III

III Performance of cement

Performance of cement 55

Light-optical microscopic recording of a polished section of Portland cement clinker

 VDZ Activity Report 2009 – 2012

Chemical and mineralogical The shift of the market towards cements sive use of these types of cements. Table
with several main constituents has result- III-2 shows the average chemical compo-
composition ed in a broad product range. Beside classi- sitions of Portland pozzolanic, Portland
cal cement types increasingly Portland limestone and Portland composite cements
The cements produced by the VDZ mem- composite cements are produced as CEM produced by the German VDZ members in
bers are investigated on their chemical com- II. Besides the dominant S-LL (blastfur- the reporting period.
positions by means of X-ray fluorescence nace slag and limestone) combination, main
analysis on a regular basis. Table III-1 lists constituents such as fly ash from hard coal Table III-3 summarises the chemical com-
the average values, as well as the highest or burnt oil shale are being increasingly positions of the Portland slag and blastfur-
and lowest values of the chemical compo- used. Due to the current standardisation nace cements (CEM II/A-S, CEM II/B-S,
sition for the various strength classes of the status, technical application approvals from CEM III/A and CEM III/B). Cement types
Portland cements according to DIN the German Institute for Building Technol- CEM III/C, CEM IV and CEM V are still
EN 197. ogy are still required for the comprehen- of minor importance in the German market.

Table III-1: Chemical composition of Portland cements, data in mass %, containing the loss on ignition
Portland cements
Strength 32,5 42,5 52,5
class min mean max min mean max min mean max
56
SiO2 19.04 20.73 22.92 18.64 20.63 23.18 18.96 20.86 23.53
Al2O3 3.50 4.82 5.83 2.86 4.73 5.77 3.38 4.53 5.68
TiO2 0.19 0.26 0.30 0.11 0.25 0.30 0.09 0.24 0.50
P 2O 5 0.04 0.19 0.60 0.03 0.19 0.54 0.02 0.19 0.45
Fe2O3 1.35 3.06 6.58 0.16 2.93 7.06 0.15 2.49 5.45
Mn2O3 0.04 0.07 0.10 0.02 0.07 0.19 0.02 0.06 0.19
CaO 60.41 63.20 65.91 59.72 63.56 66.37 61.41 64.44 67.82
MgO 0.83 1.61 3.39 0.59 1.52 3.80 0.57 1.31 3.27
SO3 2.34 2.99 3.77 1.96 3.03 3.79 2.22 3.20 3.98
K2O 0.53 0.92 1.44 0.35 0.80 1.45 0.25 0.73 1.40
Na2O 0.14 0.20 0.30 0.06 0.20 0.40 0.04 0.19 0.47
Na2Oequiv. 0.51 0.81 1.17 0.35 0.73 1.21 0.25 0.67 1.26
Loss on
0.78 1.97 3.08 0.49 2.09 4.06 0.28 1.84 3.75
ignition

Table III-2: Chemical composition of Portland pozzolanic, Portland limestone and Portland composite cements, data in mass %, containing the
loss on ignition
Portland pozzolanic, Portland limestone and Portland composite cements
Cement type CEM II/B-P CEM II/A-LL CEM II/A-M (S-LL) CEM II/B-M (S-LL) CEM II/B-M (V-LL)
min mean max min mean max. min mean max min mean max min mean max
SiO2 28.82 30.36 32.02 17.46 18.94 21.59 20.64 20.90 21.04 19.71 21.23 24.20 18.60 21.12 24.56
Al2O3 6.94 8.06 8.56 3.29 4.62 6.14 5.07 5.35 5.72 4.61 5.46 6.41 4.73 6.01 7.23
TiO2 0.33 0.40 0.46 0.18 0.24 0.41 0.25 0.27 0.28 0.26 0.32 0.44 0.20 0.28 0.33
P2O5 0.11 0.19 0.36 0.04 0.16 0.42 0.05 0.13 0.22 0.03 0.14 0.38 0.14 0.18 0.27
Fe2O3 3.25 3.58 3.93 1.21 2.51 3.76 2.17 2.43 2.70 1.62 2.32 3.02 2.54 3.15 4.26
Mn2O3 0.07 0.11 0.14 0.03 0.07 0.18 0.06 0.14 0.26 0.06 0.13 0.29 0.05 0.10 0.20
CaO 44.43 46.12 50.70 58.67 62.08 63.90 58.34 59.60 60.54 56.74 58.38 60.70 53.26 57.93 62.36
MgO 1.11 1.63 2.48 0.61 1.63 3.30 2.21 2.76 3.32 2.07 2.72 3.96 1.20 1.50 1.97
SO3 1.93 2.65 3.34 2.07 2.89 3.65 3.07 3.34 3.58 2.35 2.80 3.33 2.53 2.77 3.02
K2O 1.06 1.91 2.71 0.41 0.86 1.46 0.91 0.95 0.99 0.47 0.92 1.34 0.45 0.88 1.46
Na2O 0.53 0.84 1.20 0.08 0.17 0.32 0.19 0.30 0.37 0.13 0.21 0.32 0.16 0.21 0.28
Na2Oequiv. 1.23 2.10 2.98 0.35 0.73 1.14 0.83 0.92 0.97 0.44 0.82 1.08 0.48 0.79 1.24
S2-
n.a. n.a. n.a. 0 0.04 0.08 0.06 0.11 0.26 0.11 0.19 0.37 0.02 0.05 0.08
Loss on
3.35 4.15 5.11 3.16 5.89 7.87 3.27 3.79 4.11 2.93 5.26 7.24 5.32 5.87 6.35
ignition
n.a. = not analysed
 III Performance of cement

Table III-3: Chemical composition of Portland slag and blastfurnace cements, data in mass %, containing the loss on ignition
Portland slag and blastfurnace cements
Cement type CEM II/A-S CEM II/B-S CEM III/A CEM III/B
min mean max min mean max min mean max min mean max
SiO2 20.51 22.08 25.22 23.10 24.44 26.26 24.82 27.24 31.18 28.80 30.52 32.05
Al2O3 4.79 5.73 6.51 5.30 6.31 7.94 5.46 7.57 9.34 7.82 8.84 9.80
TiO2 0.26 0.35 0.53 0.31 0.43 0.60 0.33 0.61 1.10 0.41 0.78 1.25
P2O5 0.04 0.18 0.46 0.03 0.12 0.25 0.01 0.09 0.27 0.01 0.07 0.50
Fe2O3 1.25 2.50 3.56 1.19 2.19 3.33 0.83 1.62 2.72 0.65 1.14 1.50
Mn2O3 0.07 0.13 0.33 0.07 0.16 0.52 0.07 0.21 0.91 0.06 0.25 0.58
CaO 55.72 60.05 63.64 54.64 57.49 60.16 46.79 52.71 56.95 43.97 47.50 51.43
MgO 1.41 2.42 4.36 1.93 3.00 4.25 2.50 4.09 6.05 2.00 5.42 7.71
SO3 2.45 3.14 3.78 1.93 2.71 3.73 1.17 2.69 3.88 0.89 2.19 3.33
K2O 0.43 0.88 1.36 0.44 0.73 1.23 0.47 0.75 1.16 0.44 0.71 1.08
Na2O 0.11 0.23 0.42 0.15 0.24 0.39 0.14 0.25 0.44 0.19 0.28 0.49
57
Na2Oequiv. 0.42 0.81 1.22 0.47 0.72 1.03 0.50 0.74 1.09 0.51 0.74 0.99
S2- 0.07 0.20 0.41 0.13 0.35 0.61 0.28 0.59 0.99 0.48 0.85 1.12
Loss on
0.52 1.95 3.18 0.36 2.01 3.33 0.47 1.83 3.20 0.54 1.55 3.00
ignition

