SIST-TP CEN/TR 15697:2008

SLOVENSKI STANDARD
SIST-TP CEN/TR 15697:2008
01-julij-2008
Cement - Preskušanje sulfatne odpornosti - Dokument o stanju tehnike
Cement - Performance testing for sulfate resistance - State of the art report
Zement - Prufüng der Leistungsfähigkeit hinsichtlich des Sulfatwiderstands - Bericht zum
Stand der Technik
Ciment - Essais de performances relatifs à la résistance aux sulfates - État de l'art
Ta slovenski standard je istoveten z: CEN/TR 15697:2008
ICS:
91.100.10 Cement. Mavec. Apno. Malta Cement. Gypsum. Lime.
Mortar
SIST-TP CEN/TR 15697:2008 en
2003-01.Slovenski inštitut za standardizacijo. Razmnoževanje celote ali delov tega standarda ni dovoljeno.

---------------------- Page: 1 ----------------------

TECHNICAL REPORT
CEN/TR 15697
RAPPORT TECHNIQUE
TECHNISCHER BERICHT
April 2008
ICS 91.100.10

English Version
Cement - Performance testing for sulfate resistance - State of
the art report
Ciment - Essais de performances relatifs à la résistance Zement - Prufüng der Leistungsfähigkeit hinsichtlich des
aux sulfates - État de l'art Sulfatwiderstands - Bericht zum Stand der Technik
This Technical Report was approved by CEN on 6 November 2007. It has been drawn up by the Technical Committee CEN/TC 51.
CEN members are the national standards bodies of Austria, Belgium, Bulgaria, Cyprus, Czech Republic, Denmark, Estonia, Finland,
France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, Netherlands, Norway, Poland, Portugal,
Romania, Slovakia, Slovenia, Spain, Sweden, Switzerland and United Kingdom.
EUROPEAN COMMITTEE FOR STANDARDIZATION
COMITÉ EUROPÉEN DE NORMALISATION
EUROPÄISCHES KOMITEE FÜR NORMUNG
Management Centre: rue de Stassart, 36  B-1050 Brussels
© 2008 CEN All rights of exploitation in any form and by any means reserved Ref. No. CEN/TR 15697:2008: E
worldwide for CEN national Members.

---------------------- Page: 2 ----------------------

CEN/TR 15697:2008 (E)
Contents Page
Foreword.3
Introduction .4
1 Sulfate resistant cements .6
2 Sulfate resistance test procedures.7
3 Review of most appropriate methods to assess specimen deterioration in laboratory tests.16
4 Review of most appropriate methods to accelerate the test procedure.18
5 The importance of test method reproducibility.20
6 Suggested features of a standardised sulfate resistance test method for cements.21
Bibliography .25

2

---------------------- Page: 3 ----------------------

CEN/TR 15697:2008 (E)
Foreword
This document (CEN/TR 15697:2008) has been prepared by Technical Committee CEN/TC 51 “Cement and
building limes”, the secretariat of which is held by NBN.
This CEN/TR is a state of the art review of the international research literature dealing with testing/assessing
the sulfate resistance performance of cements and related binders. It outlines the difficulties faced by
CEN/TC 51 in applying a prescriptive approach to the specification of sulfate resistant cements and identifies
the different mechanisms and forms of deterioration that occur during sulfate attack. This report compares the
advantages and disadvantages of different test specimen types (paste, mortar or concrete), different exposure
conditions and different techniques used to assess specimen deterioration. The importance of test method
reproducibility is reviewed with reference to the experimental work carried out by CEN/TC 51 during the 1990s.
The report lists the key parameters that must be controlled in any robust standardised method and makes
suggestions for the main features of a pan-European performance test.

