SCHOOL OF CIVIL AND ENVIRONMENTAL ENGINEERING

Private Bag 3, Wits 2050, South Africa | Tel: +27 11 717 7104 | Fax: +27 11 717 7045

Excel . Together .

Durability of Concrete

1
 Introduction: What is durability of concrete?
Durability is the ability of a material or structure to withstand its design service conditions for its design life
without significant deterioration. It is its resistance to weathering action, chemical attack, abrasion and other
degradation processes
degradation processes.

Durability is mainly related to the transport properties and chemical composition of concrete. Transport
properties are a function of concrete penetrability (which is a function of pore structure - including porosity,
pore size distribution and pore continuity) and cracking. Depending on the concrete’s moisture condition and
environmental exposure, moisture and harmful substances can be transported into it by various mechanisms
such as absorption, permeation, capillary suction and diffusion.

Penetrability is the ease with which an ion, molecule or fluid may move through the concrete. It is a function
long + complex too

of interconnection between the pores, the pore size distribution and its tortuosity. It is affected by several _

parameters such as the water/binder (w/b) ratio and the degree of hydration of the cement paste. A reduction
of penetrability and elimination of cracking or minimization of the cracking increases durability of concrete.

The two main factors affecting concrete durability are (see Fig. 1):
a) the aggressiveness of the exposure environment, and
b) the quality of the concrete (which is affected by both intrinsic and extrinsic factors)

When designing for concrete durability, it is important to remember that even though improved durability may
be achieved with increase in concrete strength, this is not always the case unless the two parameters are
in mind
designed for. design for durability + concrete strength

Concrete durability

how to explain each Aggressiveness of the

know The concrete system
factor ! exposure environment

Extrinsic factors Physical attack Chemical attack

Intrinsic factors
 Production and/or  Abrasion  Nature & concentration of
 Concrete pentrability
construction processes e.g.  Erosion aggressive agents
 Binder type
mixing, placing,  Cavitation  Internal chemical
 Binder content
compaction  Freeze-thaw instability (incompatibility
 Water-to-binder (w/b) ratio
 Curing (temperature and  Salt crystalization between mix constituents)
 Other constituents e.g.
moisture environment)  Effects of cracking due to  Coupling with effects of
aggregates, admixtures, etc
 Early-age temperature loading or thermal / hygral temperature and relative
 Design mix proportioning
history effects humidity
Figure 1: Factors influencing durability of concrete (from Fulton’s Concrete Technology, 10th Edition)
galvanic cell

cursdeSalytic
{
anode
- I cathode

Common deterioration mechanisms in concrete oxidation

* J
'
reduction ·

acceptors
occurs
Ce donor) Ce
cells

Corrosion of steel in concrete electrolyte

Corrosion is a dissolution process that occurs when metals are exposed to corrosive environments. It is an
electrochemical process in which iron (Fe) enters into solution at the anode and oxygen (O2) is reduced at the
cathode.

The conventional carbon steel grades commonly used as reinforcing steel in concrete are susceptible to
corrosion when exposed to the atmosphere. However, steel embedded in concrete is naturally protected by:

2
 • the high alkalinity of the cement matrix (pH > ~12.5) - the high pH maintains the stability of a thin
protective corrosion layer on the embedded steel surface hence protecting the steel from further
corrosion. In this state, the steel is said to be in a passive state.
• the barrier effect of the concrete cover.

The basic anodic and cathodic reactions in a steel corrosion process, irrespective of the cause of corrosion, are
as follows:

• Anodic reaction (oxidation): Fe → Fe2+ + 2eˉ

• Cathodic reaction (reduction): 2H2O + O2 + 4eˉ → 4OHˉ

In addition to the anode (where corrosion occurs and from which current flows) and cathode (where no
corrosion occurs and to which current flows), the following are also essential to sustain the corrosion process
(Fig. 2):

• Electrolyte: a medium capable of conducting electric current by ionic current flow. In the case of
liquid phase
within hardened concrete, the (alkaline) pore solution constitutes the electrolyte and,
concrete
pores ;
• Metallic path: connection between the anode and cathode, which allows current return and completes
contains water

alkali
,
the circuit. Steel serves this purpose in reinforced concrete.
dissolved

hydrates + calcium hydroxide

NaOH and KOH Ca(OH) z Ingress of chloride ions
and/or carbon dioxide
Concrete surface

O2, H2O
Cover
concrete
Corrosion OH– flow
products

ANODE Electron flow CATHODE

Steel reinforcement
Figure 2: A schematic illustration of the corrosion process in concrete

Corrosion of steel in concrete results in the formation of rust (corrosion products) which have a higher volume
(and lower density) compared to that of steel. It is this increased volume that results in delamination, cracking
and spalling of the concrete cover.