Quantitative phase analysis ured. The constantly repeating measure- land cements show that large ambiguities
ment and evaluation of raw materials, clink- still occur based just on the sample prepa-
of cements ers and cements allow changes in the phase ration. However, in contrast, analyses that
The quantitative X-ray diffraction analysis composition of the materials to be deter- were performed under identical prepara-
has continued to increase in importance in mined with high precision. In some plants, tion, measurement and evaluation condi-
the cement industry over the past years. the data of the X-ray phase analyses is now tions, e.g. production controls, provided
This method of investigation is now avail- no longer just used to monitor the on- more precise data.
able in most plants and is also being in- going production, but also for process con-
creasingly installed in laboratory automa- trol or for internal quality assurance in Analysis of composite cements
tion. The evaluation of the measured data some cases. Apart from the phase analysis of clinker or
is therefore no longer restricted to individ- Portland cement, the quantitative X-ray dif-
ual monitoring-specific parameters, such Determining the absolute phase content re- fraction analysis also principally enables
as the free lime content in clinker, but mains challenging. Interlaboratory tests to the description of the mixture composition
allows for a complete phase analysis. determine the accuracy overwhelmingly of composite cements. In order to achieve
provide unsatisfactory results. The German this objective, the already complex phase
Rietveld is the method of choice Chemical Society (GDCh) established a mixture of clinker and sulphate agents has
The Rietveld method, in which there is a task force to evaluate the Rietveld method to be extended by the additional main con-
comparison between the measured X-ray for X-ray diffraction measurements on Port- stituent components. It becomes particu-
diffraction pattern and that calculated from land cements and Portland cement clinker. larly challenging when X-ray amorphous
the peaks of predefined mineral phases, has The exchange of experiences between spe- (“glassy”) components are to be consid-
established itself as the method of choice. cialists from the industry, service laborato- ered, for example, for blastfurnace slag or
In addition, a multitude of measuring and ries as well as associations and universities fly ash. The modern evaluation software
structural parameters of the mineral phases is expected to result in the formulation of also provides solution possibilities in this
are mathematically included in the cal- a guideline to increase the interlaboratory case, e.g. via so-called pseudo-structures.
culations. This type of evaluation can still accuracy. Starting with the sample prepa- However, the results of these types of cal-
only provide approximate values of the ration, the instrument parameters, the phase culations must be carefully evaluated and
real composition, especially in complex identification and the selection of the struc- calibrated to test results from convention-
mineral mixes with various minerals, tural parameters, all important areas of a al analysis. The Research Institute of the
solids, degrees of crystallinity and textures. quantitative X-ray diffraction analysis are Cement Industry will comprehensively in-
addressed using the Rietveld method. For vestigate the current application possibili-
The instrument manufacturers provide soft- reasons of practical proximity, only evalu- ties of quantitative X-ray diffraction analy-
ware solutions that include evaluation rou- ation routines that require no manual post-pro- ses for quality control in cement produc-
tines that are individually tailored for the cessing will initially be considered. Initial tion as part of an AiF research project (IGF
plant and the material system to be meas- task force internal investigations on Port- no. 17397 N) that has just commenced.
 VDZ Activity Report 2009 – 2012

decomposed alite
free lime free lime
cluster
belite
belite
rim

alite

Fig. III-1: Alite crystals partially decomposed to belite and free lime Fig. III-2: Laboratory clinker nodule with dense belite rim, pro-
in the vicinity of a free lime cluster duced from raw meal nodules with Al(OH)3 rim (schematic repre-
sentation top left)

58

belite
cluster

central pore
belite
clusters

alite

Fig. III-3: Laboratory clinker nodule produced from a raw meal

nodule with Al(OH)3 particles (schematic representation top left) Fig. III-4: Belite cluster with central pore, detail from Fig. III-3
containing belite clusters with pores (arrows) in the dimension of
the Al(OH)3 particles

Impact of raw materials Raw materials with fuel properties mation of free lime clusters in the clinker.
The effects of a material, which was used Reducing conditions were present in the
and fuels on clinker direct environment of the coarse particles
as an alternative source of calcium and as
properties a fuel in the clinker burning process, the in the clinker nodules due to the fraction
microstructure of a technical clinker were of organic compounds, which were respon-
Every raw material and fuel has an impact examined. sible for the fuel properties of the mater-
on the clinker burning process and the clink- ial. The unusual microstructural relation-
er properties. As the performance of ce- The investigated clinker sample contained ship of decomposed alite and free lime clus-
ments is, to a large extent, determined by intergrown crystals of belite and free lime ters was the result of the specific proper-
the Portland cement clinker, the continu- (Fig. III-1) that presented decomposed ties of the combined raw material and
ous and detailed monitoring of the clinker alite crystal. This phenomenon is a typical fuel.
properties is state-of-the-art in modern ce- indicator for reducing burning conditions.
ment production. In recent years, the Re- However, the structure features unusually The importance of rapid clinker cooling
search Institute of the Cement Industry has only occurred in the direct vicinity of clus- with regard to the impacts of reducing con-
carried out a research project and various ters of free lime crystals. ditions on clinker was evident in the results
investigations on technical clinker in order of this investigation. Due to the available
to further improve the knowledge of the Due to their high calcium content, coarse process technology, the described clinker
impact of alternative raw materials and particles of the combined raw material and sample was cooled slowly. No signs of re-
fuels on Portland cement clinker. fuel in the clinker nodules led to the for- ducing conditions were observed in another
 III Performance of cement

Sintering temperature variation Burning temperature variation Precooling variation

100 100 100
1400 °C sintering temperature 60 min burning time Slow precooling
90 1450 °C sintering temperature 90 5 min burning time 90 Rapid precooling
80 80 80
Compressive strength in N/mm2

1500 °C sintering temperature

70 Burning time: 60 min 70 Sintering temperature: 1450 °C 70 Sintering temperature: 1450 °C
Precooling: slow Precooling: slow Burning time: 60 min
60 60 60
50 50 50
40 40 40
30 30 30
20 20 20
10 10 10
0 0 0
2 28 2 28 2 28
Age in days Age in days Age in days

Fig. III-5: Impact of the burning and cooling conditions of laboratory clinkers on the compressive strength of the laboratory cements
produced thereof

59

clinker sample that was cooled much more ules with Al(OH)3 rims or homogeneous- Changing temperature profiles
rapidly. ly distributed coarse Al(OH)3 particles. The How changes of the temperature profile and
belite clusters had diameters that were com- the burning temperature in the rotary kiln
The more rapid cooling prevented the de- parable to those of the added Al(OH)3 par- may affect the properties of Portland ce-
composition of alite or other effects caused ticles and generally had central pores. Com- ment clinker and the Portland cement pro-
by reducing burning conditions. parable clusters or rims of belite were not duced thereof was also investigated. Dif-
observed in the laboratory clinker made ferences in the burn-out behaviour of
Al-rich ash from the homogeneous raw meal. various fuels can change these parameters
The effects of artificial fuel ashes on the when using new materials, as the flame
microstructures of laboratory clinkers were During the burning of the nodules, calcium shape and temperature may be influenced.
investigated in an AiF research project (no. aluminates and, finally, clinker melt had
15251 N). One objective was to simulate formed from the enriched Al(OH)3 and cal- Laboratory clinkers of the same chemical
the inhomogeneous distribution of alumin- cium from the raw meal. The melt distrib- composition were exposed to different sin-
ium oxide in the kiln feed due to coarse uted in the clinker microstructure. Hollow tering temperatures (1 400 °C, 1 450 °C,
Al-rich ash particles or due to the accumula- spaces remained in place of the former 1 500 °C) in a laboratory kiln or were burnt
tion of Al-rich ash particles on the surface Al(OH)3 particles. However, a part of the for different lengths of time at the same
of clinker nodules. Al-rich ashes especial- calcium was also removed with the melt so maximum temperature (5 min, 60 min) for
ly are introduced into the clinker burning that only the low-calcium silicate phase, the the investigations. Furthermore, changes
process by certain alternative fuels. belite, but not the calcium-rich alite could in the length of the precooling zone in the
form in the direct vicinity of the Al(OH)3 rotary kiln, which extends from the hottest
Al(OH)3 particles with a few hundred µm rims and the coarse Al(OH)3 particles. point of the flame to the kiln outlet, were
diameter were added to a raw meal in simulated. In some cases the laboratory
order to simulate a coarse grained alumin- The extent to which this effect observed in clinkers were removed from the laborato-
ium-rich fuel ash. A layer of Al(OH)3 pow- the laboratory also occurs in technical clink- ry kiln at the burning temperature and cooled
der was applied to the surface of raw meal ers must still be investigated. However, the in air (rapid precooling). In other cases the
nodules in order to simulate the accumu- results indicate that belite clusters in clink- laboratory clinkers were left in the kiln
lation of aluminium-rich particles on the er are not exclusively the result of local after burning until this had cooled to
nodule surface. Raw meal nodules with a SiO2 enrichments. Mechanisms that lead 1 250 °C (approx. 30 min) and then removed
homogeneous composition were also pro- to local depletion of CaO can noticeably and cooled in air (slow precooling).
duced for comparison. The overall chemi- influence the clinker microstructure. X-ray
cal composition of all three nodule types diffraction investigations on the laborato- Laboratory cements were produced from
was identical. ry clinker showed no change in the miner- these laboratory clinkers by grinding them
alogical composition compared to the lab- to comparable fineness and mixing them
The aluminium-containing clinker phases oratory clinker produced from homogene- with a predefined quantity and composi-
C3A and C4AF were homogeneously dis- ous raw meal. This is not surprising as the tion of a sulphate agent. Small prisms (1.5
tributed in the microstructure of the clink- same total quantity of calcium aluminates cm x 1.5 cm x 6 cm) were produced from
er nodules produced from all nodule types. is expected for a constant total quantity of these cements on which the compressive
However, rims (Fig. III-2) and clusters of Al2O3. The properties of the resulting lab- strengths were determined after 2 and 28
belite (Figures III-3 and III-4) were ob- oratory cements were not affected. days.
served in the clinkers produced from nod-
 VDZ Activity Report 2009 – 2012

Despite the significant changes of the burn-

ing conditions that the laboratory clinkers HH AH
were exposed to, there were no significant 0 100
Initial setting time
differences in the 28-day compressive in min.
strengths of the resulting laboratory ce- 20 80
< 25
25 – 50
ments (Fig. III-5). 50 – 75
40 60 75 – 100
However, the 2-day compressive strengths 100 – 125
were noticeably lower for clinkers that 125 – 150 Fig. III-6:
60 40
Surface plot of
were burnt for long periods and/or at high 150 – 175
the initial set-
> 175
temperatures than for clinkers that were 80 20 ting time de-
burnt for short periods at intermediate to Test point of pending on
low temperatures (Fig. III-5). The precool- 100 0 the central the sulphate
ing variation did not have a significant ef- 2 3 4 5 6
composite design content and the
mix proportion
fect on the 2-day compressive strengths. Sulphate content, mass % of the sulphate
carrier
However, the extreme test conditions must
be considered when evaluating the prac-
tical relevance of the results. Fuel-related
fluctuations in the burning temperature or
60
sintering zone lengths are lower in practice
than in the tests in which extreme condi-
tions were supposed to be examined.

