20 year-old specimens showed that the overall condition
of the concrete was very good; however, appreciable sulfate attack was found in two
areas. The deteriorated concrete was mushy and disintegrated easily. The sulfate content
of ground-waters, due entirely to alkali sulfates, was up to 10.000 mg/lt. x-ray diffraction
analysis of deteriorated concrete specimens showed large amounts of gypsum and
ettringite formed at the expense of C-S-H, CH and mono-sulfate hydrate, which are
normally present in mature portland cement pastes. This is an unmistakable evidence of
strong sulfate attack on concrete. Contamination of the cement paste by aggregate is
responsible for the presence of quartz pack in the XRD pattern.

Fig.7.1. X-ray diffraction analysis of cement paste specimens taken from deteriorated concrete of
the Fort Peck Dam.

7-65


Note : The X-Ray Diffraction Technique (XRD) offers a convenient way to determine the
mineralogical analysis of crystalline solids. If a crystalline mineral is exposed to X-rays of a
particular wavelength, the layers of atoms diffract the rays and produce a pattern of peaks which is
characteristic of the mineral. The horizontal scale (diffraction angle) of a typical XRD pattern gives
the crystal lattice spacing, and the vertical scale (peak height) gives the intensity of the diffracted
ray. When the specimen being x-rayed contains more than one mineral, the intensity of
characteristic peaks from the individual minerals are proportional to their amount.

Verbeck reported the results of a long-term investigation on concrete performance in
sulfate soils (containing 10% Na
2
SO
4
) located at Sacramento, California.
Concrete specimens were prepared with five different types of portland cement at
three cement contents. The deterioration of concrete specimens was evaluated by visual
inspection and by measurement of strength and dynamic modulus of elasticity after
various periods of exposure. Verbecks data regarding the effect of C
3
A content of PC and
the cement content of concrete on the overage rate of deterioration are shown in Fig.7.2.
The results clearly show that the cement content (or permeability of concrete) is of
more influence on the sulfate resistance than the composition of cement. For example, the
performance of concrete containing 390 kg/m
3
of the 10% C
3
A cement is two or three
times better than concrete containing 310 kg/m
3
of the 4%C
3
A. With high C
3
A cement
(11%), the effective C
3
A content in the binder mixture can be reduced by the addition of a
mineral admixture such as fly ash, thus causing a beneficial effect on the sulfate
resistance.
Cement Content
225 kg/m
3
310 kg/m
3
390 kg/m
3
A
v
e
.

r
a
t
e

o
f

d
e
t
e
r
i
o
r
a
t
i
o
n
0
40
60
100
120
140
80
20
Ave. C
3
A Content
(
%
/
y
e
a
r
)
(%)
Cement Type
0 2 4 6 8 10 12
V II III I
E
x
p
a
n
s
i
o
n

(
%
/
y
e
a
r
)
1
2
a
c
b
Cement Content ( kg/m )
3
91 136 182 227 272 318
Fly ash C A 3
40 %
20 %
0 %
0 %
0 %
7 %
9 %
11 %
5.1%
2.2 %
1)
2)
a)
b)
c)
91 kg/m
3
= 200 1b/yd
3
Same cement

Fig.7.2. Effect of cement type and content and fly ash addition on sulfate resistance of concrete.

7-66


More recently, Mehta reported an interesting case of sulfate attack showing that soil,
ground-water, seawater and industrial waters are not the only sources of sulfate attack. It
was reported that due to air pollution, the sulfates present in rain water can cause
deterioration of mortar or concrete (located even 18 to 30m) above ground level. This is
likely to happen when the material is permeable, and when during design and construction
adequate precautions are not taken for drainage of rain water.

7.2. Alkali-Aggregate Reaction
Expansion and cracking, leading to loss of strength, elasticity and durability of
concrete can also result from chemical reactions involving alkali ions from portland cement
(or from other sources), hydroxyl ions and certain reactive forms of silica within the
aggregate. In recent literature, the phenomenon is referred to as alkali-silica reaction. Map
cracking or pattern cracking frequently accompanied by exudation of a viscous alkali-
silicate fluid gel, or surface pop-outs and spalling are manifestations of the phenomenon, a
description of which was first published in 1940 by Stanton from investigations of cracked
concrete structures built during the late 1920s ( to the early 1940s) in California. Since
than, numerous examples of concrete deterioration from other parts of the world have
been reported to show that the alkali-silica reaction is at least one of the causes of distress
in structures located in humid environments, such as dams, bridge piers and sea walls.
Most of the structural failures due to alkali-aggregate reaction are reported to be
occurred within 1 to 10 years. However, when less reactive forms of silica are used, the
signs of deterioration may be observed even after 20-25 years.
7.2.1. Cement and Aggregate Types Contributing to the Reaction
Raw materials used in portland cement manufacture account for the presence of
alkalies in cement, typically in the range 0.2 to 1.5 percent equivalent Na
2
O. The use of
(exhaust gases of rotary kiln for preheating of raw mix) sophisticated heat exchangers
(such as suspension preheater) to heat raw mix before it enters the kiln is increasingly
rised. This process also tends to incrase the alkali content of modern cements.
Both laboratory and field data have showed that portland cements containing more
than 0.6 percent equivalent Na
2
O, when used in combination with an alkali-reactive
aggregate, caused large expansions due to the alkali-aggregate reaction. ASTM C 150

7-67


therefore designated the cements with less than 0.6% equivalent Na
2
O as low-alkali, and
with more than 0.6% equivalent Na
2
O as highalkali. In practice, low alkali cements are
usually found sufficient to prevent damage due to the alkali-aggregate reaction (AAR)
irrespective of the type of reactive aggregate. However, in concretes containing very high
cement contents, even less than 0.6% alkali in cement may prove harmful. Europe
experiments shown that if the total alkali content of the concrete from all sources is below
3 kg/m
3
, damage will probably not occur.
The presence of hydroxyl ions and alkali-metal ions are esential for the expansive
phenomenon. Due to the large amount of CH in hcp, the concentration of hydroxylions in
the pore fluid remains high even in low-alkali cements; in this case the expansive will
therefore be limited by the short supply of alkali-metal ions unless these ions are furnished
by any other source, such as alkali-containing admixtures, saltcontaminated aggregates,
and penetration of seawater of deicing solutions containing sodium chloride (NaCl) into
concrete.
In regard to alkali-reactive aggregates, depending on time, temperature, and particle
size, all silicate or silica minerals, as well as silica in hydrous (opal) or amorphous form
(obsidian, silica glass), can react with alkaline solutions, although a large number of
minerals react only to an insignificant degree. Feldspars, pyroxenes, amphiboles, micas,
and quartz, which are the constituent minerals of granites, gneisses, schists, sandstones,
and basalts, are classified as innocuous minerals. Opal, obsidian, cristobalite, tridymite,
chalcedony, cherts, cryptocrystalline volcanic rocks (andesites and rhyolites), and strained
or metamorphic quartz have been found to be alkali reactive, generally in the decreasing
order of reactivity. A comprehensive list of substances responsible for deterioration of
concrete by alkali-aggregate reaction is given in Fig. 5-14b. A few cases of reaction
between alkali and carbonate rocks are also reported in the literature, but they will not be
discussed here.
7.2.2. Mechanisms of Expansion
Depending on the degree of disorder in the aggregate structure and its porosity and
particle size, alkalisilicate gels of variable chemical composition are formed in the
presence of hydroxyl and alkali-metal ions. The mode of attack in concrete involves
depolymerization or breakdown of silica structure35 of the aggregate by hydroxyl ions
followed by adsorption of the alkali-metal ions on newly created surface of the reaction
products. Like marine soils containing surface-adsorbed sodium or potassium, when the