The two main causes of steel corrosion in concrete are:

a) Carbonation
b) Chloride attack

3
 Fe

FeO

Fe3O4
Fe2O3
Fe(OH)2

Fe(OH)3
Fe(OH)3·3H2O
0 1 2 3 4 5 6 7
Volume, cm3 (relative)

Craking formation of lines on concrete

surface

Corroding
reinforcing bar

Spalling formation of
fragments that
break off from the
larger
concrete
body

Delamination separation betwee

layers/lack
of
bonding between
adjacent
layers

Carbonation-induced corrosion movement from↑ concentration to

↓ concentration
Carbonation involves the following process:
(i) Ingress of carbon dioxide (CO2) from the environment into concrete. This is a diffusion process.
(ii) The CO2 reacts with moisture in the concrete pore solution for form weak carbonic acid (H2CO3) i.e.
CO 2 + H 2 O → H 2 CO 3
(iii) H2CO3 then reacts with calcium hydroxide, Ca(OH)2 (a cement hydration product which contributes to
the high pH of the concrete pore solution) for form calcium carbonate precipitate (CaCO3) i.e.:
H 2 CO 3 + Ca (OH )2 → CaCO 3 + 2H 2 O
This reaction leads to the pH of the concrete pore solution dropping from ~12.5 to ~8.5 which renders the
thin protective layer on the steel surface unstable (this protective layer is stable at high pH values). The
protective layer is therefore destroyed and the steel is exposed to further corrosion.

The depth of carbonation (i.e. the carbonation front) can be estimated spraying phenolphthalein solution (a pH
indicator) on a freshly exposed concrete surface. Phenolphthalein solution:
• remains colourless if the pH of the concrete is ~8.5 to 9.0
• changes colour to pink/purple if the pH of the concrete is > ~9.0
uncarbonated concrete

9
pH >

8
8, 5 < PH29
carbonated
concrete

4
 carbonation
Factors NOT
affecting rate

{
② Dry conditions curban dioxide with water to carbonic acid .
mixes Carbonic calcium hydroxide
> form acid attacks
-

relatedelative
to in the pore solution. No moisture = no carbonic acid formation = no carbonation

humidity
② very wet conditions (water saturated) >
- ↑ amount of water = carbon dioxide cannot enter [Why? Because C2 travels
very slowly
I
through water cannot
penetrate the concrete
deeply enough to cause carbonation.

Factors affecting carbonation rates

↑temp If temp gets

%

chemical
high
:
20 C
Temp speed & too moisture
> rxns in
evaporates
=
concrete
-
.
up .

↓ CO2
°

solubility in water. Temp of 20 C creates optimal balance of rxn

speedA moisture retension.

>
Relative humidity :
CO2 must dissolve in water to
produce H2COs ↑
humidity (80% )
-

are
concrete
pores
=
·

( 40%)
completely filled with water =
CO2 is blocked from
diffusing into concrete · ↓
humidity < = not
enough water for

carbonation to occur . 50-70 % relative

humidity =
pores are
partially
filled with water
(allowing
. run to
occurs + pores are

into
are
open enough to allow CO2 to
penetrate concrete
.