Cement sulphate
optimisation
The initial setting time of cements is de- sign of Experiments (DoE) enables the num- of the initial setting time depending on the
fined by the reactive cement components ber of necessary tests for sulphate carrier sulphate content and the mix proportion of
and their interactions. The optimisation of optimisation to be considerably reduced the anhydrite and hemihydrate sulphate car-
the sulphate content, in particular, with re- and also allows optimal mixes to be estab- rier. The Fig. indicates that the optimum
spect to the quantity and reactivity of the lished more quickly. A central composite area for both the supply quantity and the
tricalcium aluminate (C3A) present in the design (CCD) is created for this purpose ratios of sulphate carrier in the mixture can
clinker plays a decisive role in cement pro- using DoE, with which the correlation be clearly localised after only nine tests.
duction. Mortars and concretes can only be between the processability of the cement, Further tests can be used to investigate the
processed with a sufficient workability time its sulphate content and the sulphate car- setting time behaviour of the cement paste
with an optimal supply of sulphate. rier mix can be calculated. in this area of composition and to deter-
mine the optimal composition of the ce-
Sulphate carrier optimisation can be per- Approach ment with regard to the sulphate carrier
formed by measuring the initial setting time An orthogonal central composite design, quantity and composition.
of the cements produced in the laboratory. which considers the sulphate content in ce-
For this different sulphate quantities in dif- ment and the mix proportion of anhydrite/ This described process also enables the sul-
fering ratios of gypsum, anhydrite or hemi- hemihydrate, is suitable for sulphate car- phate carrier optimisation of cement with
hydrate are added to the clinker and the ini- rier optimisation. Cements consisting of regard to other parameters, such as com-
tial setting time of the sample is then de- Portland cement clinker, other main con- pressive strength according to DIN EN 196-1
termined with a penetrometer. Unlike the stituents (if applicable) and anhydrite (AH) or isothermal hydration heat release.
Vicat method according to DIN EN 196-3, and/or hemihydrate (HH) are then produced
in this case a needle with a diameter of according to the statistical design (see points
in Fig. III-6). The homogenisation of the
Influencing the effective-
3 mm is used, which enables a higher res-
olution of the setting times. In addition, samples in a ceramic ball mill has proven ness of chromate
considerably less cement is required for the itself to be an optimal solution in this re- reducers
penetrometer tests. gard.
Tin(II) sulphate is frequently used for the
Design of experiments The water demand of the cements to be op- chromate reduction of cement and cemen-
Sulphate carrier optimisations are general- timised must be determined prior to start- titious mixtures. However, in practice, it
ly expensive due to the high number of ne- ing the penetrometer measurements. This has been found that the content of water-sol-
cessary tests. The impact of mix ratio of an- takes place according to a method based uble chromate of cements with added tin(II)
hydrite to hemihydrate on the processabil- on DIN EN 196-3, which was esspecially sulphate will sometimes rise again consid-
ity of the cement for a predefined sulphate developed for determining the water re- erably, after only a few months. The chro-
content is generally investigated first for a quirement for small sample quantities. The mate reducing effect, surface changes and
sulphate optimisation. The next step then initial setting time of the cement paste is phase conversion reactions of various tin(II)
involves the determination of the impact then determined with the penetrometer and sulphate compounds were investigated over
of the sulphate content for fixed mix the test results are statistically evaluated. a period of nine months in the Research In-
ratios of the sulphate carrier. Using the De- For example, Fig. III-6 shows a surface plot stitute in order to clarify the cause.
 III Performance of cement

Long-term effectiveness
Mixtures of a cement CEM I 42,5 R with
about 14 ppm water-soluble chromate were
produced with various tin(II) sulphate com-
pounds for the investigations. The com-
pounds differed with regard to their con-
tent of residual moisture and residual sul-
phuric acid, the particle size and morph-
ology.

All of the investigated chromate reducers

showed a reduced effectiveness over the
nine months storage period, i.e. the meas-
Fig. III-7: Fine-
ured contents of water-soluble chromate grained tin(II)
increased over time. The effectiveness of sulphate
the coarse-grained tin(II) sulphate com-
pound decreased constantly over the stor-
age time. The fine-grained tin(II) sulphate
compound displayed a good reducing effi-
cacy over a period of six months. However,
61
its effectiveness subsequently sunk to a
level comparable with the other compounds.

Surface reactions
The behaviour of two industrially produced
tin(II) sulphate compounds were investi-
gated in more detail in order to clarify
whether chemical reactions were respon-
sible for the loss of effectiveness. The
fine-grained compound had a BET surface
Fig. III-8:
of about 1.2 m2/g, residual moisture of 0.4
Coarse-grained
mass % and residual sulphuric acid of 0.3 tin(II) sulphate
mass %. The particles were predominant-
ly rounded and flattened and formed ag-
glomerates with diameters of between about
5 and 30 µm (Fig. III-7). The coarse-grained
compound displayed a BET surface of
about 0.05 m2/g and contained less than
0.1 mass % residual moisture and less than
0.1 mass % residual sulphuric acid. The
particles were mostly euhedral acicular with
lengths of between 20 and 200 µm (Fig.
III-8).

Mixtures of Portland cement CEM I 42,5 R

Fig. III-9: Sur-
were produced with a tin(II) sulphate face of SnSO4
compound in proportions of 80 mass % ce- after 7 months
ment and 20 mass % SnSO4 for the series of storage in a
of investigations. Tin(II) sulphate particles mixture with
of both mixtures were investigated on a cement
monthly basis with the aid of X-ray diffrac-
tometry. Furthermore, the mixtures were
characterised with the aid of X-ray photo-
electron spectroscopy (XPS) after a stor-
age time of nine months. This method of the high tin(II) sulphate fraction was ex- solution to be inhibited when processing
surface analysis provided information on pected to help verify any phase conversions cement or for testing with regard to the con-
the elemental composition and the chemi- with the aid of X-ray diffractometry. tent of water-soluble chromate. Surprising-
cal bonding of the elements present. The ly, the conversion of the investigated coarse-
unusually high fraction of tin(II) sulphate The investigations showed that tin(II) sul- grained particles took place more quickly
in the mixtures ensured that a sufficient phate in cement is subject to chemical con- than the fine-grained material. The chem-
fraction of tin(II) sulphate particles could version processes. Thin layers of various ical reactions continued over the storage
be observed for the REM investigations tin(IV) compounds with low solubility ini- time until the tin(II) sulphate completely
and that sufficient cement was also avail- tially form on the surface of the SnSO4 converted to tin(IV) compounds.
able for possible interactions. In addition, particles (Fig. III-9). This allows their dis-
 VDZ Activity Report 2009 – 2012