7-68


alkali silicate gels come in contact with water, they sell by imbibing a large amount of water
through osmosis. The hydraulic pressure so developed may lead to expansion and
cracking of the affected aggregate particles, the cement paste matrix surrounding the
aggregates, and the concrete. Solubility of the alkali silicate gels in water accounts for their
mobility from the interior of aggregate to the microcracked regions both within the
aggregate and the concrete. Continued availability of water to be concrete causes
enlargement and extension of the microcracks, which eventually reach the outer surface of
the concrete. The crack pattern is irregular and is therefore referred to as map cracking. It
should be noted that the evidence of alkali-aggregate reaction in a cracked concrete does
not necessarily prove that the reaction is the principal cause of cracking. Among other
factors, development of internal stress depends on the amount, size, and type of the
reactive aggregate present and the chemical composition of the alkali-silicate gel formed.
It is observed that when a large amount of a reactive material is present in a finely divided
form (i.e., under 75 m), there may be considerable evidence of the reaction, but
expansions to any significant extent do not occur. On the other hand, many case histories
of expansion and cracking of concrete attributable to the AAR are associated with the
presence in aggregate of sand-size alkali-reactive particles, especially in the size range 1
to 5mm. Satisfactory explanations for these observations are not available due to the
simultaneous interplay of many complex factors; however, a lower water adsorption
tendency of alkali-silica gels with a higher silica/alkali ratio, and relief of hydraulic pressure
at the surface of the reactive particle when its size is very small, may partially explain
these observations.
7.2.3. Control of Expansion due to ASR
The most important factors affecting the expansion caused by ASR can be listed as
follows:
(1) The alkali content of the cement and the cement content of the concrete;
(2) The alkali-ion contribution from sources other than portland cement, such as
admixtures, salt-contaminated aggregates, and penetration of seawater or deicing salt
solution into concrete;
(3) The amount, size, and, reactivity of the alkali-reactive constituent present in the
aggregate;
(4) The availability of moisture to the concrete structure;
(5) The ambient temperature.

7-69


Thus, the knowledge of he factors affecting the alkali-aggregate reaction can be used
to control its effects in concrete. The following approaches can be used;
- control of pH in the pore solution,
- control of alkali concentrations,
- control of amount of reactive silica,
- control of moisture,
- alteration of alkali-silica gel.

Mineral admixtures are commonly stated to control the expansion associated with
AAR (Fig. 7.3).

0 3 6 9 12
20%
10%
0% flyash
Time (months)
E
x
p
a
n
s
i
o
n

(
%
)
0.02
0.04
0.06
0.08
0.10
0.12
0.14
ASTM C 227
Expansion Limits
Fig.7.3. Effect of flyash addition on the progress of alkali-aggregate reaction in mortar specimens.

It is suggested that pozzolanic reaction in the paste consumes its CH and thereby
lowers the pH of the pore solution. Moreover, the highly reactive silica that is
characteristic of a finely divided pozzolan may consume the alkali in the cement in a rapid
but harmless alkali-pozzolan reaction leading to the formation of Ca bearing low
alkali/silica ratio gel of less expansive character. Besides, it is denoted that alkalies in slag
and pozzolans are acid-insoluble (well bound alkalies) and probably are not available for
reaction with aggregate.
When cement is the only source of alkali ions in concrete and alkali-reactive
constituents are suspected to be present in the aggregate, the use of low-alkali portland

7-70


cement (less than 0.6 % equivalent Na
2
O) may offer the best protection against alkali
attack. If beach sand or sea-dredged sand and gravel are to be used, they should be
washed with sweet water to ensure that the total alkali content from the cement and
aggregates in concrete does not exceed 3 kg/m
3
.
In the future the alkali content of cements can be expected to rise due to changes in
manufacturing technology and environmental regulations. Thus, the use of low-alkali
cements may not remain a feasible means of controlling the alkali-aggregate reaction.
Thus, the total alkali content in concrete can be reduced by replacing a part of high-alkali
cement with cementitious or pozzolanic admixures such as granulated blast furnace slag,
volkanic glass (ground pumice), calcined clay, fly ash, rice husk as or silica fume. It should
be noted that, similar to the well-bound alkalies in most feldspar minerals, the alkalies
present in slags and pozzolans are acid-insoluble and probably are not available for
reaction with aggregate.
In Iceland, only alkali-reactive volcanic rocks are available and the cement raw
materials are such that only high-alkali portland cement is produced, all portland cement is
now blended with condensed silica fume.
Another (even beter) solution is the avoidance of a suspectible aggregate based on
petrographic analyses and service records, although this too is not always practicable or
an economical solution. River gravels are often the most dangerous because they may
contain relatively small amounts of reactive rocks causing high, localized expansions
disrupting the matrix. With mildly reactive aggregates, it is possible to sweeten the
reactive aggregate with 25 to 30% limestone or any other nonreactive aggregate if
economically available. Besides, since moisture availability is essential for expansion to
occur. Control of the access of water to concrete by prompt repair of any leaking joints is
therefore highly desirable to prevent excessive expansion.
A low W/C ratio concrete is less permeable and may help to limit the supply of water
needed to cause the alkali-silica gel to swell, but will only slow down the reaction. Lithium
and barium salts have been used as additives to control alkali-silica expansions, but this is
not an efficient and ecoomical solution. The effect of these additions is not exactly clear; it
is thought that the metal ions may enter the alkali-silica gel and reduce its ability to inbibe
water. Lithium containing glasses are known not to cause expansion in mortars.


7-71


7.2.4. Methods of Identifiying Potential Aggregates
ASTM has available four methods to identifying potentially reactive siliceous
aggregates. They are:
1- ASTM Test Method For Potential Alkali Reactivity of Cement Aggregate
Combinations (Mortar Bar Method)(C 227),
2- ASTM Test Method For Potential Reactivity of Aggregates (Chemical Method)
(C 289),
3- ASTM Guide for Petrographic Examination of Aggregates for Concrete (C 295),
4- More recently Accelerated Mortar Bar Test Method for Alkali-Silica Reactivity
was proposed (C 1260). This test more reliability than either ASTM C 227 and C289.