>
-

↑ Sco2] : rate & which CO2 diffuses into concrete is driven

by Concentration gradient between outside air

A the ↑ [CO2]
inside A
accelerating the
concrete
surrounding
of environment pushes into faster
.
in CO2 concrete ,

d ifference in the concentration of substance across distance

carbonation a
process.
a

>
-

w/b ratio > 0

,
00 :
wlb ratio determines
porosity of concrete
. ↑ wlb ratio ( > 60% ) = more water in the mix than
interconnected
necessary for hydration= as excess water
evaporates
= it leaves behind a network of
ycapillary pores =
↑
porosity =

concrete becomes more

permeable =
CO2 easily penetrates deeply into structure

of fly ash materials CaOHz their pozzolanic

slag
use consume reactions .
supplementary during
>
-
or : cementitious ,

consumption of CalOH) =
↓CaCOH)2 to resist CO2 = concrete carbonates & loses its pH faster than
pure PC mixtures

cement comes into contact wl water => solid matrix

primaniy composed of calcium silicate hydrates (CSH) which

contribute the time

to
strength over
 Note: In colour, the top portion appears colourless while the bottom portion is coloured purple

Carbonation, being a diffusion-controlled process, is governed by Fick’s 1st law of diffusion and is generally
modelled by means of the simplified expression: d c = k c t where dc is the depth of carbonation (mm), kc is
the coefficient of carbonation (mm/√(year)) and t is the time of exposure (years).

Factors affecting rate of carbonation:

 Favourable conditions for increased carbonation rates include:
 temperatures ~ 20 °C
 relative humidity ~ 50-70%
 increased CO2 concentrations
 water-to-binder (w/b) ratios at or above 0.60
 use of fly ash or slag
 Carbonation does not occur if the concrete is water-saturated or in very dry conditions because:
 moisture is required to form carbonic acid which attacks the Ca(OH)2 and
>
-

 Diffusion of CO2 through water is very slow H2COs

Carbonation-induced corrosion is characterized by a generalized/microcell type of steel corrosion.

Chloride-induced corrosion
Contrary to carbonation-induced corrosion where corrosion is induced by reduction in pH in order to destroy
the protective passivating layer on the embedded steel surface, in the case of chloride-induced corrosion, the
protective layer is destroyed when the chlorides at the steel surface accumulate and reach a critical
concentration – referred to as chloride threshold.

The chloride threshold is not a universal value and varies from concrete to concrete. However, a value of 0.4%
(by mass of cement) is conventionally used in cases where accurate values are not available.

Possible sources of chlorides in concrete include:

(i) Internal sources: contaminated raw materials, admixtures (e.g. CaCl2)
(ii) External sources: sea water (including airborne chlorides), contaminated soils, de-icing salts

Chloride ingress is also a diffusion process but is governed by Fick’s 2nd law of diffusion. Crank’s solution is
used to determine the chloride diffusion coefficient (Dc):

∂C ∂ 2C
=D 2
∂t ∂x

5
   x 
C x ,t = C s 1 − erf  
  2 Dt 

where Cx,t is the chloride concentration at the depth x after a given time of exposure t, Cs is the chloride
concentration at the surface of the concrete, D is the chloride diffusion coefficient and erf is the mathematical
error function.

Depth and concentration of chloride penetration in concrete can be determined as follows:

1. Drill concrete cores from an in-situ structure.
2. Cut thin slices (~5-10 mm) from the core in a direction perpendicular to the direction of chloride
penetration.
3. Each thin slice is ground into fine powder (i.e. pulverized).
4. Extract chlorides from the powder using an acid e.g. HNO3.
5. Perform a chemical titration on the resultant solution using a suitable titrant e.g. AgNO3.

Chloride ingress is mainly affected by concrete quality which is affected by factors such as:
 Type of binder used (NOT BINDER CONTENT!) – blended cements are preferred to plain Portland
cements
 w/b ratio
 curing and compaction
 presence of cracks, etc

Chloride-induced corrosion is characterized by a pitting/macrocell type of steel corrosion.

Synergy between
working
-
together
carbonation and chloride ingress
When carbonation takes place in concrete into which chlorides have already penetrated, the two processes
interact to accelerate the depassivation (i.e. destruction of the thin protective layer on the steel surface) of the
reinforcement. The lowering of the pH due to carbonation (from ~12.5 to ~8.5) causes some of the chlorides
bound to the cement (chemically and physically) to be released thereby increasing the amount of free chloride
ions available for depassivating the steel reinforcement.