3

---------------------- Page: 4 ----------------------

CEN/TR 15697:2008 (E)
Introduction
Under the terms of EU Mandate 114, committee CEN/TC 51, cement, building limes and other hydraulic
binders, is required to develop standards for ‘common cements’ and also for cements with special properties
such as low heat cements, calcium aluminate cements and sulfate resisting cements.
EN 197-1: Composition, specifications and conformity criteria for common cements was adopted in 2000 and
was the first harmonised European Standard to be adopted for a construction product.
Since 2000, European Standards for masonry, low heat, and low early strength blastfurnace cements, very
low heat special cements and calcium aluminate cements have been published. The development of a
prescriptive EN for sulfate resisting cements has been complicated by national differences in the types of
cement that are recognised to have sulfate resisting properties. Note, however, that all nationally
standardised sulfate resisting cements meet the requirements of EN 197-1:2000 and that the absence of a
specific standard for sulfate resisting cement has not constituted a barrier to trade.
In order to overcome these national difficulties, and also to permit new types of cement to be recognised in the
future, work was directed towards the development of a performance test for sulfate resistance. Work
commenced in 1991 and following a preliminary assessment of the French NF-P-18-837 procedure and the
German, so called flat prism method, a decision was taken to concentrate on developing the French
procedure. The method measures the expansion of 20 mm x 20 mm x 160 mm prisms in a sodium sulfate
2-
solution containing 16 g/l SO .
4
During five co-operative testing exercises involving up to thirteen laboratories, the method was refined with the
objective of improving reproducibility and also discrimination between sulfate resisting and non-sulfate
resisting cements. In 1998 it was concluded that further development would require a more fundamental
approach and efforts were directed towards obtaining EU funding for ‘pre-normative’ research. These
applications were not successful.
In early 2004 a meeting was arranged with representatives of the NANOCEM programme to explore a more
fundamental approach to the problem of sulfate resistance and sulfate resistance testing. The aspects of
particular interest to CEN/TC 51 were:
a) understanding sulfate attack mechanisms in relation to the type of cement and the
concentration/temperature conditions;
b) establishing a relationship between laboratory tests and field performance;
c) methods to accelerate the test;
d) using parameters other than deformation measurement to monitor the progress of the sulfate attack;
e) understanding the role of thaumasite in sulfate attack.
The NANOCEM group has formulated a research programme that addresses the above aspects and work on
this programme commenced in 2006 within the framework of a larger programme funded by the Marie Curie
Training Network. In parallel with this programme, CEN/TC 51 asked committee WG 12 (Additional
Performance Criteria) to prepare a CEN Technical Report outlining the current state of the art concerning
sulfate resistance testing.
A literature search identified over 250 relevant papers and reports published during the period 1970 to 2006.
To assess the different sulfate resistance techniques employed and their possible influence on the
4

---------------------- Page: 5 ----------------------

CEN/TR 15697:2008 (E)
performance of different cement/binder types, the testing details from 129 papers were entered into an Access
Database. The papers selected for entry into the database were those which contained original research data
and detailed information concerning test conditions.
This report draws on the information contained in these 129 papers plus a further 50 papers and reports not
selected for entry into the database. In the interests of brevity the current report only includes references to
selected references that are either key papers or contain specific information. It is intended that a statistical
analysis of the database and a full listing of the papers studied will be made available as a supplementary
document of CEN/TC 51 / WG 12.
5

---------------------- Page: 6 ----------------------

CEN/TR 15697:2008 (E)
1 Sulfate resistant cements
Portland cement concrete can undergo attack by sulfate bearing solutions such as natural groundwater or
those contaminated by industrial activity. Attack can result in expansion, strength loss, surface spalling and
ultimately disintegration.
The resistance that a cement matrix provides to sulfate attack depends on a number of factors which include:
• nature of the reaction products formed with the sulfate solution and in particular, whether their formation
results in disruptive expansion;
• impermeability of the matrix (including the important paste-aggregate interfacial zone) which provides a
barrier against penetration by sulfate ions;
2-
• concentration of sulfate ions (in this report expressed as g/l SO );
4
• mobility of the sulfate containing groundwater;
+ 2+ 2+
• nature of the accompanying cation e.g. Na , Mg , Ca etc;
• pH of the sulfate bearing ground water/solution;
• presence of other dissolved salts such as chlorides;
• temperature of the exposure;
• degree of pre-curing before exposure, although in the field this is only likely to affect the performance of
the concrete surface;
• presence of finely divided limestone (calcium carbonate) in the aggregate, or carbonate ions dissolved in
the groundwater, which may promote the formation of thaumasite under low temperature conditions.
Almost all developed countries have product specification standards for sulfate resisting cement(s). With a
few exceptions these are prescriptive standards that specify cement composition. The permitted compositions
are based upon long-standing laboratory test results and also satisfactory performance in the field. National
differences reflect different exposure conditions and also differences in the nature of the available cement
constituents.
Poor performance under sulfate exposure conditions is normally associated either directly or indirectly, with
the formation of ettringite. In the hydrated matrix of a CEM I cement, the source of reactive alumina is
normally the monosulfate phase according to the reaction:
2+ 2-
C Al (OH) .SO .6H O + 2Ca + 2SO + 20H O => C Al (OH) (SO ) .26H O
4 2 12 4 2 4 2 6 2 12 4 3 2
(monosulfate)     (ettringite)
Any unreacted C A is also a potential source of ettringite.
3
Monosulfate will normally also be present in composite cements containing blastfurnace slag, fly ash or
natural pozzolana but in the hydrated matrix of these cements, alumina is also present in phases such as
hydrotalcite or hydrogarnet or substituted in C-S-H in which latter forms. It does not appear to be available to
form an expansive reaction product [1].
Strength loss and disintegration are also associated with decalcification of C-S-H, which is an important
mechanism during attack by MgSO solutions but which also occurs to a lesser extent in Na SO solutions [2].
4 2 4
6