The investigation results clearly show that this can then be provided as a reference ce- Hydration products
tin(II) sulphate is subject to chemical con- ment for chromate determination. The isolated investigation of reaction prod-
version processes in cement. Free lime, ucts of siliceous fly ashes formed in cemen-
moisture and the production-related morph- ASR test cement titious systems is practically impossible.
ology of the tin(II) sulphate particles in- Test cement with a high alkali content is Ashes were therefore mixed with various
fluence the speed of the relevant reactions used in order to test the alkali reactivity of calcium hydroxide contents and hydrated
and thereby the chromate reducing effec- aggregates according to Part 3 of the Alka- in order to obtain corresponding hydration
tiveness of tin(II) sulphate. In view of the li Guidelines. All the testing laboratories products of fly ash. This allowed the reac-
many variables, a forecast with regard to should use the same cement in order to re- tion of siliceous fly ashes to be adjusted.
the effectiveness of a specific chromate re- duce fluctuations between the various lab- The arising hydration products were inves-
ducer is virtually impossible. Determin- oratories during these determinations. The tigated using various methods. Ettringite
ations of the content of water-soluble chro- VDZ provides a uniform ASR test cement formed when the ash contained sufficient
mate are therefore essential. Compensat- with additional information on the cement amounts of sulphate. The fly ashes also
ing the expected loss of effectiveness by a properties and indicates the additional con- formed mono and hemi carboaluminate as
corresponding overdose of tin(II) sulphate siderations when using the cement. crystalline reaction products as well as strat-
has established itself in practice. This lingite to some extent. Increasing the water-
allows the chromate reduction to be en- Reactivity of main solid ratio increased the conversion of the
sured up to the specified expiration date pozzolanic reaction and more calcium
with proper storage.
constituents of cement
hydroxide was used.
62 In 2011 about 68 % of all cements produced
Reference material and in Germany had several main constituents. The reaction of the siliceous fly ashes in
cements In order to successfully continue this trend, fly ash-containing cements was also inves-
a more detailed understanding of the reac- tigated. The cements were produced in the
Reference material for determining tions of the cement constituents used in the laboratory by intensively mixing the rele-
chlorine hardened cement paste matrix is required, vant ash with a CEM I cement (cf. Table
The increasing diversity and use of second- in particular for the production of power- III-4). The formation of various carboalu-
ary fuels lead to higher requirements on the ful cements with high fractions of addition- minates was determined by using X-ray
analytics of these materials. The determi- al main constituents apart from clinker. Sev- diffraction analysis. The type (mono or
nation of chlorine is a central parameter, in eral research projects were involved in pro- hemi carboaluminate) depended on the
particular when burning plastics. A refer- ducing and improving the properties of ce- chemical composition of the fly ash. Fig.
ence material to determine the chlorine in ments with several main constituents in the III-10 shows the structure of a CEM II/
secondary fuels was provided based on an reporting period. B-V cement with 30 mass % siliceous fly
initiative by the VDZ Analytical Chemis- ash after a hydration period of 28 days,
try working group in cooperation with plas- Siliceous fly ash as a cement main magnified 5 000x. The reaction products of
tics producers. This material is a polymer constituent the fly ash are clearly visible. The forma-
that was homogenously dosed in defined The hydration reactions of eight different tion of strength-related products from the
quantities by the targeted amount of PVC. siliceous fly ashes were thoroughly inves- reaction of the fly ash was most intense be-
tigated in a research project supported by tween days 7 and 28 of testing. The micro-
An interlaboratory test was performed in the AiF (IGF no. 16206 N), which allowed structure of the hydrated samples became
the autumn of 2010 in cooperation with the conclusions to be drawn on their reactivi- looser as the ash content of the cement in-
Gütegemeinschaft Sekundärbrennstoffe ty. The chemical compositions, the densi- creased and there was a reduction in the
und Recyclingholz e.V. (BGS). This in- ty and the granulometric properties of the quantity of pozzolanic reacting fly ash par-
volved 43 test laboratories determining the fly ashes used are listed in Table III-4. ticles.
chlorine content and calorific value of the
untreated material. The average chlorine Suspension tests Conversion determinations and
content of the reference material was ap- The suspension of siliceous fly ashes and compressive strengths
prox. 0.8 mass % and the average calorif- of fly ash-calcium hydroxide mixtures in The conversion of the relevant siliceous fly
ic value was approx. 39,200 kJ/kg. The various test solutions allowed the corrosion ashes in cementitious systems was able to
material can be purchased from the VDZ. of the ashes to be accelerated and their hy- be determined by methanol-salicylic acid-
dration reactions to be investigated in quick hydrochloric acid (MSS-) digestion of
Reference material for determining time. More potassium than sodium was the hydrated hardened cement paste. The
chromate contained in the reaction products of the conversion of the ash in hardened cement
In the reporting period, the VDZ Analytic- fly ashes during suspension in an alkali test paste with 30, 50 or 70 mass % siliceous
al Chemistry working group also worked solution consisting of 0.54 mol/l KOH and fly ash was between 15 and 27 % after
on developing a reference material for de- 0.11 mol/l NaOH, which generally corres- 7 days of hydration and between 21 and
termining water-soluble chromate in ce- ponded to the pore solution of a CEM I 39 % after 28 days. The relevant ash con-
ments. A non-reduced CEM I cement free cement. The aluminium content in the version reduced as the ash content in
of minor additional constituents was select- eluate of the fly ash-calcium hydroxide cement increased. This is predominantly
ed for this purpose and analysed by the mixture was lower as the sulphate contained due to the reducing pH value of the pore
members of the working group. Its storage in the ash increased. The aluminium was solution and the lack of reaction partners
stability was also investigated. The chro- required for the formation of AFt and AFm in systems with high ash content and ex-
mate content of cement is expected to be phases, when sufficient sulphate was avail- plains the rising porosity of the hardened
determined in an interlaboratory test by the able, and was therefore no longer dissolved cement paste with increasing fly ash frac-
cement industry laboratories in 2012 so that in the eluate. tions.
 III Performance of cement

Table III-4: Chemical composition and granulometric parameters of the siliceous fly ashes and the Portland cement used for the production of fly
ash-containing cements
Components Unit A1 A2 A7 A8 A10 A11 A13 A14 CEM I
Loss on ignition 4.89 2.58 2.28 3.27 2.88 7.12 3.77 4.20 3.02
Silicon(IV) oxide 40.96 39.48 45.07 50.48 51.82 50.17 51.57 48.72 20.07
Aluminium oxide 26.56 22.56 28.31 30.25 24.86 24.12 26.83 28.90 4.34
Iron(III) oxide 9.98 14.27 12.13 6.04 7.30 6.22 6.65 4.66 2.54
Calcium oxide 7.73 11.28 2.14 0.75 5.74 3.31 3.62 7.47 64.02
Sodium oxide 1.48 1.04 1.40 1.13 0.66 1.03 0.85 0.17 0.18
Potassium oxide 1.79 2.28 3.25 3.92 1.97 3.74 3.07 0.60 1.15
mass %
Phosphorous(V) oxide 0.94 0.06 0.58 0.44 0.94 1.13 0.58 1.06 0.07
Magnesium oxide 3.82 3.02 2.58 1.92 1.93 1.76 1.65 1.45 1.44
Manganese(III) oxide 0.21 0.21 0.18 0.08 0.09 0.07 0.06 0.06 0.04
Titanium oxide 1.09 0.81 1.14 1.26 1.14 1.07 1.24 1.64 0.19
Sulphate a SO 3 0.65 2.09 0.94 0.40 0.67 0.22 0.34 0.26 3.66
Free lime 1.27 1.10 0.39 n.v. 0.82 0.14 0.22 0.89 n.v. 63
Reactive SiO2 33.8 33.8 36.5 44.3 39.0 44.1 40.6 34.1 n.a.
Density g/cm3 2.70 2.79 2.64 2.53 2.20 2.36 2.58 2.64 3.12
x’ (RRSB function) µm 27.27 32.04 25.46 13.20 32.42 26.01 27.79 36.39 12.41
n (RRSB function) – 0.77 0.87 0.80 1.10 0.90 0.70 0.80 0.88 0.87
n.v. = not verifiable
n.a. = not analysed

The impact of the chemical composition of

the fly ash on its strengthening contribu-
tion in cement increased as the fly ash con-
tent in the cement increased. The content
of reactive SiO2, Al2O3, Na2O and K2O of
the fly ash generally had a positive influ-
ence on the compressive strength of cement
containing fly ash. In contrast, reactive CaO
had a negative impact on compressive
strength (cf. Fig. III-11).

Hydration of siliceous fly ashes

with mineral additives
The addition of various mineral additives
to cement containing fly ash and the mix-
ing of various ashes had an, in part, strong
impact on the hydration of these systems.
The addition of 5 mass % sodium sulphate
had a positive effect on strength for cements
with low-sulphate ashes. The addition of
5 mass % of an aluminium-containing sub-
Fig. III-10: CEM II/B-V with 30 mass % siliceous fly ash A8 after 28 days of hydration with
stance (e.g. aluminium sulphate or meta-
a w/c value of 0.50; fracture surface magnified 5 000x
kaolin) was more effective for cements with
calcium-rich ashes than for those with alu-
minium and silicon-rich ashes. In addition,
it was determined that mixing equal parts
of calcium-rich ash (A2) with low-calcium
ash (A7) had a similar effect with regard
to strength as for the addition of additives Calcined clay as cement main of the AiF research project (IGF no. 16566)
(cf. Fig. III-12). constituent was the production of calcined clays, which
Calcined clays can be used as pozzolanic are suitable for use as a cement main con-
cement main constituents in accordance stituent. The performance of these clays as
with the requirements of DIN EN 197-1, if cement main constituents should also be
they have a minimum reactive silicon di- determined and evaluated.
oxide content of 25 mass %. The objective
 VDZ Activity Report 2009 – 2012

70 70
Reactive Si, mol Reactive Al, mol
Compressive strength in N/mm2

65 65

60 60

55 55
0.55 0.60 0.65 0.70 0.20 0.25 0.30 0.35 0.40

70 70
Reactive Na and K, mol Reactive Ca, mol
Compressive strength in N/mm2

65 65

64
60 60

55 55
0.03 0.06 0.09 0.12 0.05 0.10 0.15 0.20

Fig. III-11: Compressive strength of the CEM II/B-V cements with 30 mass % siliceous fly ash after 91 days depending on the chemical compo-
sition of the fly ash