ASTM C289 is a quick chemical test which measures the solubility of silica when the
powdered aggregate is treated with NaOH. The results are used to assess reactivity.
Certain minerals may interfere with the test and petrographic examination should be done
to help evaluate results. Thus, the test is not a reliable indicator of alkali reactivity. A
potentially deleterious aggregate should be tested for deleterious expansion in the mortar
bar test (C227). In this test, the coarse aggregate is reduced to a fine grading by crushing
(used to cast mortar bars 25x25x285mm) using the cement that will be used in the job.
The bars are stored in moist condition (over water) at 38 C to accelerate the ASR.
Expansion should not exceed 0.05% after 3 months or 0.1% after 6 months. This is not
always a reliable test since deleterious expansions can occur after 6 months (see Fig.9.3.).
Although the test can be continued longer if necessary, the time taken to get the required
information is a serious drawback.
The mortar bar test is also used in a modified form for the evaluation of sand-gravel
reactivity (ASTM C342). Storage is at 55 C and a drying period is also included. A mortar
bar test, using pyrexglass as aggregate, is the basis for (ASTM C441) to evaluate the
effectiveness of a pozzolan in controlling the alkali-aggregate reactions. Expansions
should be reduced to 25% of those measured in mortar bars without the pozzolan.
ASTM C295 (Petrographic examination) can be done rapidly but is somewhat
subjective in nature and depends on the experience and capabilities of the petrographer
examining the aggregate. The results depend on the judgement of the petrographer in
contrast to fixed measuring procedures for classifying the aggregate. ts major use is for
alerting the materials engineer to the presence of potentially reactive rock and mineral

7-72


types and is highly recommended for use in conjuction with other tests and field
performance evaluations.
ASTM C1260 (Accelerated Mortar-Bar Expansion)
Both ASTM C227 and C289 tests fail to identify many slowly reactive aggregates
such as gneiss, schist or quarterite, whose alkali reactivity results from strained
microcrystalline quartz. Thus, recently accelerated mortar bar test was designed to identify
slowly and highly reactive aggregates, based on field performance recordes.
Aggregates for this test are sized and washed to meet the specified grading, than
cast in to mortar bars of 0.5 W/C and stored in a moist cabinet at 23 + 2C. After one day
the specimens are demolded and comparator reading is obtained then stored for 24 h in
water brought to 80 + 2C. An additional comparator reading is obtained then the bars are
transferred to 1 N NaOH solution and again stored at 80 + 2C in sealed containers for 14
days. Periodic comparator readings, say, 1, 3, 7, 10 and 14 days are made while being
careful to prevent cooling and drying during the taking the readings. Test criteria are
suggested as follows:
1. Expansion greater than 0.2% indicates potentially deleteriously reactive aggregate.
2. Expansions between 0.1 and 0.2% are uncertain but are known to include
aggregates that have reacted deleteriosly in field concrete.
3. Expansions less than 0.1% represent innocuous aggregates.

The following points may be emphasized in the procedure:
- The alkali content of cement sed in making the mortar bars is comparatively
unimportant because NaOH in the immersion solution diffuses into the mortar
bars and excerts overwhelming control on reactions.
- The test result indicates only that aggregate is potentially reactive, not that it will
react deleteriously in field concrete. Cement alkali content exerts control on field
performance.
- The test has successfully identified in conformance with field observations, the
deleterious nature of reactivity with the slowly reactive aggregates. Overall the
procedure is rapid.





7-73


7.3. Alkali-Carbonate Reaction (ACR)
The expansive alkali-carbonate reaction occurs between alkali solutions (almost
always alkali (Na&K) hydroxides derived from the cement) in the cement paste matrix, and
limestone aggregate particles of particular but nevertheless somewhat variable
compositions and textures. The rocks in question are dolomitic limestones (MgCO
3
/
CaCO
3
) containing some clay, but not all the reactive rocks in this category cause
deleterious expansions. It is known that expansive rocks have the following features:
1) very fine grained dolomite (small crystals),
2) considerable amounts of fine-grained calcite,
3) abundant interstitial clay, and
4) the dolomite and calcite crystals evenly dispersed in a clay matrix.

The reaction is not properly understood. It has been established that dedolomitization
occurs according to the following equation:
CMC
2
+ NH (or KH) MH + CC + NC
Dolomite (from paste) brucite calcite soluble carbonates

This reaction is expansive and the rate at which it proceeds depends on the crystal
size and concentration of the accompanying calcite. Dedolomitization is fastest when there
is about 50% calcite in the carbonate fraction, with a particle size of less than 40 m. The
exact role of the clay is unknown; it seems that clay causes a structural weakening of the
carbonate skeleton so that expansion caused by dedolomitization can result in internal
cracking. Another suggetion is that dedolomitization exposes clay in a special active
state which can react with alkali metal ions to form a swelling clay. However, it is also
possible that the clay acts as a semi-permeable membrane, thereby setting up an osmotic
cell involving the soluble alkali carbonates produced on dedolomitization. The critical
amount of clay is dependent on the amount of dolomite in the carbonate fraction.
7.3.1. Manifestation of ACR Distress
Concrete that is affected by ACR expands and cracks. Depending on the perfection
of the petrological characteristics of the rock that promotes expansion, and the
convergence of the environmental and materials-related factors that promote expansion,
among which are availability of moisture and alkali content in the concrete, the concrete

7-74


may in a matter of months exhibit severe surface cracking accompanied by closing of
joints. Pattern (map) cracking are reported after 3 years of exposure.
When all of the factors favoring ACR combine to produce near-ideal conditions for
causing the reaction, spectacular expansion and deterioration can occur within 3 years. In
that instance the concrete in sidewalks and curbs expanded about 1.2% with consequent
compressive failures, blow-ups and shoving of adjacent asphalt pavement. The alkali
content of cement and the cement content of the concrete are reported as 1% Na
2
O equ.
and 300 kg/m
3
, respectively.
In ACR affected field concrete, there is no definitive feature of the concrete or of
the geometry of the cracking that precisely identifies the cause of the cracking as being
alkali carbonate-rock reaction. In contrast, gel deposit, if present, may provide a strong
indication that cracking is due to alkali-silica reaction but, not exclude the possibility that
both reactions can be present. The ACR crack geometry depends on the response of
concrete to internal expansive force based on the size and shape of the structural member
and the direction of maximum moisture availability. Generally, expansion of plain, cast-in-
place concrete as in sidewalks, floors, decks and footings where there is a moisture
gradient from top to bottom or from side tos ide results in pattern cracking. Where moisture
conditions are uniform or when concrete element is thick, or of low surface area to volume
ratio, bulk expansion of the concrete is evidenced by closing of joints, extrusion of filler
material, shoving or buckling of adjacent materials or crushing of adjacent weaker
concretes. Cracks may be prominent or faint-visible only following wetting of the surface to
produce some contrast and concrete between the cracks can be hard and intact.
Reinforced and post tensioned concrete, precast and prestressed concrete and other
kinds of prefabricated concrete can develop their own individualized cracking responses to
internal expansion that may not be similar to those produced in slabs on grade. For
example, elangote prestressed or post tensioned members typcally develop long cracks
parallel to the direction of tensioning or prestressing.
In a particular type of exposure condition, the manifestation of distress due to ACR
does not involve gross expansion of the concrete and widespread cracking, but instead,
the damage is confined to the surface region and is the result of application of deicing
salts. A parking deck containing reactive carbonate aggregate throughout was reported to
be deeply scaled exposing numerous cross-fractured coarse aggregate particles in the
scaled surface. Only the coarse aggregate particles in the near surface region had
cracked, extending cracks into the mortar to connect with similar cracks from other

