Aging behaviour of Alphabond and Calcium Aluminate Cement bonded castables

Dagmar Gierisch*, Andreas Buhr, Jerry Dutton

Almatis GmbH, Ludwigshafen, Germany

Margarete Hösler, Brunhilde Franz

RHI Urmitz AG & Co. KG, Urmitz, Germany

ABSTRACT addition of citric acid to increase setting time can

The aging behaviour of factory mixed and packed test considerably reduce the early strength of castables after a
castables is investigated by storing in a warehouse for 9 few days aging.
months and under severe conditions in a climate chamber for The remarkable increase in setting time during aging
4 weeks. Complete bags are also stored in the climate with citric acid as an additive has also been shown by Krebs
chamber. The castables are tested periodically for mixing [4], who investigated the aging of a medium cement castable
behaviour (wet out time), self-flow behaviour, setting start containing 70% alumina cement. The pot life of the castable
(by ultrasonic device) and strength. increases even after one day and shows clear differences
after eight days, e.g. from 5 to 45 minutes without citric acid
Introduction and 15 to 210 minutes with 0.02% citric acid.
The use of the hydratable alumina binder Alphabond 300 Parr et al. [5] investigated the aging of a Bauxite based
has increased dramatically over the past few years. It is used lcc with 70% alumina cement by single component aging
in both cement free castables and in combination with high and entire castable aging. Due to the complex and interfering
purity calcium aluminate cement for low (lcc) or ultra low aging reactions, the latter is recommended to achieve
(ulcc) cement castables. In comparison to calcium aluminate representative results. The samples have been stored in
cements, Alphabond 300 is more hygroscopic and must be layers of 25 mm thickness in a climate chamber for up to 60
handled with care. When not in use, the product should be days. The initial flow and the setting time have been
stored in sealed packaging in cool, dry conditions. checked, but not the flow at 30 and 60 minutes. The authors
In the 1990´s extended tests were carried out to distinguish two aging periods - up to 7 days and after 7 days.
investigate the shelf life of Almatis calcium aluminate In the first period, both flow and setting vary but stabilise
cements and Alphabond 300 including Brazilian warehouse afterwards. The flow at first increases and later shows a
storage and jungle room conditions in a climate chamber for gradual deterioration. The working time increases, and the
more severe conditions. Following these tests, a special changes are more marked than the initial flow. After 30 days
packaging has been introduced for Alphabond 300. Based at 20°C and 70% relative humidity, the castable becomes
upon the results of these tests, a shelf life of 12 months is unsuitable for use.
stated in the Almatis datasheets for both calcium aluminate Sugimoto et al. [6] report differences in aging behaviour
cements and Alphabond [1,2], if the material is stored under of sodium metaphosphate (SMP) and sodium polyphosphate
adequate dry conditions. (SPP), both of which are used as dispersing agents in
Dispersion and setting control additives are another chamotte based low and ultra low cement castables with