---------------------- Page: 7 ----------------------

CEN/TR 15697:2008 (E)
Current sulfate resisting cements standardised in CEN member countries can be divided into two categories:
1) Portland (CEM I) cements with a maximum permitted C A content.
3
2) Portland composite cements containing appropriate levels of glassy blastfurnace slag, fly ash or
natural pozzolana.
Low C A sulfate resistant cements provide a chemical resistance to sulfate attack. The products that reaction
3
with sulfates is not expansive and consequently the matrix is not disrupted facilitating further attack. The
unreactive nature of the hydration products of low C A cements is attributed to a low level of monosulfate
3
and/or the formation of an iron-rich form which is slow reacting and produces a ‘non-expansive’ form of
ettringite [3].
Portland composite cements (i.e. CEM II, III, IV and V types) provide resistance to sulfate attack which is
predominantly micro-structural in nature [4 to 8]. This is derived from the significantly lower permeability of the
hydrated matrix. Additional positive factors are:
• reduced level of free calcium hydroxide in the matrix which reduces calcium availability for ettringite
formation and also the formation of gypsum when the matrix is exposed to concentrated sulfate solutions;
• formation of hydrates containing alumina which are non-reactive to sulfate solutions.
The reduced availability of calcium may also result in the formation of ettringite with a morphology and
distribution throughout the hydrated matrix which is not expansive [9].
One factor that is often overlooked is that resistance to external sulfates is normally positively influenced by
the level of SO in the binder; the higher the level in a range between ~ 1 % to ~ 4 %, the greater the
3
resistance. This applies to concrete produced from CEM I cements [10] and also particularly to slag and fly
ash containing concretes. Where the ash or slag is added to the mixer [2], [11, 12] the SO level is lowered by
3
dilution and the hydrated matrix is more vulnerable to attack by penetrating sulfates in comparison with a
binder with an optimised SO level. The improved resistance can be attributed to the increased level of
3
sulfated phases, such as ettringite, formed during initial hydration, which are stable in the presence of an
elevated sulfate level.
2 Sulfate resistance test procedures
2.1 General review
The sulfate resistance properties of cement can be assessed by preparing realistic concretes specimens and
placing them in conditions which are representative of field conditions. Unfortunately, unless the concretes
are of low quality (high w/c, poorly compacted) several years exposure will be required to provide any
meaningful discrimination between sulfate resistant and non-sulfate resistant cements [13, 14]. Consequently,
there is a need for accelerated test procedures that provide discrimination within a timescale of weeks or
months.
The first laboratory test procedure to determine the sulfate resistance properties of a cement was the Le

Chatelier - Anstett procedure, [15] in which cement paste is hydrated, crushed and dried and then interground
with 50 % (by mass) of gypsum. The expansion of a moist cylinder, formed from the interground mixture is
determined at 1 day, 28 days and 90 days. The method is severe and cements with a low potential to form
expansive products such as calcium aluminate cement and supersulfated cement perform well, while low C A
3
sulfate resisting Portland cements perform poorly.
The first test procedure to attain the status of a national standard was the ASTM C 452 procedure, which was
adopted in 1964. In this test, cement is blended with finely divided gypsum to bring the SO level to 7,0 % and
3
0
the expansion of 25 mm x 25 mm x 285 mm mortar bars (1:2,75, w/c 0,485) placed in water at 23 C is
7