7-75


reactive coarse aggregate particles with the result that the surface concrete was lost.
Although the deck has reactive aggregate throughout, the alkali content of the concrete, in
the absence of deicing salts, is insufficient to sustain reaction and to develop the overall
expansion and tensile cracks producing the map cracking, that is usually observed.
NaCl was reported to have exacerbating effect on ACR. Using concrete prisms,
made with a 1.1% alkali content and a known reactive aggregate, stored in saturated NaCl
solution at 21 C, ~0.19% expansion at 300 days was recorded. The same concrete stored
in pure water at the same temperature showed 0.12% expansion at 300 days.
7.3.2. Comparison of Alkali Carbonate Rock Reaction with Alkali-silica
Reaction as Regards Generation of Crack Damage
In ASR, the first damage to concrete can be produced by both
a) internal cracking of aggregate particles extending cracks into the paste and mortar, and
b) extension of existing cracks and production of new cracks as a consequence of the
migration and subsequent expansion of the ASR gel, initially produced by the reacting
particles, away from the particles and into cracks and voids in the concrete.
In contrast, direct damage due to ACR is produced by the cracking and extension of
cracks into the mortar, generated within the reactive coarse aggregate particle itself.
The initial or direct damage produced in both ASR and ACR may than be
exacerbated by the action of cyclical freezing and thawing. Moreover, it should be
recognized that, although bulk expansions of concrete produced ..by the ACR may be
lower than those produced in severe ASR, the ultimate damage to the concrete can be of
the same order as produced by ASR if the concrete is exposed to cyclical freezing and
thawing action.
7.3.3. Factors Affecting Expansion of Concrete with Reactive Carbonate
Rock and Factors Affecting Expansion of Reactive Carbonate Rock
by Itself
Some carbonate rocks that have reactive texture and that exhibit varying degrees of
unestrained expansion in the rock cylinder test (ASTM C586) may exhibit a vary different
expansion ranking when tested restrained by incorporation in concrete. It is shown that the
degree to which reactive carbonate rocks will cause expansion in concrete is related, on
the one hand, to the restrained imposed by the concrete, and, on the other, to the volume
of dolomite in the rock, upto a point, and the internal textural restrained, of rigidity, of the

7-76


carbonate rock in question.
A reactive carbonate rock with low textural restraint can be highly expansive
unrestrained, but when restrained, either experimentally or within concrete, it is more
compressible and less expansive.
It is found that carbonate rocks with different or equivalent chemical/mineralogical
composition could have vastly different expansion characteristics in concrete if their
structural fabrics differ to such an extent that their elastic properties are quite different.
The petrological factors heading to the perfection of the lithology for producing
expansion in concrete are:
1- Clay content or insoluble residue content in the range of 5 to 25%,
2- Calcite to dolomite ratio of approximately 1:1,
3- Increase in dolomite value up to a point at which interlocking texture becomes a
restraining factor, and
4- Small size of the discrete floating dolomite rhombs.

The expansion of concrete containing high alkali reactive carbonate aggregate is
promoted by:
1- Increasing coarse aggregate size,
2- Moisture availability,
3- Higher temperature,
4- High alkali content of the concrete and high pH of the liquid phase in cement
pores,
5- High proportion of reactive rock in the coarse aggregate,
6- Lower concrete strength.
7.3.4. Mechanism of Reaction and Expansion
The main chemical reaction is decomposition or dedolomitization of dolomite
[CaMg(CO
3
)] to calcite (CO
3
) and brucite [Mg(OH
2
)], as represented by the following
reaction in which M represents an alkali element, such as K, Na or Li.
CaMg(CO
3
) + 2MOH Mg(OH)
2
+ CaCO
3
+ M
2
CO
3

A further reaction that occurs in concrete is that the alkali carbonate produced is the
initial reaction may them react with the Ca(OH)
2
, produced as a normal product of cement
hydration, to regenerate the alkali hydroxide, e.g.,
Na
2
CO
3
+ Ca(OH)
2
2NaOH + CaCO
3


7-77


Although the starting and end products of the reaction have been well ocumented
and established the overall bulk chemical and mineralogical changes taking place in
dedolomitization reaction are not well understood. However, it seems that the primary
dedolomitization reaction is the prerequisite that enables other processes to work to
produce the expansion.
An idea common to some proposed mechanism is that, in one way or another, water
uptake, absorbtion and incorporation is important in producing expansion.
It is proposed that dedolomitization exposes active clay minerals that are present as
inclusions within the dolomite euhedra in some rocks.
Exchange sites on the clay surface absorb sodium ions, and water uptake by the
new clay results in swelling. Thus, it is the active clay released during dedolomitization
that contributes to the expansion.
The role of clay minerals in the fabric of the rock, which may contribute to the
expansion mechanism, is to provide mechanical pathways to the dolomite rhombs by
disrupting the structural framework of the rock, thus weakening the carbonate matrix and,
by that means, helping to promote expansion. Thus, the presence of clays is not essential
to the chemistry of the dedolomitization reaction but is an enabling factor, mechanically
promoting expansion when dedelomitization occurs.
It is also concluded that both the limit and the rate of the dedolomitization reaction
and expansion depend directly on the pH value of the pore water in the concrete. At pH
lower than 12, the reaction and subsequent expansion, proceeds very slowly or not all.
7.3.5. Evaluating the Potential for AC Reactivity
a. Field Servis Record:
Field performance information on concrete incorporating the aggregate under
consideration can furnish direct answers to the questions that the laboratory evaluation
tests attempt to answer indirectly.
Provided that certain essential information is available and certain exposure
conditions are fulfilled, an evaluation of the potential for aggregate reactivity based on the
investigations of concrete structures in the field can be most unequivocal.
When investigating and samplimg a concrete structure (ASTM C823) known to
contain aggregate from the source in question, for example, by prior petrographic
examination, the alkali content of the cement and the cement content of concrete must be
known. In general, if the cement used in concrete was low alkali cement (<0.6%Na
2
O