component of modern castables which can be susceptible to 70% alumina cement, calcined alumina and silica fume. The
aging. The shelf life of the Almatis dispersing aluminas test castables have been lab mixed and stored in a climate
ADS/W and M-ADS/W series has also been tested and a chamber for up to 60 days. The hardening time was
shelf life of 12 months is satisfactory if stored under evaluated but not the flow behaviour. SMP retards the
adequate dry conditions. hardening by aging, while SPP remains more stable. The
However, it is well known in the industry, that castables increase of hardening time during aging becomes more
do show aging during storage which can be more or less pronounced with higher silica fume contents and lower
pronounced depending on the storage conditions and the cement contents (lcc vs. ulcc). During aging the chains of the
castable composition. In the case of Alphabond 300 SMP are broken up by hydrolysis, which reduces the pH and
containing castables limited experience was available with hampers the cement reaction.
regard to aging. This is of particular interest for castables This study will present the results of factory mixed and
which are to be used for on site installations, and which may packed castables, which have been exposed to aging by
be transported and stored for an extended period of time storage under dry conditions in a warehouse and under
before finally being used for the installation. Therefore this defined conditions in a climate chamber, but always as
study has been performed and the aging behaviour of complete bags as initially packed.
different castables has been investigated.
The impact of artificial aging of a 70% alumina cement Test castables
lab sample on the flow and setting behaviour of EU standard Four different self-flowing castables based on tabular
mortars and a silica fume containing low cement castable alumina have been tested for their aging behaviour (table 1):
(lcc) is reported by Mathieu et al. [3]. For the mortar, the - LCC-PACA: a low cement castable with
flow at 30 and 60 minutes and the setting have already polyacrylate/citric acid as additives,
increased after the first days aging of the cement. For the lcc, - LCC-ADS/W: with dispersing aluminas ADS 3 and
less additives are required for a pre-aged cement when ADW 1 as additives,
compared to fresh cement. If the additive package is adjusted - ULCC-M-ADS/W: an ultra low cement castable with
based on fresh cement, it might be too much already for dispersing aluminas M-ADS 1 and M-ADW 1 as
cement, which has aged for just a few days. For example the additives,
 Tab. 1. Composition and properties of self flowing test castables at start status
Castable SFL SFL SFL SFL
ULCC-M-
Component LCC-PACA LCC-ADS/W NCC-AB
ADS/W
Coarse fraction
Tabular T60/T64 % 70 70 70 70
(up to 6 mm)
T60/T64 -45 MY Li % 8 8 14 10
Fine fraction
T60/T64 -20 MY % 7 7 7
CTC 22 % 10 10 10
Reactive Alumina
CTC 20 % 7
Silica fume Elkem 971U % 5
CA-14 M % 5 5 3
Binder
Alphabond 300 % 1 3
Citric Acid % 0,05
Polyacrylate % 0,05
ADS 3 % 0,40 0,75
Additives
ADW 1 % 0,60 0,25
M-ADS 1 % 0,80
M-ADW 1 % 0,20
Water % 5,4 4,9 4,8 4,9
wet mixing time min 4 4 4 6