---------------------- Page: 8 ----------------------

CEN/TR 15697:2008 (E)
determined at 14 days. ASTM C150, Specification for Portland cement, permits a cement to be classed as a
Type V sulfate resistant cement if expansion at 14 days is less than 0,040 %. This is an optional requirement
and cement can be classified as Type V if the C A content is less than 5 % (and the sum of C AF + 2C A is
3 4 3
less than 25 %). The method is not suitable for cements containing constituents such as blastfurnace slag
and pozzolanas as firstly, the short timescale of test does not permit adequate hydration of the secondary
constituent and secondly the sulfate attack is ‘internal’ and does not take into account the reduced
permeability associated with constituents such as slag and pozzolanas.
Current thinking regarding accelerated test procedures [16, 17] is that the mechanism of deterioration in the
accelerated test should be representative of those observed in service. Neither the Le Chatelier- Anstett nor
the ASTM C 452 test procedures meet these criteria and they are not suitable for the assessment of cement
with secondary constituents.
Accelerated tests that provide a realistic mechanism of deterioration should take into account:
• resistance to penetration by sulfate solutions (impermeabilty) provided by the cementitious matrix;
• degree of curing that can be expected before the matrix is subjected to a critical level of attack, as
opposed to superficial surface damage; and
• sulfate environment, in terms of sulfate ion concentration, pH and temperature, which should not be too
far removed from conditions likely to be encountered in the field.
Accelerated sulfate resistance tests may be carried out using the following types of cementitious matrix:
• paste;
• mortar;
• concrete.
The advantages and disadvantages of these three test matrices are summarised in Table 1.
8

---------------------- Page: 9 ----------------------

CEN/TR 15697:2008 (E)

Table 1 - Advantages and disadvantages of different cementitious matrices for sulfate resistance
testing
Matrix Advantages Disadvantages
Paste Small specimen size reduces space Unless low w/c is used, or other measures
requirements. adopted, such as agitation during setting,
bleeding will result in heterogeneity.
Can be applied worldwide without need to
obtain suitable aggregate. A low w/c (< 0,35) is likely to result in an
extended testing time.
Test results are independent of aggregate
type Does not include the important aspect of
permeability of the paste aggregate interface.
Samples can be examined using
techniques such as chemical analysis, Water requirement of paste does not relate very
XRD, SEM etc without dilution by well to water demand in mortar or concrete.
aggregate.
Mortar
A well characterised single source of test May be criticised for ‘not being concrete’.
sand can be used by all laboratories within
a large geographic area.
Specimen size can be small thus
accelerating the test and reducing the size
of storage tanks required etc.
Most laboratories will have test equipment
suitable for the strength testing of mortar
specimens.
Mortar proportions can be adjusted and w/c
increased to accelerate the test without
unacceptable bleeding occurring.
Mortar may be gauged to constant
workability/flow.
Concrete Results may be more directly applicable to Large specimens are required in order to
field concrete. accommodate coarse aggregate.
Concrete may be gauged to constant Space requirements for specimens storage are
workability to take into account the water high.
demand characteristics of cement or
Duration of test will be long as a result of large
addition.
specimen cross section.
Impractical (and uneconomic) to use a
standardised aggregate thus reducing
reproducibility of test procedure.
Large capacity test machines required for
strength testing.

From Table 1, it can be seen that mortar testing offers practical advantages when seeking to develop a
standardised test procedure for sulfate resistance testing.
Specimens are normally fully immersed in the sulfate solution, but as discussed in section 2.3, procedures
have been developed that require specimens to be partially immersed in the solution. Supporters of this
approach believe that partial immersion better simulates the mechanism of deterioration which occurs in the
9