7-78


equilibrium) then the concrete cannot provide information on the propensity for reaction of
the aggregate with high alkali cement. However, in the view of the facts,
(1) the most reactive carbonate rocks can be produce deleterious expansion with
cements of 0.4% total alkalies, and
(2) The more important measure is the alkali content per unit volume of concrete and
not the alkali content of the cement.
Then if the cement used is a low alkaline one, if the cement content of concrete is
high, the alkali content of concrete may equal or exceed that of concrete in which cement
content is low but the alkali content of cement is high. A threshold guidline (based on
equivalent Na
2
O in kg/m
3
) must be established for judging whether the aggregate should
be considered to have been used in a high or low alkali concrete.
If the structure shows no distress and it has been established that the alkali content
of the concrete would be sufficient to cause reaction if the aggregate is potentially reactive,
then the effects of the aggregate size, proportion of reactive aggregate, and moisture
availability to the concrete must also be considered singly and in combination before
concluding that the aggregate is non-deleteriously expansive.
Larger-sized reactive aggregate causes greater expansion than smaller-sized
reactive aggregate. For a given high alkali cement content a reactive aggregate that has
not caused distress, may cause deleterious expansion when used in larger size.
The greater the proportions of reactive aggregate of the total aggregate, the greater
the expansion. Besides, potentially reactive aggregates may not cacuse extensive
expansion if the concrete is in a relatively dry condition. Where exposure to moisture is
accompanied by exposure to NaCl containing deicing salts, which are known to
exacerbate ACR, thean the conclusion that the aggregate is non-deleteriously expansive is
enhanced.
b. Concrete Prism Expansion Test:
The surest measure of the potential for deleterious expansion of a carbonate
aggregate in concrete is the measurement of the length change of concrete prism, as by
ASTM C 1105 or Canadian Standart CSA A23.2-14A.
ASTM 1105 is intendeal primarly for evaluating a particular cement-aggregate
combination prior to use in concrete.
CSA 23.2-14A Potential Expansivity of Cement Aggregate Combinations (Concrete
Prism Expansion Method), because of stipulations regarding use of the same gradiation
and same (high) alkali content, can be considered to be a test aimed at maximizing the

7-79


conditions for detection of potentially expansive rocks and facilitating comparisons among
test results.
Both methods use similarly sized test specimens 75x75 by 300 to 450 mm (CAN) or
285 mm (ASTM) and storage conditions of 100%RH and 23 C. However, ASTM C1105
specifies that if the job proposed aggregate has material >19mm, it shell not be used;
based on the assumption that plus 19 mm material does not differ in composition, lithology
and presumed expansive characteristics from the minus 19 mm material. However, in
those respects, where plus 19 mm material differs from minus 19 mm material; instructions
are given for testing the coarser fraction. In ASTM C1105, job cement is used when the
test is used for its primary purpose.
In CSA 23.2-14A the coarse aggregate is separated on 20-, 14-, 10-, and 5-mm
sieves, oversize and undersize .. discarded, and equal masses of the three sizes
recombined. The alkali content of cement shall be 1.0+ 0% Na
2
O equ. but raised by the
addition of NaOH to the mix water to 1.25% Na
2
O equ. by mass of cement to accelerate
expansion, in furtherance of the objective to identify potentially expansive rocks, rather
than to reflect job conditions.
Expansion limits for the CSA and ASTM prism tests are tabulated in Table.

Table 7.3. Concrete Prism Test, Expansion Limits and Alkali Content
Deleterious Expansion Limits
at Indicated Time (%)

Test
Na
2
O equ.
Alkali
Content
3 months 6 months 1 year
ASTM C1105 Job cement 0.015 0.025 0.03
CSA A.23.2-14A 1.25 % 0.01 ------ 0.025

Work by Roges and Hooton calls attention to the fact that the specified storage
conditions for both the CSA and ASTM concrete prism test allow leaching of the alkalies
from concrete resulting in reduced expansion. A known reactive aggregate produced
~0.15% at 52 weeks in the moist room at 23C, still well above the specification limit
0.025% but replicate prisms stored over water at 38C in a sealed box produced greater
than 0.3% expansion at 52 weeks. Determination of the water-soluble alkalies remaining in
the prism at 130 weeks indicated a loss of 63% versus 42% for the moist room versus the
sealed box storage, respectively. The authers recommended that the solution to detecting

7-80


alkali leaching problems during storage is the use of a reference reactive aggregate of
known performance. If the reference aggregate expansion is not within prescribed limits,
the test results on the unknown would be invalid, or suspect.





























7-81


7.4. Attack by Acids and Bases
Hydrated cement paste is an alkaline material and therefore specific attack by
alkaline materials will not normally be encountered. High concentrations of alkaline
materials that may come in contact with concrete in industrial processes cause
deterioration by processes other than direct chemical reaction with hydroxide ions. The
situation is entirely different for acidic solutions, which will readily attack concrete. The
effects of acidic solutions on concrete may be summarized as follows:
The hydrogen ion will accelerated the leaching of CH
Ca(OH)
2
+ 2H
+
Ca
+2
+ 2H
2
O

And, if highly concentrated, C-S-H may also be attacked, forming silica gel:

3CaO.2SiO
2
.3H
2
O+6H
+
H
2
O 3Ca
2+
+ 2(SiO
2
.nH
2
O) + 6H
2
O


The nature of the anion accompaning the hydrogen ion may further aggravate the
situation. Both sulfuric acid and carbonic acid are common constituents of acid rains and
acid groundwaters. The sulfuric ion will abviously participate in sulfate attack and thus
sulfuric acid is particularly corrosive. As it was mentioned earlier, carbonic acid is also very
corrosive due to the formation of soluble calcium bicarbonate.
Any acid forming soluble calcium salts will be particularly aggressive, while if the acid
forms an insoluble calcium salt its precipitation during acid attack can protect the concrete
against further deterioration. Some examples are given below:

Table 7.4. Acid Attack on Concrete
Acid Formula Likely Occurrence
Aggressive Acids Forming Soluble Calcium Salts
Hydrochloric acid HCl Chemical Industry
Nitric acid HNO
3
Fertilizer manufacture
Acetic acid CH
3
.COOH Fermentation processes
Formic acid H.COOH Food processing
Lactic acid C
2
H
4
(OH).COOH Dairy industry
Tonnic acid C
76
H
52
O
46
Tonnic industry, peat waters
Acids Forming Insoluble Salts
Phosphoric acid H
3
PO
4
Fertilizer manufacture
Tartaric acid [CH(OH).COOH]
2
Winemalcing



7-82


Sugar, although not an acid, dissolves CH and attacks slowly both C-S-H and
C-A-Hs. For this reason, sugar in solution is very aggressive to concrete and should not
be allowed to come in direct contact for more than brief periods.
It shoud be noted that corrosive chemicals can only attack concrete when water is
present. Thus, concrete can be used to store dry chemicals.






