Fifteen different clays from German clay

deposits were selected and comprehensive-
ly characterised for the research project.
60 The mineralogical composition of the clays
is indicated in Table III-5. Low quantities
50 of the clays were tempered under various
Compressive strength in N/mm2

burning conditions (variation of the burn-

40 ing period and temperature) and then chem-
5 % AI2(SO4)3 ically and mineralogically characterised.
5 % Na2SO4
30
5 % CaSO4
Larger quantities of clay were tempered
2 % CaCI2 after the optimal burning conditions had
20
5 % Metakaolin been determined for each clay. The sam-
5 % Blastfurnace slag ples were then ground to equal fineness and
10 Mix A2/A7 CEM II/A-Q cements, each with 20 mass %
A2-CEM IV/B
0 tempered clay, and CEM IV/B cements,
0 10 20 30 40 50 60 70 80 90 100 each with 40 mass % tempered clay, were
produced in the laboratory. A CEM I ce-
Time in days
ment, which met the requirements of a test
cement in accordance with DIN EN 450-1,
was used as clinker component.
Fig. III-12: Strength development according to DIN EN 196-1 of CEM IV cements with 50
mass % siliceous fly ash A2 with various mineral additives, as well as CEM IV cements with Calcined clay samples
25 mass % fly ash A2 and 25 mass % fly ash A7 The phase change of the clay samples dur-
ing the burning process was traced using
X-ray diffraction analysis. For example,
Fig. III-13 shows the phase change of a
kaolinitic clay sample. A burning temper-
ature of 1 000 °C and a burning duration of
30 min were determined as the optimal
burning conditions for this clay. Under these
burning conditions, the clay minerals kao-
lin and illite/muscovite contained in the
sample decomposed, while reactivity-re-
 III Performance of cement

Table III-5: Mineralogical phases of the clay samples used estimated using X-ray diffraction analysis, as well as information on the determined
burning conditions (temperature and burning period) and the related reactive silicon dioxide content, mass %

Clay 10

Clay 12

Clay 13

Clay 14

Clay 15
Clay 11
Clay 1

Clay 2

Clay 3

Clay 4

Clay 5

Clay 6

Clay 7

Clay 8

Clay 9
Quartz + T T T + + ++ + ++ ? +++ ++ ++ + +
Microline ? T – – – – T T – – T T – ? ?
Calcite T T T T – + – – – – – – – + –
Chlorite group T ? ? ? +++ ++ – – – – – – + – –
Kaolinite ++ + T T ? ? ++ ++ +++ ++++ ++ ++ ++ + +
Montmorillonite ++ +++ +++ +++ ? ? + ++ – – – – – + ++
Muscovite/Illite + ++ ++ ++ +++ ++ + T + + ++ + 2) +++ ++ ++
Goethite ? – – – – – + – – – – + – ++
Ankerite + T T – – + 1)
– – – – – – – ++ –
Siderite T – – – – – – – + ? T – + ? +
Gypsum T – – – – – – – – – – – – – –
65
Albite – – – – T – – – – – – – – – –
Lepidocrocite – – – – – – – – – – – – + – –
Pyrite T – – – – – – – – – – – – T –
Temp. in °C 650 600 900 750 850 1 000 1 200 1 200 600 1 000 1 200 1 200 950 600 950
Duration in min 30 30 30 30 30 30 30 5 30 30 5 5 5 30 5
Reactive SiO2 41.9 28.3 52.6 47.3 51.0 35.2 29.4 35.0 38.5 35.4 41.3 32.5 35.2 33.8 34.6
T = Traces 1)
Iron-rich ankerite
? = Not verifiable 2)
Poorly crystallised

ducing high-temperature phases such as

mullite and christobalite had not yet formed. 50000
Mu – Mullite Si – Sillimanite
The reactive silicon dioxide content of the 45000 Q– Quartz L– Limonite
I– Illite/Muscovite C– Christobalite
calcined clay was strongly dependent on 40000 K– Kaolinite H– Hematite
the composition of the respective clay as S– Siderite
well as the selected burning conditions. 35000
Table III-5 shows the burning conditions
Absolute intensity

30000 Mu
for each clay that led to high fractions of Mu C
Mu Mu H Mu Mu
H Si Si
reactive components. The corresponding 25000 1200–30
1200–5
reactive silicon dioxide contents are also 20000 1000–30
provided. With regard to the reactive com- 1000–5
ponents, it was generally noticeable that 15000 800–30
the content of reactive silicon-rich compo- 10000
650–30
600–30
nents increased as the burning temperatures 500–5
5000
increased, while the content of reactive alu- I K QS 500–30
K Q
minium-rich components reduced. Reac- 0
I KQ L S KQ K QQ Q Q QK S Q Q K Q Original
tive aluminium-rich components were more 10 20 30 40 50 60
likely to form at lower temperatures and Diffraction angle in ° 2 
reactive silicon-rich components at higher
temperatures. Fig. III-13: X-ray diffraction patterns of a kaolinitic clay tempered under various burning
conditions (burning conditions right in the Fig.: Temperature in °C – burning duration in
Cements containing clay minutes)
CEM II and CEM IV cements containing
calcined clays were produced with calcined
samples mentioned in Table III-5, which
contained about 28 to 51 mass % reactive
silicon dioxide. The cements were investi-
gated with regard to their performance based termined on mini-prisms, 15 mm x 15 mm The compressive strengths of the CEM II
on DIN EN 196. As expected, the water de- x 60 mm in size, with a w/c value of 0.5 cements after 2 days were between 24 and
mand of the cements increased as the frac- pursuant to the standard, while the CEM 29 N/mm2 and the standard strengths be-
tion of clay increased. The compressive IV cements had a w/c value of 0.60. The tween 48 and 66 N/mm2. Most of the ce-
strengths of the CEM II cements were de- results are displayed in Fig. III-14. ments could therefore be allocated to strength
 VDZ Activity Report 2009 – 2012

slag and siliceous fly ashes were investi-

gated with regard to their cement proper-
70 ties and the durability of the concretes pro-
60 duced with these cements. The investigat-
Fig. III-14:
ed combinations ranged from CEM II ce-
Compressive strength in N/mm2

Strength devel-
50
CEM II/A-Q opment accord- ments through to CEM IV and CEM V ce-
40 (w/c = 0.50) ing to DIN EN ments and even non-standardised cements.
196-1 of the The preparation of the test design and the
30 CEM II/A-Q evaluation of the results were performed
cements with
20 with the aid of statistical methods.
CEM IV/B 20 mass % cal-
10 (w/c = 0.60) cined clay (w/c
= 0.50) and of Raw materials
0 the CEM IV/B Two different Portland cements (Z) from
0 5 10 15 20 25 30 cements with strength class 42,5 R, which fulfilled the
Time in days 40 M.-% cal- requirements according to DIN EN 450-1,
cined clay (w/c
= 0.60)
were used to produce cements containing
blastfurnace slag and fly ash. Two blastfur-
nace slags (S) with differing activity indi-
ces (AI) were also used. Two different
grades of fineness (4 200 cm2/g and 5 500
66 S
cm2/g according to blaine) of blastfurnace
slag were used. Two types of siliceous fly
ashes (V) from German coal-fired power
stations, which had different chemical com-
positions, were also used. The chemical
CEM V composition and the granulometric param-
CEM III eters of the raw materials are provided in
Table III-6.

CEM II Parameter study

The Design of Experiments (DoE) was used
to prepare a test matrix with 108 cements
Fig. III-15:
Mixture design with 15 different compositions (cf. Fig. III-
of test 15) for which the 28-day compressive
(K = Portland strengths as well as the quantity of released
K cement; S =
V
hydration heat after 2 and 7 days were de-
Blastfurnace termined. The results were statistically evalu-
CEM I CEM IV slag; V = silice-
ous fly ashes)
ated to produce a model with the highest
possible regression for each parameter,
which only contained significant influenc-
ing parameters. The results of the determi-
nation of the 28-day compressive strength
are displayed in Fig. III-16. The cement
compositions, standardised according to
class 52,5 N. The compressive strengths of Ternary cements with fly DIN EN 197-1, are highlighted in grey in
CEM IV cements with a w/c value of 0.60 order to better classify the cement compo-
were between 9 and 13 N/mm2 after 2 days
ash and blastfurnace slag sitions. The Fig. indicates that it was pos-
and between 24 and 49 N/mm2 after 28 as main components sible to produce cements with a 28-day
days. compressive strength over 42.5 N/mm2 for
The production and use of cements with a a wide range of investigated cement com-
With regard to the strength development of clinker fraction under 65 mass % is becom- positions. The quality of the blastfurnace
the cements containing clay, higher reac- ing increasingly common due to the efforts slag had a considerable impact in this re-
tive silicon dioxide contents did not auto- of the cement industry to reduce CO2 emis- gard. The use of a particularly reactive blast-
matically lead to higher compressive sions per ton of cement. However, to date, furnace slag (left diagram in Fig. III-16)
strengths. Whereas the original clay min- only a small number of practical experi- enabled the use of up to 40 mass % of
erals present in the clay had a clear impact ences have been gained with cements with siliceous fly ash in the cement.
on the strength development of the cements. lower clinker fractions in Germany, apart
The strength of the tempered products in- from blastfurnace cements. The Research The lower the Portland cement clinker con-
creased as follows: Muscovite/Illite < Mont- Institute of the Cement Industry, together tent in the cement, the higher the impor-
morillonite < Kaolinite. with the FEhS – Institut für Baustoff- tance of the reactivity of the other main
Forschung e.V., carried out an AiF research constituents. In the event of unfavourable
project (IGF no. 16148 N) in which a large combinations (low clinker and blastfurnace
range of possible cement compositions in slag reactivity), it was not always possible
the ternary system of clinker, blastfurnace to achieve a 28-day compressive strength
 III Performance of cement