---------------------- Page: 10 ----------------------

CEN/TR 15697:2008 (E)
field. Section 2.4 briefly reviews procedures in which the specimens have been subjected to alternate wetting
and drying in order to accelerate the penetration of sulfate solutions into the specimen.
Several researchers [18 to 20] have found that initial carbonation of specimens markedly improves sulfate
resistance. It is likely that the inadvertent and variable carbonation of specimens is responsible, at least in
part, for the poor reproducibility of existing sulfate resisting test methods.
The following performance indicators are most commonly used to assess the sulfate resistance properties of a
test specimen:
 change in length (linear expansion);
 change in compressive strength;
 change in flexural strength;
 change in mass;
 change in appearance.
Less commonly, resistance has been monitored using non destructive techniques such as elastic dynamic
modulus and ultrasonic pulse velocity [21 to 24].
The advantages and disadvantages of the different performance indicators are discussed in section 3.
2.2 Review of test procedures in which specimens are fully immersed in sulfate solution
2.2.1 Mortar tests
Table 2 summarises the main characteristics of mortar test procedures which have either been standardised
(ASTM C1012, GOST 4798), have progressed to the status of a draft EN that describes the methods used
during the round robin activities of WG 12/TG1, or have been widely used, mainly in Germany (Wittekindt,
SVA and Koch Steinegger) [25].
The test procedures share the following characteristics:
 specimens have a high surface to volume ratio;
 with the exception of the GOST test procedure the use of highly concentrated Na SO solutions
2 4
2 2-
(16 g/l SO to 34 g/l SO ;
4 4
 replacement of the Na SO solution at monthly intervals (apart from the ASTM C 1012 timings which vary
2 4
according to age);
 use of a nationally (or European) standardised test sand;
 with the exception of the ASTM C 1012 procedure, assessment of sulfate resistance at an early age e.g.
56 days in the Wittekindt test;
 rather poor reproducibility.
The ASTM C1012 test procedure is the only internationally recognised and standardised test procedure. It
has undergone a number of modifications since it was introduced in 1984. The early version used a mixed
MgSO (4,3 %) and NaSO (2,5 %) solution. In 1987 this was replaced with a 5 % Na SO solution as it was
4 4 2 4
considered that the mixed solution gave confusing results. The procedure also differs in that the test bars are
° °
cured for 24 h at a temperature of 35 C and subsequently in limewater at 23 C. The bars are placed in the
10

---------------------- Page: 11 ----------------------

CEN/TR 15697:2008 (E)
5 % Na SO solution when a compressive strength of (20 ± 1) MPa is achieved by 50 mm mortar cubes cured
2 4
under the same conditions.
It should be noted that Poland incorporated the draft EN test procedure that describes the methods used
during the round robin activities of WG 12/TG1 into their specification for sulfate resisting cement PN-B-
19707:2003. Cements are classed as sulfate resisting if expansion is less than 0,5 % at 1 year.
Many researchers have chosen to use mortar test procedures, which do not follow those outlined in Table 2
but utilise ‘standard sized' specimens e.g. the 40 mm x 40 mm x 160 mm prisms used in the EN 196-1
strength test [24 to 28]. These prisms lend themselves to the determination of both flexural and compressive
strength and also to expansion if studs are cast into the ends or affixed with epoxy. The larger cross section
of the bars does delay deterioration compared to the smaller 20 mm x 20 mm or 10 mm x 40 mm section bars,
particularly if the w/c is relatively low.
In all the test procedures outlined in Table 1, the pH of the test solution will vary in a cyclical manner [5, 29].
Several researchers [30, 31, 8] have shown that if the pH of the solution is controlled by means of the addition
of dilute sulfuric acid then sulfate attack can be induced within a reasonable timescale without the need to
resort to high sulfate concentrations. The increase in pH occurs primarily as a result of the counter-diffusion
- 2-
of OH ions from the specimens necessary to maintain electroneutrality as SO ions diffuse inwards and react
4
with the matrix.
As well as accelerating the test it can be argued that controlling the pH also simulates more closely field
conditions where concrete is exposed to a mobile sulfate containing environment. However, it may not model
stagnant situations.
11

---------------------- Page: 12 ----------------------

CEN/TR 15697:2008 (E)

Table 2 - Comparison of sulfate resistance tests based on mortar
ASTM C1012: round robin round robin Wittekindt (VDZ SVA GOST 4798 Koch &
2004 modified) Steinegger
Test parameter Expansion Expansion Expansion Expansion Expansion Relativeflexural Relative flexural
strength strength
Specimen size 25 x 25 x 285 20 x 20 x 160 20 x 20 x 160 10 x 40 x1 60 10 x 40 x 160 10 x 10 x 30 10 x 10 x 60
(mm)
Surface : 0,17 0,21 0,21 0,26 0,26 0,47 0,43
volume ratio
2 3
(mm /mm )
Proportions 2,75 :1 : 0,485 3 : 1 : 0,50 3 : 1 : 0,50 3 : 1 : 0,60 3 : 1 : 0,50 1 : 3,5 : 0,40 1 : 3 : 0,6
Sand : cement : Blends with slag
water or pozzolana
gauged to same
flow as control PC
Sand ASTM C 109 test EN 196-1 EN 196-1 German test EN 196-1 Silica sand 0,4 - German test
type/source sand (Ottawa Standard sand Standard sand sand I and II Standard sand 0,5 mm sand I and II
silica sand) (fine and coarse)
Studs Stainless steel Stainless steel Stainless steel Stainless steel Stainless steel
or brass or brass
(projecting or (projecting or
recessed) recessed)
Compaction Tamping Jolting (10 jolts) Jolting (60 jolts) Vibration Vibration Jolting (20 jolts)
Curing 1 day in mould at 1 day in mould 1 day in mould, 1 day in mould 2 days in mould 2 days in humid 1 day in humid
°
35 C then in 27 days in water 27 days in 13 days in water 12 days in air, 13 days in air 20 days in
° ° ° °
water at 23 C at 20 C Ca(OH) at 20 C Ca(OH) water. water at 20 C
2 2
° °
until 50 mm cube solution at 20 C solution at 20 C
strength of 20
MPa
(continued)
12