7-83


7.5. Hydration of Free CaO and MgO
Hydration of free (crystalline or uncombined) CaO and MgO, when found in
excessive in cement, can cause expansion and cracking in concrete. The expansive
effects of MgO in cement was recognized as early as 1880s, when a number of concrete
structures in France and Germany failed roughly two years after construction.
It was reported that the expansive cements used in those constructions contained 16
to 30% MgO. This led to restrictions in the maximum permissible MgO in cement. It is
known that not all MgO present in cement is expansive. The MgO combined with other
oxides, and the x-talline MgO (periclase) in Portland cement clinker exposed to 1400 to
1500 C are essentially inert to moisture at room temperature, because when heated
above 900C the reactivity of periclase reduces sharply. In spite of this, the ASTM
Standart specification for cement limits the MgO content of cement to 6%. The
corresponding value in TS is 5%. However, no cases of structural distress due to the
presence of periclase in cement are reported from the countries (such as Brasil) where
manufacture of cement with more than 6% MgO is allowed (due to the composition of raw
materials).
The expansion due to hydration of crystalline CaO has been known for a long time.
The expansive effect of hard-burnt CaO in Portland cement was first recognized in the
U.S. in the 1930s. when some concrete pavements cracked and failed 2 to 5 years after
construction. Laboratory tests indicated that the paste prepared from a low MgO PC
containing 2.8% free lime showed considerable expansion. However, in concrete relatively
larger amounts of crystalline CaO are needed to cause significant expansion, because of
restraining effect of aggregate. In modern cements, due to better manufacturing controls,
the free lime content seldom exceeds 1%.
CaCO
3
~900 C CaO + CO
2


Soft-burnt lime (hydrates relatively rapidly)
Due to calcinations at elevated temperatures hard-burnt lime forms, which hydrates
slowly and causes expansion in hcp. The reason why free CaO and MgO cause expansion
lies in the fact that, these hydrate slowly to form their hydroxides, where, the hydration
reactions are expansive.



7-84



CaO + H
2
O Ca(OH)
2
(~1.3 times expansion)
MgO + H
2
O Mg(OH)
2
(~1.4 times expansion)

The soundness of cement may be detected by Lechatelier ring or autoclave test.
Another cause of expansion due to composition of cement is the presence of large
amounts of gypsum. Gypsum found in large amounts results in the formation of ettringite
even after setting. It is known that ettringite formation is accompanied with expansion. This
is why the amount of gypsum interground with clinker during the production of cement, is
limited to ~2-5% by weight of clinker. The best way to detect the gypsum content of the
cement is chemical analysis. It is assumed that sulfurtrioxide (SO
3
) reported in chemical
analysis of PC comes largely from gypsum added to the clinker. In order to prevent the
flash set without causing excessive expansion the SO
3
content of cement (depending on
C
3
A content of it) is limited to 3.5%.



















7-85


7.6. Corrosion of Sewer Pipes
Concrete sewer pipes may not only be corroded by substances in the groundwater
(sulfates or acids), but they are also more prone to corrosion from the sewage itself.
Furthermore, the thinness of the concrete pipe wall can make the effect of corrosion critical
in a much shorter time. If an unbalanced water pressure exits, flow of water through the
pipe wall may cause problems of leaching or salt depositions if wall thicknesses are less
than 30 to 40 cm. the W/C ratio of the concrete will obviously be an important parameter.
Domestic sewage or animal manures usually harmless to concrete unless the
reinforcing stell is inadequately protected. The greatest problem in sewer pipes is
corrosion caused by bacteria. The metabolic activity of many bacteria forms acids that may
attack concrete, but the most harmful results of bacteria action in sewers is the formation
of hydrogen sulfide (H
2
S) in the sewage from the reduction of sulfate compounds by
anaerobic bacteria. The hydrogen sulfide gas dissolves in water films in the upper part of
the wall, where there will be sufficient air to allow oxidation back to sulfate compounds by
aerobic bacteria. In this way high concentrations of sulfuric acid can be formed locally
which can corrode the upper part of the pipe very rapidly.

Protective treatment of sewer pipe surfaces will prolong the life of installations.
Coatings of bitumens, tars and resins have been used. Chemical treatments of the
concrete surfaces are another possibility, and, several are available which, when exposed
to the concrete, precipitate insoluble salts to fill the pores or provide a resistant surface
coating. Examples are water glass (sodium silicate), insoluble soaps, fluride salts, and iron
compounds. The surfaces can also be treated with carbon dioxide gas or silicon
tetrafluoride vapor. These gases form highly insoluble CaCO
3
or CaF
2
, respectively which
seal the surfaces.

7-86


7.7. Corrosion of Reinforcing Steel in Concrete
Concrete Reinforcement is invariably made from mild steel, which is suspectible to
corrosion. The formation of rust is an expansive reaction that will lead to cracking and
spalling of the concrete above the rusting steel. Thus, the corrosion of embedded steel is
found to be on of the principal causes of concrete deterioration. The failure of reinforced
concrete structures due to corrosion of reinforcement generally occurs in the age of ~10 to
40 years. However, failures as early as two years are also reported rarely.
The damage to concrete resulting from corrosion of reinforcement is generally in the
form of expansion, cracking and eventually spalling of the cover. In addition to loss of
cover, the reinforced concrete member may suffer structural damage due to loss of bond
between steel and concrete and loss of rebar cross-sectional area. Sometimes these
effects are to the extent that structural failure is inevitable.
It seems that when reinforcement is protected from air by a sufficiently thick cover of
low permeability concrete, the corrosion of steel and the problems associated with it would
not be faced. However, this expectation is not fully met in practice. This is evident from the
unusually high frequency with which even properly built reinforced and prestressed
concrete structures continue to suffer damage due to steel corrosion. The magnitude of
damage is especially large in marine structures and structures exposed to deicing agents.
For example, in 1975 it was reported that the U.S. Interstate Highway Syatem alone
needed $6 billion for repair and replacement of reinforced concrete bridge decks.
Approximately 4.800 of the 25.000 bridges in the state of Pennsylvania were found to be in
need of repair.
A review of the mechanisma involved in concrete deterioration due to corrosion of
reinforcement and measures for the control of the phenomenon are given here.
7.7.1. Mechanism Involved in Concrete Deterioration by Corrosion of
Embeded Steel
The corrosion of steel in concrete is an electrochemical (galvanic) process that
requires a flow of electrical current for the chemical corrosion reactions to proceed. The
electro-chemical potentials to form corrosion cells may be generated in two ways:




7-87


1. Composition cells may be formed when two dissimilar metals are embeded in
concrete, such as steel rebars and aluminium conduit pipes, or when significant variations
exist in the surface characteristics of the steel.



2. Concentration cells may be formed due to the differences in concentrations of
dissolved ion in the vicinity of steel, such as alkalies, chlorides and oxygen.
As a result one of the two metals (or some parts of the metal when only one metal is
present) becomes anodic and the other cathodic. The fundamental chemical changes
occurring at the anodic and cathodic areas are as follows:
Anode process
Fe 2e
-
+ Fe
2+

(metallic iron)

Cathode process
2
1
O
2
+ H
2
O +2e
-
2(OH)
-


Oxidation and reduction can occur only if electron transfer can take place, so that the
net redox reaction can proceed with the precipitation of ferrous hydroxide
Fe
2+
+ 2(OH)
-
Fe(OH)
2


The spontaneous oxidation of ferrous hydroxide to hydrated ferric hydroxide (rust)
occurs rapidly.
2Fe(OH)
2

O
2
, H
2
O 2 Fe(OH)
3
. 3H
2
O





7-88



Fig 7.4. Electrochemical process of steel corrosion in moist and permeable concrete
The galvanic cell constitutes an anode process and a cathode process. The anode
process cannot occur until the protective or the passives iron oxide film is either removed
in an acidic environment (e.g. carbonation of concrete) or made permeable by the action of
Cl
-
ions. The cathode process cannot occur until a sufficient supply of oxygen and water is
available at the steel surface. The electrical resistivity of concrete is also reduced in the
presence of moisture and salts.
The occurrence of corrosion is localized anodic areas, due to development of anodic
and cathodic areas can be caused by a variety of conditions, such as different impurity
levels in the iron, different amounts of residual strain, or different concentrations of oxygen
or electrolyte in contact with metal.
The transformation of metallic iron to rust is accompanied by an increase in volume
which, depending on the state of oxidation, may be as large as 600% of the original metal
causing expansion and cracking in concrete. This volume increase is believed to be the
principal cause of concrete expansion and cracking.