Table III-6: Chemical composition and granulometric parameters of the raw materials used
Components Unit Z1 Z2 HS 1 1) HS 2 1) FA 1 FA 3
IR in HCl/Na2CO3 0.63 0.41 0.06 0.22 a.b. a.b.
Loss on ignition at 950 °C 2.90 2.30 1.10 0.40 2.39 4.57
TOC mass % a.b. a.b. a.b. a.b. 1.57 4.68
CO2 2.25 1.42 0.32 < 0.20 0.06 0.05
Water 0.65 0.84 0.86 < 0.20 0.31 < 0.10
Silicon(IV) oxide 20.48 20.06 32.5 39.1 48.19 54.58
Aluminium oxide 3.99 4.38 14.0 11.8 23.22 26.15
Titanium oxide 0.21 0.20 0.81 0.47 0.94 1.31
Phosphorous(V) oxide 0.13 0.08 0.04 0.02 0.03 0.80
Iron(III) oxide 2.56 2.47 0.66 0.52 12.14 4.95
Manganese(III) oxide 0.06 0.04 0.32 1.80 0.13 0.07
Magnesium oxide 0.81 1.37 9.93 9.03 2.02 1.00
mass %
Calcium oxide 64.28 64.73 39.8 35.8 5.14 2.48
67
Sulphate as SO3 2.89 3.45 0.055 0.084 0.90 0.21
Potassium oxide 0.59 1.17 0.57 1.11 2.30 1.12
Potassium oxide 0.20 0.20 0.50 0.36 0.96 0.22
Sodium equivalent 0.59 0.97 0.87 1.08 2.47 0.95
Glass content a.b. a.b. 100.0 98.6 73.5 64.2
Reactive silicon dioxide a.b. a.b. a.b. a.b. 36.4 41.3
Particle density g/cm 3
3.142 3.116 2.917 2.923 2.675 2.543
Spec. surface acc. to Blaine cm /g
2
3 543 4 620 4 230 4 260 3 010 4 970
x’ (RRSB function) µm 20.86 12.41 15.17 15.69 33.13 23.32
n (RRSB function) – 0.85 0.87 0.79 0.73 1.00 0.84
n.a. = Not available
1)
Activity index after 28 d: HS 1 = 104; HS 2 = 88

Cements standardised Cements standardised

according to DIN EN 197-1 according to DIN EN 197-1
S S
Standard compressive strength Standard compressive strength
< 32,5 < 32,5
32,5–42,5 32,5–42,5
42,5–52,5 42,5–52,5
> 52,5 > 52,5

Cement Z 2 Cement Z 2
Blastfurnace slag HS 1 Blastfurnace slag HS 2

0 0 0 0

Z 0 V Z 0 V

Fig. III-16: Attainable 28-day compressive strengths of cements with clinker component Z 2 and blastfurnace slag HS 1 (left) and HS 2 (right)
 VDZ Activity Report 2009 – 2012

Table III-7: Composition of the cements selected to investigate sulphate resistance and for con- to those of concretes with cements approved
crete investigations, data in mass % for exposure class XF3 according to DIN
Cement Z1 Z2 HS 1 HS 2 FA 1 FA 3 1045-2. All of the concretes tested accord-
ing to the CDF method displayed scaling
SVA 1 25 – 5 – – 70
values of over 1.5 kg/m2. This was charac-
SVA 2 31 – – 17 52 – terised by a severe initial scaling during the
SVA 3 – 20 – 45 – 35 first freeze-thaw-cycles, which is primari-
ly due to the unfavourable preliminary stor-
SVA 4 30 – – – – 70
age period and conditions of the test spe-
SVA 5 – 36 – 12 52 – cimens (20 °C / 65 % rel. h.) for cements
SVA 6 31 – 52 – 17 – with longer hardening period.

SVA 7 56 – 27 – 17 – Ecological considerations

SVA-8 70 – – – 30 – The CO2 emissions for the possible com-
binations of Portland cement clinker, blast-
Concrete 1 – 20 80 – – –
furnace slag and siliceous fly ash could be
Concrete 2 – 60 – – 40 – estimated based on the boundary condi-
Concrete 3 – 30 35 – 35 – tions selected for Design of Experiments.
The following input variables were select-
Concrete 4 – 30 35 – 35 –
ed: for blastfurnace slag 0.1 t CO2/t blast-
68 Concrete 5 – 56 12 – 32 – furnace slag, for Portland cement 0.89 t
Concrete 6 – 56 36 – 12 – CO2/t cement and for siliceous fly ash
0.05 t CO2/t of fly ash. Fig. III-17 displays
Concrete 7 – 31 52 – 17 – the calculated CO2 savings potential for the
Concrete 8 – 40 45 – 15 – example of ternary cements with Z 1 as the
clinker component, HS 2 as the blastfur-
nace slag component and both of the inves-
tigated fly ashes. A standard compressive
strength of more than 42.5 N/mm2 was mod-
elled for cements consisting of the above-men-
tioned components to the left of the white
dotted line, while a standard compressive
strength of more than 52.5 N/mm2 was mod-
of at least 42.5 N/mm2 in the investigations Concrete investigations elled for cements left of the blue dotted
performed during the research project. Based on the results of the cement investi- line. The composition of standardised ce-
gations, eight cements were selected and ments is represented in grey and serves as
It must be taken into account that the in- two different concretes were produced us- an orientation. The calculation established
vestigated cements were produced by mix- ing each of the cements in order to perform that cements with standard compressive
ing of the three components Portland ce- concrete tests. The compositions of the ce- strengths over 42.5 N/mm2 and over 52.5
ment, blastfurnace slag meal and siliceous ments for the concrete investigations are N/mm2 could be produced with up to 77 %
fly ash. Strength optimisation aligned to listed in Table III-7. Concrete type 1 cor- and 55 % lower CO2 emissions respective-
the individual raw materials by adjusting responded to the provisions of the German ly compared to Portland cement. In par-
the meal fineness or by sulphate optimisa- Institute for Building Technology for ticular, the non-standardised area between
tion was not performed. It must therefore approval tests for binders to be approved the CEM III and CEM V cements was ex-
be assumed that cements with less than 40 for use in concrete in exposure class XF3 tremely promising, both in regard to the
mass % Portland cement clinker and with (z = 300 kg/m3; w/c = 0.60). Concrete type achievable compressive strengths as well
a mortar compressive strength of more than 2 was composed according to the minimum as the potential CO2 savings. However, the
42.5 N/mm2 after 28 days can be produced concrete composition requirements of DIN area between the CEM IV and CEM V/A
under practical conditions. 1045-2 for concretes in exposure class XF4 cements is not covered by the standard pro-
(z = 320 kg/m3; w/c = 0.50; LP = 4.5 vol. visions. But, the production of cements in
Further investigations suggest that market- %). The fresh concrete properties of the in- strength class 42,5 is also possible within
able cements can be produced in the inves- vestigated concretes were within the usu- this composition range.
tigated composition ranges with regard to al range for both concrete compositions
the early strength development and the ad- produced. With the exception of both of the
ditional requirements of DIN EN 197-1. investigated CEM V/B cements (20 mass Sulphate resistance at
% Z; 45 mass % S; 35 mass % V) all ce- magnesium exposure
Eight cements from the test program were ments in type 1 concretes resulted in 28-
investigated with regard to their sulphate day compressive strengths which allowed Unexpected damage was determined on
resistance according to the SVA method. the targeted realisation of concrete strength some originally sulphate-resistant concrete
The composition of these cements is listed class C25/30. and mortar test specimens during the la­b­
in Table III-7. The expansions of the mor- oratory and field investigations when
tar prisms were below the limit value of With regard to frost resistance, the weather- these were exposed to magnesium as well
0.500 mm/m. These cements therefore dis- ing damage of the concretes determined as sulphate in aggressive water.
played a high sulphate resistance. based on the cube test method was similar
 III Performance of cement

The damage occurred even though the mag-

nesium content was clearly below the limit
Cements standardised
value of 300 mg/l for exposure class XA1. according to DIN EN 197-1
Test specimens based on Portland and Port-
S
land limestone cement in combination with
fly ash from hard coal as concrete addition CO2 savings potential as
% of CEM I
were affected.
77–66 %
66–55 %
The material requirements and boundary
55–44 %
conditions under which a sulphate attack 44–33 %
of 1 500 mg sulphate/l and magnesium of
up to 300 mg/l may result in damage to Cement Z 1
0 0
mortars and concretes, if these have been Blastfurnace slag HS 1
produced with cement and fly ash from hard 28-day compressive strength
42.5 N/mm2
coal pursuant to the fly ash regulation ac- 52.5 N/mm2
cording to DIN 1045-2 have been investi-
gated in a research project (IGF no.
16710 N).