Factors affecting initiation and propagation of steel corrosion in concrete (summary):

 Cover thickness
 Cover condition e.g. cracked or uncracked
 Cover quality i.e. resistance to penetration of corrosion-inducing and sustaining species (Clˉ, CO2, O2,
H2O). Cover quality is a function of w/b ratio, curing, binder type, level of compaction
 Chloride binding capacity of the cement used
 Binder type
 Chloride threshold – for chloride-induced corrosion
 CO2 concentration – for carbonation-induced corrosion
 Temperature (up to 40 ºC)
 Moisture content of the concrete

Effects of reinforcement corrosion on RC structures

 Rust stains – unsightly

6
  Reduction in steel reinforcement diameter – reduced load-carrying capacity
 Cover cracking, delamination and spalling
 Structural failure

Sulphate attack
Sources of sulphates in concrete:
(a) Internal – SO3
(b) External – sea water (SO42- and Clˉ), soils (natural or from fertilizers), industrial effluents, sewage
pipes and treatment plants.
sodium sulphate potassium
↑ ~
Forms: Na2SO4, MgSO4, KSO4, CaSO4
↓ し calcium
magnesium
The sulphates attack C3A (3CaO·Al2O3·12H2O - from Portland cement, PC) and its hydrates (Ca(OH)2,
abbreviated CH) forming ettringite in already hardened concrete. The formation of ettringite is expansive
resulting in cracking.

In the case of external sulphate attack, sulphates from an external source penetrate concrete and dissolve into
solution. The sulphates then react with CH to form gypsum in the pore solution, which in turn converts mono-
Talcium
sulphate to ettringite.
hydnexiae

Chemical reactions (in the case of attack by calcium sulphate):

C3A from PC reacts with gypsum to form ettringite (i.e. insoluble calcium sulphoaluminate):

^ ^
C 3 A + 3C S⋅ H 2 + 26H → C 3 A ⋅ 3C S⋅ H 32
(gypsum) (ettringite)

As the amount of gypsum (sulphates) dissolved in the concrete pore solution decreases, ettringite becomes
unstable and converts to mono-sulphate:

^ ^
2C 3 A + C 3 A ⋅ 3C S ⋅ H 32 + 4H → 3C 3 A ⋅ 3C S⋅ H 12
(ettringite) (mono-sulphate)
If there is an excess of sulphates dissolved in solution, the mono-sulphate reverses to ettringite resulting in
expansion and cracking in hardened concrete. If the dissolved sulphates are from excess gypsum, this results
in internal sulphate attack.
^ ^ ^
C 3 A ⋅ 3C S⋅ H 12 + 2C S⋅ H 2 + 16H → C 3 A ⋅ 3C S⋅ H 32

If however the excessive dissolved sulphates in concrete are from an external source of sulphates (e.g. Na2SO4),
the sulphates react with CH to form gypsum in the pore solution, which in turn converts mono-sulphate to
ettringite i.e. external sulphate attack.

Delayed ettringite formation (DEF) is a form of sulphate attack known to occur in concretes subjected to
elevated temperature curing in excess of 70 ºC e.g. heat-cured precast concrete elements. When fresh concrete
is subjected to high temperatures, the ettringite that normally forms during hydration is destroyed releasing

7
sulphate ions which then get adsorbed on the CSH (calcium silicate hydrate). In the presence of moisture, the
sulphate ions are slowly released from the CSH over many years and react with mono-sulphate re-forming
ettringite and causing disruption in the hardened concrete.

Sulphate concentration in water (and concrete pore solution) is greatly increased by evaporation (low RH).

Damage (cracking) due to expansion is accelerated with wetting and drying cycles due to salt crystallization
in the concrete pores.

It can be controlled by:

• Use of sulphate resisting Portland cement (SRPC) which has low C3A (<3.5%)
• Use of cement extenders (supplementary cementitious materials, SCMs) e.g. 50% GGBS. The SCMs
lower concrete penetrability, reduce Ca(OH)2 content and dilute C3A.
• In the case of DEF, select appropriate blended cement that does not lead to an excessive temperature
rise due to hydration.

Test method for sulphate resistance

Mortar bars 25 × 25 × 280 mm are cast and cured until 20 MPa strength is attained. They are then immersed
in 5% Na2SO4 solution and kept at 23 ºC. Expansion after 6 months immersion should be < 0.10%. This method
is for external sulphate attack e.g. to evaluate concretes containing extenders, varied w/b ratios.