Fig 7.5. Volume expansion in metallic iron depending on the oxidation state.

7-89


It should be noted that the anodic reaction involving ionization of metallic iron will not
progress far unless the electron flow to the cathode is maintained by consumption of the
electrons at the cathode. Thus, the presence of both air and water at the surface of the
cathode is absolutely necessary. Besides, in alkaline environment of concrete, ordinary
iron and steel products are covered by a thin iron-oxide film which is impermeable and
strongly adherent to the steel surface, thus making the steel passive to corrosion. This
means that metallic iron will not be available for the anodic reaction until the passivity of
steel is destroyed.
In the absence of chloride ions in the solution, the protective film on steel is reported
to be stable as long as the pH of the solution stays above 11.5. Since hydrated portland
cements contain alkalies in the pore fluid and about 20 weight percent solid calcium
hydroxide, normally there is sufficient alkalinity in the system to maintain the pH above 12.
In exceptional conditions (e.g., when concrete has high permeability and alkalies and most
of the calcium hydroxide are either carbonated or neutralized by an acidic solution), the pH
of concrete in the vicinity of steel may be reduced to less than 11.5, thus destroying the
passivity of steel and setting the stage for the corrosion process.
The critical reduction of pH occurs when CH is converted to CC by carbonation. In a
well-cured and low W/C ratio concrete, the depth of carbonation rarely exceeds 25 mm,
and therefore a concrete cover of 25-40 mm over rebars should provide adequate
protection from corrosion in most instances. Where more severe conditions of exposure
are encountered or concrete with a fairly high permeability is used, the cover should be
increased to at least 50 mm.
Even in fully carbonated concrete, where the reserve of alkalinity has been lost,
corrosion of the reinforcing steel may not be a problem. The same applies to concrete
made with pozzolanic cements or high alumina cements, where CH is much reduced in
quantity or even is absent. Under such conditions the rate of corrosion depends on the
availability of oxygen and water at the surface of the steel. Concretes with a low
permeability will limit the rate of corrosion by limiting the rate of diffusion of oxygen. If the
concrete pores are filled with water, diffusion of oxygen will be further reduced. Thus, very
low or vary high moisture contents will prove beneficial in avoiding corrosion, while
intermediate moisture contents may be dangerous.




7-90


Effects of Chloride Ions:
Chlorides may enter into concrete from three major sources:
1- CaCl
2
added to the concrete as an accelerating admixture,
2- From deicing salts solutions, seawater or seapary,
3- From salt contaminated aggregate.

In the presence of chloride ions, depending on the Cl

/OH

ratio, the protective film of

reinforcement may be destroyed even at pH values considerably above 11.5. When
Cl

/OH

molar ratios are higher than 0.6, steel seems to be no longer protected against
corrosion, probably because the iron-oxide film becomes either permeable or unstable
under these conditions. For normal concrete structures, the threshold chloride content to
initiate corrosion is reported to be in the range 0.6 to 0.9 kg of Cl

per cubic meter of

concrete. Furthermore, when large amounts of chloride are present, concrete tends to hold
more moisture, which also increases the risk of steel corrosion by lowering the electrical
resistivity of concrete. Once the passivity of the embedded steel is destroyed, it is the
electrical resistivity and the availability of oxygen that control the rate of corrosion.
7.7.2. Control of Corrosion
Since water, oxygen, and chloride ions play important roles in the corrosion of
embedded steel and cracking of concrete, it is obvious that permeability of concrete is the
key to control the various processes involved in the phenomena. Concrete-mixture
parameters to ensure low permeability, e.g., low water/cement ratio, adequate cement
content, control of aggregate size and grading, and use of mineral admixtures, etc. were
discussed earlier.
Accordingly, ACI Building Code 318 specifies a maximum 0.4 water/cement ratio for
reinforced normal-weight concrete exposed to deicing chemicals and seawater. Proper
compaction and curing of concrete are equally essential. Design of concrete mixtures
should also take into account the possibility of a permeability increase in service due to
various physical-chemical causes, such as frost action, sulfate attack, and alkali-aggregate
expansion.
For corrosion protection, ACI Building Code also limits, the chloride contents of
concrete mixes. For example, maximum watersoluble Cl

ion concentration in hardened

concrete, at an age of 28 days, from all sources (including aggregates, cementitious
materials, and admixtures) should not exceed

7-91


- 0.06 percent by weight of cement for prestressed concrete,
- 0.15 percent by weight of cement for reinforced concrete exposed to chloride in
service,
- 0.30 percent by weight of cement for other reinforced concretes,

Reinforced concretes that will remain dry or protected from moisture in service are
permitted to contain up to 1.00 percent Cl

be weight of cement. Besides ACI Building

Code specifies minimum concrete cover of 50 mm for walls and slabs and 63 mm for other
members when concrete is exposed to corrosive environment. Current practice in the
North Sea is to provide minimum 50 mm of cover on conventional reinforcement and 70
mm on prestressing steel.
Also, ACI 224R specifies 0.15 mm as the maximum permissible crack width at the
tensile face of reinforced concrete structures subject to wetting-drying or seawater spray.
The CEB Model Code recommends limiting crack width of 0.1 mm at the steel surface for
concrete members exposed to frequent flexural loads, and 0.2 mm to others.
The FIP (International Prestressing Federation) recommendations specify that crack
widths at points near the main reinforcement should not exceed 0.004 times the nominal
cover (i.e., maximum permissible crack width for a 500-mm cover is 0.2 mm, which
increases to 0.3 mm by increasing cover to 75-mm).
No direct relation is found between the crack width and corrosion; however, it
appears that by increasing the permeability of concrete and exposing it to numerous
physical-chemical processses of deterioration, the presencd of cracks would eventually
have a deleterious effect.
The repair and maintenance or replacement costs of concrete elements damaged by
corrosion of reinforcement have become a major expense. As a result, the extra initial cost
of providing a waterproof membrane, or a thick overlay of an impervious concrete mixture
on newly constructed or repaired surfaces of reinforced or prestressed concrete structures
having large and have flat surfaces is preferred.
Waterproof membranes, of sheet-type variety are used. These membranes should be
protected from physical damage or wear by asphaltic concrete overlays. Thus, their
surface life is limied to the life of the asphaltic layer, which is about 15 years. Overlay of
watertight concrete 40 to 65 mm thick provides a more durable protection to penetration of
aggressive fluids into concrete.
Typically, concrete mixtures used for overlay are of low slump, very low water/cement