---------------------- Page: 13 ----------------------

CEN/TR 15697:2008 (E)

Table 2 (concluded)
ASTM C1012: round robin round robin Wittekindt (VDZ SVA GOST 4798 Koch &
2004 modified) Steinegger

Solution Type Na SO Na SO Na SO Na SO Na SO (NH ) SO Na SO
2 4 2 4 2 4 2 4 2 4 4 2 4 2 4
Solution 33,8 16 ± 0,5 16 ± 0,5 29,8 29,8 1 and 2 29,8
concentration
2-
(g/l SO )
4
Solution 23 20 20 20 (and recently 20 20
°
temperature ( C) 5)
Specimen 461 to 358 408 408
surface :
solution volume
2
ratio (cm /l)
Frequency of 1, 2, 3, 4, 8 ,13, Every 28 days Every 28 days Every 28 days Every 28 days Every 30 days
solution 15 weeks, 4, 6, 9,
replacement 12 months
pH controlled No No No No No
Test limit High sulfate Not established Not established ≤ 0,5 % ≤ 0,5 % Flexural strength Flexural strength
resistance < expansion at 56 expansion at 91 > 80 % of control >70 % of control
0,10 % expansion days days at 1, 2, 4, 6 at 77 days
at 12 months months

Coefficient of 15 to 22 > 10 10 Satisfactory Satisfactory
repeatability (%)
Coefficient of 25 to 40 > 50 20 to 50 poor poor
reproducibility
(%)

13

---------------------- Page: 14 ----------------------

CEN/TR 15697:2008 (E)
2.2.2 Cement paste tests
Cement paste specimens have most frequently been selected by researchers who wish to examine the
specimens using techniques such as SEM and XRD [32 to 35].
A paste test procedure using 12,7 mm cubes and a pH controlled sulfate solution was first proposed by Mehta
and Gjorv in 1974 [36]. A relatively high w/c of 0,5 is used and bleeding is minimised by mixing the paste
using a high speed agitator. To promote the hydration of additions such as slag and fly ash, the pastes are
° 2-
cured at 50 C for 7 days before immersion in a Na SO solution with 27,1 g/l SO . In the initial work the pH
2 4 4
was controlled at 6,2 using 0,1N H SO but in more recent work [37] the procedure was modified to make it
2 4
°
less aggressive. The paste cubes are cured at 40 C for 14 days, the solution pH is neutral pH (7 ± 0,5) and
the sulfate concentration is reduced to 13,6 g/l. The pH adjustment may be achieved using automatic titration
equipment linked to a pH meter.
Sulfate resistance is determined by comparing the mean compressive strength of 10 paste cubes after 28
days immersion in the sulfate solution with that of 10 cubes broken at the end of the initial curing period. A
strength loss of 25 % is rated as satisfactory.
2-
Chemical analysis and X-ray diffraction of cement pastes tested using the Mehta procedure with a SO
4
concentration of 27,1 g/l, showed considerable gypsum formation but only traces of ettringite [38]. The
authors conclude that the test is based on the reaction between free lime and sodium sulfate. However, the
evidence available in the literature [9] indicates that ettringite causes the initial disruption and that this
facilitates further reactions, which can in some cases be more damaging. Ettringite is stable at a pH of 11,5,
unstable at pH 10 and highly unstable at a pH of 6 [29]. Thus, although ettringite is almost certainly formed at
the reaction front in the interior of the specimen, where the pH will be in excess of ~12, as the reaction front
progresses and the pH is reduced in this region (as a result of the controlled pH of the external solution) the
ettringite will decompose and the dominant reaction product will be gypsum. All published data from tests
using a controlled pH sulfate solution show the expected relationship between CEM I
...