Five cements were tested (CEM I, CEM II/

69
A-LL, CEM II/B-S, CEM II/B-M (S-LL), Z 0 V
CEM III/A) in combination with two fly
ashes from hard coal as well as a CEM III/
B-SR and a CEM I-SR for reference. How- Fig. III-17: Calculated CO2 savings potential depending on the cement composition and the
ever the majority of the tests were per- modelled standard compressive strengths 42.5 and 52.5 N/mm2
formed with Portland and Portland lime-
stone cement. Initial damage in the form
of structure-deteriorating thaumasite for-
mation occurred after about 6 months for
test specimens based on Portland limestone
cement that were stored at a lower tempera­
ture (8 °C) and that were exposed to a mag- Nano-optimised binding measures (e.g. exposure classes XA3 and
nesium concentration of 300 mg/l. The test XM3).
specimens based on pure Portland cement
agents with high acid
showed comparable behaviour after about resistance The Research Institute of the Cement In-
12 months. An insufficient structure den- dustry predominantly concerned with the
sity at the beginning of the sulphate attack Experience with ultra-high performance con- structure optimisation of an ultra-high per-
was decisive with regard to the occurrence cretes has revealed that particle optimisa- formance concrete, whose composition
of the damage. Despite the rapid formation tion with fine particles can considerably im- (M3Q) is indicated in Table III-8. On the
of a thin brucite layer close to the surface, prove the structural density, strength and du- one hand, the structure density must be op-
sulphate was still able to penetrate the struc- rability of concretes. In mid-2009, a joint timised by minimising the packing densi-
ture to form secondary ettringite and thau- research project titled “Cold-curing ceram- ty and the porosity. On the other, a partial
masite in cases of the normal, short prelim- ics by nano-technological structure optimi- replacement of Portland cement with com-
inary storage time of 14 days during SVA sation” was commenced on this basis, sup- binations of blastfurnace slag and fly ash,
testing. All the tested cements remained ported by the Federal Ministry of Education or fine blastfurnace slag and fine fly ash, is
undamaged in cases of a longer prelim- and Research (BMBF). Besides the Research expected to increase acid resistance. The
inary storage period of the test specimens Institute of the Cement Industry, nine addi- required model calculations on the pack-
(90 days) or when stored at 20 °C. Both of tional partners are also involved in the pro- ing density and structure optimisation were
the reference cements CEM I-SR and ject. They represent the entire value chain, performed by a project partner, the Univer-
CEM III/B-SR did not show any weak­ from producers of individual raw materials sity of Kassel. The Fraunhofer Institute for
nesses under any of the test conditions, over formulators and producers of building Silicate Research had the task of develop-
which indicates that the presence of mag- products, through to consumers. ing silica with defined particle sizes of about
nesium (≤ 300 mg/l) does not pose any 100 to 800 nm for an extensive structural
additional attack potential for SR cements. Objective: High acid resistance density, which can be used as redispersible
An additional, more detailed report on this The new binder formulations should hard- microparticles.
research project is provided in Chapter V en at temperatures less than 100 °C with-
in the sulphate resistance section. out expensive autoclave treatment and Optimised concrete compositions
should enable the production of chemical- A specifically purchased high-performance
ly and mechanically highly resistant con- mixer enabled the production of homogen-
crete components and coating mortar. The ous, reproducible binder mixtures and con-
focus is on applications for which concrete cretes. A total of about 30 different com-
according to DIN EN 206-1/DIN 1045-2 positions with Portland cement fractions
has not previously been able to be used, or between 10 and 34 vol. % were produced
only with additional expensive protective in the Research Institute and investigated
 VDZ Activity Report 2009 – 2012

Table III-8: Selected concrete compositions together with their compressive strengths, porosities and acid resistance
Components/Parameters Unit Reference Optimisation Optimisation Optimisation Optimisation Optimisation
M3Q A B C D E

CEM I Vol. % 34 14 14 14 14 10

Blastfurnace slag 1/2 Vol. % – 3 / 10 2.3 / 7.5 2.4 / 8 8 / – 8 / –

Fine blastfurnace slag Vol. % – 7 5.3 5.6 3 5

Fly ash 2 Vol. % – – – – – 5

Silica Vol. % 10 10 12 12 14 13

Quartz meal Vol. % 10 10 9 8 9 1

Quartz sand Vol. % 46 46 50 50 52 58

Packing density, calc. Vol. % 81.4 82.4 84.0 85.1 85.7 86.1

7-day compressive strength N/mm² 136 127 121 122 122 98

28-day compressive strength N/mm² 167 159 145 146 141 127

Gel porosity, ≤ 0.01µm Vol. % 3.3 3.3 3.4 3.3 4.0 5.0
70
Capillary porosity Vol. % 1.7 0.8 0.9 1.3 1.4 1.6
> 0.01-100 µm

Acid resistance – 0 ++ ++ + + ++
Water binder value = 0.193; superplasticizer (PCE): 1.2 mass % of binder
(Binder: Mass of CEM I, blastfurnace slag, fine blastfurnace slag, fly ash and silica)

exceeded those of the reference composi-

1 tion. This was presumably due to the low-
0.9
est Portland cement fraction of 10 vol. %
CEM I of all of the investigated compositions. The
0.8
Blastfurnace slag 1 use of blastfurnace slag was more favour-
0.7 Blastfurnace slag 2 able than the use of fly ash for the targeted
Passage in mass %

0.6 Fine blastfurnace slag highest possible compressive strength. One

0.5 Fly ash 1
reason for this may be the differing par-
Fly ash 2
0.4 Fine fly ash ticle size distributions, which are represent-
0.3 ed in Fig. III-18. The two fly ashes were
0.2 clearly and the air classified fine fly ash
0.1 slightly coarser than the Portland cement
component. The blastfurnace slags barely
0
0.01 0.1 1 10 100 1000 differed from each other and differed only
slightly to the Portland cement. They were
Particle size in µm
therefore very well suited for substitution.
The fine blastfurnace slag was the finest
Fig. III-18: Particle size distribution of the blastfurnace slags, fly ashes and Portland cement
component by some distance and was there-
used in the concrete compositions
fore particularly well suited for packing
density optimisation.
with regard to their strength and porosity. For example, for concrete pursuant to op-
In this case the Portland cement fraction timisation A, when substituting about Selected samples were investigated by the
was predominantly substituted with blast- 60 % of the cement volume with blastfur- University of Kassel project partner for
furnace slag and fine blastfurnace slag in nace slag or fine blastfurnace slag com- their acid resistance according to the Ger-
order to achieve the highest possible acid pared to the reference mixture M3Q, the man sewerage construction guideline 2001.
resistance at pH1. capillary porosity was halved to 0.8 vol. % Based on the positive results, individual
and the acid resistance increased signifi- compositions were used by project partners
Table III-8 shows a selection of concrete cantly. However, as indicated in Table III-8, to produce concrete pipes with an internal
compositions, the correspondingly achieved the calculated packing densities did not cor- diameter of 20 cm and a length of about
compressive strengths, porosities and an relate with the obtained porosities and com- 1 m. Test specimens of these concrete pipes
estimate of the relevant acid resistance. The pressive strengths. The highest packing are currently also being tested for their
selection includes cement compositions for density was calculated for optimisation E. acid resistance. Furthermore, additional
which relatively high packing densities, However, the compressive strengths ob- investigations on the fresh concrete and
compressive strengths and acid resist- tained in the age of 7 and 28 days were durability properties (incl. sulphate resist-
ances were achieved. clearly under those of the reference com- ance, frost resistance) are in progress in the
position, while the porosities considerably Research Institute until Autumn 2012.
 III Performance of cement

Impact of additives on Table III-9: Chemical composition of the raw materials used: HS = Blastfurnace slag; KA = White
cement clinker; KB = Portland cement clinker; N = Limestone; A = Metakaolin; data in mass %
compressive strength and
Komponente HS 1 HS 2 KA KB N A
carbonation propensity of
blastfurnace cement SiO2 31.98 38.36 23.25 21.46 2.59 51.54