7-92


ratio (using superplasticizing admixtures), and high cement content.
Portland cement mortars containing polymer emulsion (latexes) also show excellent
impermeability and have been used for overlay purposes. (Now, styrene-butadiene-type is
preferred since vinylidene chloride-type latexes are suspected to be the cause of corrosion
problems in some cases).
Reinforcing bar coatings and cathodic protection provide other approaches to prevent
corrosion and are more expensive than producing a low-permeability concrete through
quality, design, and construction controls. Protective coatings for reinforcing steel are of
two types:
- anodic coatings (e.g., zinc-coated steel) and
- barrier coatings (e.g., epoxycoated steel).
Due to concern regarding the long-term durability of zinc-coated rebars in concrete,
the U.S. Federal Highway Administration in 1976 place a temporary moratorium on its use
in bridge decks. Long-time performance of epoxy-coated rebars is under serious
investigation in many countries. Cathodic protection techniques involve suppression of
current flow in the corrosion cell, either by supplying externally a current flow in the
opposite direction or by using sacrificial anodes.
7.8. Effects of Seawater on Concrete
Oceans make up 80 percent of the surface of the earth; therefore, a large number of
structures are exposed to seawater either directly or indirectly (e.g., winds can carry
seawatere spray up to a few miles inland from the coast).
Most seawaters are fairly uniform in chemical composition, which is characterized by
the pesence of about 3.5 percent soluble salts by weight. The ionic concentrations of Na
+
and Cl

are the highest, typically 11,000 and 20,000 mg/liter, respectively. However,
sufficient amounts of Mg
2+
(1400 mg/liter) and SO
4
2
(2700 mg/liter) are present to harm
cement hydration products. The pH of seawater varies between 7.5 and 8.4, the average
value in equilibrium with the atmospheric CO
2
being 8.2. Under exceptional conditions
(i.e., sheltered bays) pH values lower than 7.5 may be encountered.
Concrete exposed to marine environment may be deteriorated as a result of
combined effects of chemical action of seawater constituents on cement hydration
products, alkali-aggregate expansion (if reactive aggregate is present), crystalization
pressure of salts within concrete if one face of the structure is subject to wetting and others

7-93


to drying conditions, frost action in cold climates, corrosion of rebars and physical erosion
due to wave action and floating objects. Attack on concrete due to any one of these
causes tends to increase the permeability of concrete, thus, making it more susceptible to
further action by the same destructive agent or to other types of attack.
7.8.1. Theoretical Aspects Involved in Seawater Attack
For onreinforced concrete (with no alkali reactive aggregate) it may be anticipated
that sulfate and magnesium are harmful constituents in seawater. It may be recalled that
with groundwaters, sulfate attack is classified as severe when the SO
4
-2
concentration is
higher than 1500 mg/liter; similarly, Portland cement pastes may deteriorate by cation-
exchange reactions when Mg
+2
concentration exceeds, ~ 500 mg/liter.
Interestingly, in spite of high sulfate content of seawater, it is observed that even
when a high-C
3
A portland cement is used and large amounts of ettringite are present as a
result of sulfate attack, the deterioration of concrete is not characterized by expansion;
instead, it mostly takes the form of erosion or loss of the solid constituents from the mass.
It is proposed that ettringite expansion is suppressed in environments where (OH)

ions have essentially been replaced by Cl

ions. This view is consistent with the

hypothesis that alkaline environment is necessary for swelling of ettringite by water
adsorption.
According to ACI Building Code, sulfate exposure to seawater is classified as
moderate, for which the use of ASTM Type II Portland cement (< 8% C
3
A content) with a
0.50 maximum water/cement ratio in normal-weight concrete is permitted. It is also stated
that cements with C
3
A up to 10 %may be used if water/cement ratio is limited to 0.4.
From long-term studies on concrete exposed to seawater, the evidence of
magnesium ion attack is well established by the presence of white deposits of MH, also
called brucite and M-S-H. In seawater, well-cured concretes containing large amounts of
slag or pozzolan in cement are superior to the concrete containing only Portland cement,
partly becuse the former contain less uncombined CH after curing. The implication of loss
of lime by cement paste, whether by magnesium ion attack or by CO
2
attack, is obvious
from Fig 7.6 below.





7-94


100
75
50
25
0
10 20 30
Dissolved CH expressed as %CaO
C
o
m
p
r
e
s
s
i
v
e

S
t
r
e
n
g
t
h

(
%
)
After Moskvin
A
v
e
r
a
g
e
After Breezky
Strength loss in concrete
due to lime leaching.
strength drops ~2% when
1% CaO is removed from
hcp.
Since seawater analyses seldom include the dissolved CO
2
content, the potential
for loss of concrete mass by leaching away of calcium from hcp due to carbonic acid
attack is often overlooked.
It should be noted that in permeable concrete the normal amount of CO
2
present in
seawater is sufficient to decompose the cementitious products eventually. The presence of
thaumasite (calcium silicocarbonate), hydrocalumite (calcium carboaluminate hydrate),
and aragonite (calcium carbonate) have been reported in cement pastes derived from
deteriorated concretes exposed to seawater for long periods.

Lessons from Case Histories
For the construction of concrete sea structures, important lessons from case histories
of concrete deteriorated by seawater are given below:
1. Permeability is the key to durability. Deleterious interactions of serious
consquence between constituents of hydrated portland cement and seawater take place
when seawater is not prevented from penetrating into the interior of a concrete. Typical
causes of insufficient watertightness are poorly proportioned concrete mixtures, absence
of properly entrained air if the structure is located in a cold climate, inadequate
consolidation and curing, insufficient concrete cover on embedded steel, badly designed or
constructed joints, and microcracking in hardened concrete attributable to lack of control of
loading conditions and other factors, such as thermal shrinkage, and ASR.




7-95



7-96

2. Type and severity of deterioration may not be uniform throughtout the structure
Generally that part of structure fully immersed in seawater suffers only a little,
concrete exposed to salts in air or waterspray suffers some deterioration, especially when
permeable and concrete subject to tidal action sufferes the most.
It is observed that progressive chemical deterioration of cement paste by seawater
from the surface to the interior of the concrete follows a general pattern. The formation of
aragonite and bicarbonate by CO
2
attack is usually confined to the surface of concrete, the
formation of brucite by magnesium ion attack is found below the surface of concrete, and
evidence of ettringite formation in the interior shows that sulfate ions are able to penetrate
even deeper. Unless concrete is very permeable, no damage results from chemical action
of seawater on cement paste because the reaction products (aragonite, brucite, and
ettringite), being insoluble, tend to reduce the permeability and stop further ingress of
seawater into the interior of the concrete. This type of protective action would not be
available under conditions of dynamic loading and in the tidal zone, where the reaction
products are washed away by wave action as soon as they are formed.

3. Corrosion of reinforcing bar is, generally, the major cause of concrete deterioration
in reinforced and prestressed concrete structures exposed to seawater, but in low-
permeability concrete, this does not appear to be the first cause of cracking.