Note: Instead of Na2SO4 solution, MgSO4 solution, or a mixture of the two can also be used in the test.

Alkali-silica reaction (ASR)

Occurs when certain reactive forms of silica (e.g. opal) in some aggregates (e.g. greywacke) react with alkalis
in the concrete pore solution forming ASR gel which results in expansion within the concrete, and consequent
map cracking i.e. internal material incompatibility.

The alkalis (Na+ & K+) in the concrete pore solution originate from: cement, admixtures (e.g. superplasticizers),
aggregates (e.g. sea sand)

The necessary conditions for ASR to occur include:

(i) Sufficiently high alkalinity of the concrete pore solution
(ii) Presence of sufficient amount of reactive mineral(s) in the aggregates i.e. reactive silica
(iii) Sufficient moisture. It has been suggested that the cut-off internal relative humidity of concrete below
which deleterious ASR will not occur, is in the range 75 to 85%.

Effects of ASR on concrete structures:

• Decreases strength
• Lowers the modulus and stiffness
• Increases creep

The risk of ASR can be minimized by:

• Control the alkali content of the cement. A cement having %Na2O-eq < 0.6 is generally regarded as a low-
alkali cement. The cement content should also be limited to ~350 kg/m3.

8
 • Use of cement extenders – GGBS ≥ 40%, SF ≥ 15%, FA ≥ 20% (Percentages are w.r.t. the total mass of
the cement).
• Limiting access of water – to protect against reaction and expansion.

Test method for ASR resistance

Concrete prism test: 75 × 75 × 300 mm concrete prisms (or cores) are stored over water at 38 ºC. Expansion
at 52 weeks (~1 year) suggests a reactive aggregate.

Acid attack (and soft water attack)

The pore water solution in hardened cement is highly alkaline with a pH of around 12,5. In this environment
the calcium silicate, aluminate and ferrite hydrates, the main binding components of cement, are stable.
However, as Ca(OH)2 is leached by the soft water, the pH of the pore water decreases (from ~12.5 to about
~5.5), causing the hydrates (mainly Ca(OH)2) to become unstable and release calcium oxide (CaO) in order to
re-establish the alkalinity. In this manner the binding agent in the cement phase is attacked, causing a decrease
in strength. It can be mitigated by:
• Application of waterproof surface coatings – external and internal (e.g. in tanks, concrete pipes)
• Application of a sacrificial concrete layer

Abrasion gradual surface wear

by friction
caused
scraping sliding impact from machinery ,etc
,
, ,

gradual This problem occurs in concrete surfaces that may be subjected to abrasive surface wear. Sliding or scraping
wearing away
degradation can cause attrition on wearing surfaces.
of
aggregate • In hydraulic structures, abrasion is by solid particles carried in the flowing water
partices

• Industrial floors, pavings, roads, grain hoppers are subject to mechanical abrasion.

At initial stages of abrasion, the wearing surface shows slight roughness as the few millimetres of the surface
layer is eroded. This can deteriorate to expose aggregates in more severe cases. In general, abrasion resistance
of concrete is improved by ensuring that a particularly good quality surface is achieved. Particularly:
• Give attention to surface finishing techniques and moist curing at early ages
• Consider using surface hardening agents

The abrasion resistance of concrete is determined by a combination of strength and hardness of the aggregate
together with the strength of the aggregate-cement paste bond (indirectly measured by the compressive
strength). Concretes of high compressive strength and hard surfaces will show a good resistance to abrasion.
Extenders tend to produce soft surfaces and thus may not be suitable.

South African approach to design and testing for concrete durability

Different tests have been developed in different parts of the world to assess concrete durability. However, the
harmful unexpected way
in an

fundamental similarity of these tests is that they measure different transport mechanisms used by deleterious
substances (e.g. CO2, chlorides, H2O) to penetrate the concrete in order to assess the potential durability of a
given concrete. In SA, three durability index (DI) tests have been developed that measure the resistance of
cover concrete to the transport of fluids and ions that affect deterioration. The purpose of the tests is to produce
reliable index values to be used for characterisation of concretes. The three DI tests comprise of CCT, OPT
and WST. The values obtained from these tests are used to provide indices (CCI, OPI and WSI) which can be
used to predict likely material (concrete) performance.