Al2O3 16.61 9.13 4.99 5.33 0.55 43.94

The impact of additives on the strength de-
velopment and the carbonation behaviour TiO2 0.95 0.97 0.15 0.28 0.03 1.41
of hardened cement paste, mortar and con-
cretes with blastfurnace cements (CEM P2O5 0.01 0.01 0.04 0.24 0.10 0.11
III/A) were investigated in a Federal Min- Fe2O3 0.24 1.76 0.15 3.04 0.20 0.85
istry of Education and Research project. In
particular, the impact of different additives Mn2O3 0.41 1.10 0.02 0.11 0.06 0.01
such as metakaolin, calcium hydroxide and
CaO 36.29 36.56 67.34 67.47 53.01 0.02
silica fume was investigated. (See also
Chapter V, “Carbonation of blastfurnace MgO 10.58 8.29 0.57 1.25 0.45 0.07
cement concretes”).
SO3 0.04 0.01 0.47 0.21 0.02 0.09
Impact of different additives
S2- 1.07 1.05 – – – –
The basic investigations were performed
71
with two blastfurnace slags (HS 1 and HS K2O 0.41 1.25 1.37 0.27 0.12 0.18
2). According to the chemical composition
in Table III-9, it can be assumed that blast- Na2O 0.38 0.41 0.12 0.11 0.01 0.06
furnace slag HS 1, with a higher Al2O3 con-
tent and lower SiO2 content, will provide
a higher strengthening contribution than
blastfurnace slag HS 2. The blastfurnace
slags were ground to a fineness of approx.
3 800 cm2/g according to Blaine. A white
cement clinker (KA) and a Portland cement were also determined on standard mortar The addition of 5 mass % calcium hydrox-
clinker (KB), both with a fineness of ap- prisms under standard conditions (20 °C / ide (C) had no significant impact on the
prox. 4 000 cm2/g, were selected as cement 65% rel. h.) at the age of 28, 42, 56, 84, compressive strength of blastfurnace ce-
clinker (see Table III-9). The low-iron white 119, 208 and 393 days pursuant to Issue ments HOZ 1 and HOZ 2.
cement clinker was selected in order to bet- 422 of the DAfStb.
ter investigate amorphous silicate hydro- The addition of 5 mass % silica fume (S)
gels and any arising amorphous alumo- Fig. III-19 to Fig. III-22 display the stand- effected an increase of the 2-day compres-
silicates using NMR spectroscopy. The alu- ard mortar compressive strengths and the sive strength of around 20 % and a slight-
minium-containing additive metakaolin (A) carbonation depths of the investigated blast- ly higher compressive strength at the age
was produced by tempering kaolin for two furnace cements. The blastfurnace cements of 28 (HOZ 2) and 91 days (HOZ 1 and 2).
hours at 700 °C. Silica fume (S) and cal- without limestone (N) displayed almost the
cium hydroxide (C) were also used as same compressive strengths as those with As shown in Fig. III-20, the addition of the
additives. 5 mass % limestone (N) at the same age aluminium-containing additive metakaolin
and are therefore not represented. (A), in particular for the high-performance
The investigated blastfurnace cements (HOZ cement HOZ 3, effected an increase in the
1 to HOZ 4) contained 35 mass % Portland Compressive strength compressive strength at the age of 7 to 91
cement clinker (KA or KB) and 65 mass % As shown in Figures III-19 and III-20, the days. The compressive strength only increased
blastfurnace slag (HS 1 or HS 2). A mixture blastfurnace cements with blastfurnace slag slightly at the age of 7 days for the low-per-
of gypsum and anhydrite at a ratio of 1:1 HS 2 with the lower aluminate and higher formance blastfurnace cement HOZ 4.
was used as setting regulator. The setting silicon content (HOZ 2 and 4) generally
regulator content was set at 3.5 mass % SO3 displayed lower compressive strengths than The addition of 5 mass % calcium hydrox-
for all the cements, with regard to the frac- the blastfurnace cements with blastfurnace ide (C) had no significant impact on the
tion of main constituents. The blastfurnace slag HS 1 with the higher aluminate and compressive strength of blastfurnace ce-
cements always contained 5 mass % of an lower silicon content (HOZ 1 and 3), re- ments HOZ 3 and HOZ 4, regardless of the
additive (metakaolin (A), silica fume (S) gardless of the clinker (KA or KB). blastfurnace slag.
and calcium hydroxide (C)). The reference
cements consisted of the respective clinker As indicated in Fig. III-19, the addition of The addition of 5 mass % silica fume (S)
and blastfurnace slag and contained 5 mass the aluminium-containing additive meta- effected a slight increase of the 91-day com-
% limestone (N) instead of the additive. kaolin (A), in particular for the low-perfor- pressive strength of blastfurnace cement
mance blastfurnace cement HOZ 2, led to HOZ 3 and an increase of the 28 and 91-
The compressive strength of the blastfur- an increase of the 2-day compressive strength day compressive strength of the blastfur-
nace cements produced in the laboratory of about 50 % and by about 20 % for the nace cement HOZ 4.
was determined on standard mortar prisms high-performance blastfurnace cement HOZ
pursuant to DIN EN 196-1 at the age of 2, 1. The compressive strength was also high-
7, 28 and 91 days. The carbonation depths er at the age of 7 days.
 VDZ Activity Report 2009 – 2012

Carbonation
80
2d 7d 28 d 91 d Apart from the generally higher compres-
HOZ 1: 35 mass % KA; 65 mass % HS 1 HOZ 2: 35 mass % KA; 65 mass % HS 2 sive strength, the blastfurnace cements
70
5 mass % additive 5 mass % additive with blastfurnace slag HS 1 (HOZ 1 and
Compressive strength in N/mm2

60 HOZ 3) also displayed a considerably high-

er carbonation resistance compared to the
50
blastfurnace cements with blastfurnace slag
40 HS 2 (HOZ 2 and 4) (see Fig.s III-21 and
30 III-22).
20 As indicated in Fig. III-21, the addition of
10 5 mass % metakaolin (A) led to a slight in-
crease in carbonation resistance up to 56
0
N A C S N A C S days for the blastfurnace cement HOZ 1
and up to 119 days for blastfurnace cement
HOZ 2.
Fig. III-19: Standard mortar compressive strengths pursuant to DIN EN 196-1 of the blast-
furnace cements from clinker KA with blastfurnace slag HS 1 and HS 2 and with 5 mass %
additive (N = Limestone, A = Metakaolin, C = Calcium hydroxide, S = Silica fume) The addition of 5 mass % calcium hydrox-
ide (C) effected an increase in the carbon-
ation resistance of blastfurnace cement HOZ
72
1 up to 208 days and up to 393 days for
2d 7d 28 d 91 d blastfurnace cement HOZ 2.
80
HOZ 3: 35 mass % KB; 65 mass % HS 1 HOZ 4: 35 mass % KB; 65 mass % HS 2
70 5 mass % additive 5 mass % additive
Compressive strength in N/mm2

60

50

40

30

20

10

0
N A C S N A C S

Fig. III-20: Standard mortar compressive strengths pursuant to DIN EN 196-1 of the blast-
furnace cements from clinker KB with blastfurnace slag HS 1 and HS 2 and with 5 mass %
additive (N = Limestone, A = Metakaolin, C = Calcium hydroxide, S = Silica fume)
 III Performance of cement

The addition of 5 mass % silica fume (S)

most effectively increased the carbonation 10
28 d 42 d 56 d 84 d 119 d 208 d 393 d
resistance of blastfurnace cement HOZ 2. HOZ 1: 35 mass % KA; 65 mass % HS 1 HOZ 2: 35 mass % KA; 65 mass % HS 2
9 5 mass % additive 5 mass % additive
In contrast, no significant differences were
8
detected in the carbonation resistance of

Carbonation depth in mm
blastfurnace cement HOZ 1. 7
6
Fig. III-22 indicates that the additives meta- 5
kaolin (A), calcium hydroxide (C) and 4
silica fume (S) always tendentially increased 3
the carbonation resistance of blastfurnace 2
cement HOZ 3. Depending on the storage 1
period, the addition of 5 mass % metakao-
0
lin (A) had no significant impact on the car- N A C S N A C S
bonation behaviour of blastfurnace cement
HOZ 4. Whereas a slight increased carbon-
Fig. III-21: Carbonation depths of the blastfurnace cements from clinker KA with blastfur-
ation resistance of samples produced with nace slag HS 1 and HS 2 and with 5 mass % additive (N = Limestone, A = Metakaolin, C =
HOZ 4 was detected on adding calcium hy- Calcium hydroxide, S = Silica fume) when stored under standard conditions
droxide (C) and silica fume (S).
73

28 d 42 d 56 d 84 d 119 d 208 d 393 d

10
HOZ 3: 35 mass % KB; 65 mass % HS 1 HOZ 4: 35 mass % KB; 65 mass % HS 2
9
5 mass % additive 5 mass % additive
8
Carbonation depth in mm

7
6
5
4
3
2
1
0
N A C S N A C S

Fig. III-22: Carbonation depths of the blastfurnace cements from clinker KB with blastfur-
nace slag HS 1 and HS 2 and with 5 mass % additive (N = Limestone, A = Metakaolin, C =
Calcium hydroxide, S = Silica fume) when stored under standard conditions