9
Specimen type, size and pre-conditioning
The recommended size of the test specimens in all the three DI tests are (concrete or mortar) discs. The test
specimens for all the 3 tests are first pre-conditioned by drying in an oven at 50 ºC for 7 days before testing
(to ensure uniform moisture content). Further pre-conditioning by vacuum-saturating in 5M NaCl solution is
required in the chloride conductivity test before actual testing. Detailed test procedures for each test are given
in the Durability Index Testing Manual available on Wits-e.

Oxygen permeability test

The test involves determination of the OPI of a concrete from the rate of pressure decay through the test sample
when placed in a falling head permeameter. The rate of pressure decay of a falling head permeameter is
governed by Darcy’s law, allowing the coefficient of permeability (k) to be determined. As permeation occurs,
the decrease in pressure with time is measured. From the slope of the log of pressure head versus time, the
oxygen permeability coefficient may be determined as follows:
ωVgd Po
k= ln
RAθt p

Top screw

Solid ring

Cover plate
Permeating
Rigid sleeve gas outlet
Compressible collar

Air-tight seal
Concrete specimen

Concrete specimen in
compressible collar in
rigid sleeve
Gas inlet valve connected
to oxygen cylinder

Pressure vessel
containing oxygen

Connection to pressure
transducer and data logger

Gas outlet
valve

where k is the coefficient of permeability (m/s), ω is the molecular mass of permeating gas (kg/mol), V is the
volume of the pressure cylinder (m3), g is acceleration due to gravity (m/s2), d is sample thickness (m), R is
universal gas constant (Nm/Kmol), A is cross sectional area of specimen (m2), θ is absolute temperature (K)
and Po, P are pressures at start of test and at time t (sec) respectively (kPa). To obtain a less cumbersome
number, the negative log of k is used, and termed the ‘OPI’ i.e. OPI = − Log 10 k . Typical OPI values range
from 8 to 11.

→ Concrete resistance to gaseous ingress increases with increase in OPI.

10
Water sorptivity test
This test is used to measure the rate of movement of a water front through a concrete by capillary suction. The
vertical sides of the specimens are sealed using epoxy or sellotape and placed in a few millimetres depth of
saturated Ca(OH)2 solution and weighed at regular intervals (0, 3, 5, 7, 9, 12, 16, 20 and 25 minutes) to
determine the mass of water absorbed. The sorptivity, S, of the concrete is obtained from the slope of the
straight line of graph of mass of water absorbed versus the square root of time:

ΔM t  d 
WSI = S = 
1/ 2 

t  M sat − M o 

where ΔMt is the change in mass with respect to the oven dried mass (g), Mo is the mass of oven-dried concrete
(g), Msat is the mass of saturated surface dry concrete (g), t is the period of absorption (hr) and d is the average
thickness of sample (mm).

→ Concrete resistance to moisture ingress increases as the WSI decreases.

Chloride conductivity test

This test is used to assess the potential of a concrete to resist chloride penetration. It involves measuring the
ionic flux (current) through a sample due to an applied ~10 V potential difference. The test allows for a rapid
determination of the chloride conductivity index (CCI) value of the test concrete which is defined as:

it
CCI = σ =
VA

where σ is the conductivity of the specimen (mS/cm i.e. milliSiemens/cm), i is the measured electric current
through the specimen (mA), V is the voltage difference across the specimen (V), t is the average thickness of
specimen (cm) and A is the cross-sectional area of the specimen (cm2). The CCI (σ) is converted to a diffusion
coefficient (D) value using empirical relationships. D is then used in to predict rate of chloride ingress in
concrete.

Voltmeter
V
DC power source
Ammeter
– + A

5 M NaCl solution

. . . .
Concrete
specimen Luggin
capillaries
PVC
cap-screw

Stainless steel Rubber collar Stainless steel

CATHODE ANODE

→ Concrete resistance to chloride penetration (ingress) increases as CCI decreases.

11
Designing for durable concrete structures
What are the general steps to be followed in designing for durable concrete structures?
To be covered in class…

12