wet out time sec 15 50 25 50

10 min mm 272 286 238 270
self flow 30 min mm 242 288 248 250
60 min mm no flow 285 244 220
ultrasonic setting start min 52 62 134 206
20°C / 24h MPa 1 4 4 1
110°C / 24h MPa 7 12 15 8
400°C / 5h MPa 3 9 14 7
CMoR 800°C / 5h MPa 5 8 16 3
1000°C / 5h MPa 3 6 34 1
1500°C / 5h MPa 18 32 53 29
1650°C / 5h MPa 26 43 63 39
20°C / 24h MPa 4 17 15 3
110°C / 24h MPa 36 60 93 27
400°C / 5h MPa 38 66 108 44
CCS 800°C / 5h MPa 37 71 77 30
1000°C / 5h MPa 24 27 145 6
1500°C / 5h MPa 93 116 143 83
1650°C / 5h MPa 128 165 166 102

- NCC-AB: a no cement castable with Alphabond 300 tent. Storing under dry conditions was guaranteed, but
and ADS 3 and ADW 1 as additives. castables were exposed to seasonal temperature variations.
Before testing the material, the particular bags were stored in
The castable matrix contains T60/T64 -45 µm LI and/or the air conditioned lab at 20°C for temperature
-20 µm and the reactive aluminas CTC 22 or CTC 20 in acclimatisation. Within one week the castables were tested
combination with silica fume. All formulations were in the Ludwigshafen lab for wet out time, flow behaviour up
adjusted to give satisfactory flow values at 10 and 30 to 60 minutes, ultrasonic setting start and strength after
minutes and a start of setting below 240 minutes. various pre-firing temperatures (table 1).
The castable was wet mixed in batches of 5 kg using a
Test conditions Hobart A 200 planetary mixer at speed 1 for 4 minutes
500 kg of each test castable were dry mixed and packed (LCC`s and ULCC) and 6 minutes (NCC). The wet out time
in 25 kg valve-sealed bags (multi-layer: paper-plastic foil- was determined as described in a previous paper [7]. The
paper) at RHI Urmitz. The fresh material was shipped to the flow properties after 10, 30 and 60 minutes (F10, F30, F60)
Almatis plant in Ludwigshafen and stored in a warehouse were measured by the cone test (lower diameter 100 mm,
upper diameter 70 mm, height 80 mm). The setting shown in table 2. A new unopened bag was always used for
behaviour was measured by ultrasonic equipment. The eight each test. The castable properties were tested at a constant
channel device records the velocity of ultrasound through the water demand. The test series will continue until the
sample during the curing time by means of a transmitter and castables become unusable or the test material is consumed.
a receiver (figure 1). The hardening of the test piece gives an Also a series of tests were performed in a climate chamber
increase in velocity and the initial setting coincides with the (HEKK 2057 S from Weiss Enet B.V./NL, volume 500 l)
first rise in the velocity vs. curing time graph (figure 2). using complete unopened bags. Here different settings were
An overview of storage conditions and test intervals are chosen to compare with the warehouse aging trend.

Tab. 2. Test conditions during storage in warehouse and climate chamber and test intervals
warehouse climatic cabinet

changing conditions 75% / 35°C 85% / 35°C 95% / 35°C

climatic cabinet climatic cabinet climatic cabinet
storage warehouse tent, dry
(unopened bag) (unopened bag) (unopened bag)
rel. humidity rel. humidity rel. humidity
seasonal
75% 85% 95%
treatment temperature
temperature temperature temperature
variations
35°C 35°C 35°C
start status start status start status start status
1 month 1 week 1 week 1 week
3 months 2 weeks 2 weeks 2 weeks
test intervals 5 months 4 weeks 4 weeks 4 weeks
7 months
9 months

4000
Start warehouse
1 month warehouse
5 months warehouse
9 months warehouse
3000
velocity [m/s]

2000

1000

0
0 100 200 300 400 500 600 700
time [min]

Fig. 2a. Ultrasonic graphs of LCC-PACA / warehouse

Fig. 1. Eight channel ultrasonic device - sample carrier with
storage
transmitter and receiver
 5000 of 15, which was achieved initially. The same trend towards
longer wet out after aging occurs in the climate chamber
particularly at 95% rel. humidity. LCC-PACA shows no
4000 flow at 60 minutes when fresh material is used. The flow at
10 and 30 minutes remains relatively stable over a period of
5 to 7 months, although a slight decrease after 7 months
velocity [m/s]

3000
storage shows a deterioration of flow which results in no
flow after 9 months. The castable is still usable but only at a
2000
0.4% higher water demand. The decrease in flow
corresponds to the results obtained from test castable stored
under extreme conditions in the climate chamber, where the
1000 Start warehouse castable is no longer usable after 2 weeks at 85% or 95% rel.
1 month warehouse humidity.
5 months warehouse
9 months warehouse
A slight decrease of setting start is observed. The
0 corresponding ultrasonic graphs are shown in figure 2a.
0 100 200 300 400 500 600 700 Here, the warehouse results do not correlate with the climate
time [min]
chamber, which shows the reverse trend.
Fig. 2b. Ultrasonic graphs of LCC-ADS/W / warehouse Strength properties remain stable over the whole period.
storage Even with the increased water demand the strength values
are still at the same level. At 75% rel. humidity and 35°C no
2000 change in strength is observed over a period of 4 weeks, but
Start warehouse at 85% rel. humidity the strength is reduced by about 50%
1 month warehouse after 2 weeks storage. For both, 85 and 95% relative
5 months warehouse
9 months warehouse humidity, the castable is unusable after 2 weeks.
1500
LCC-ADS/W requires less water when compared to
LCC-PACA (4.9 v.s 5.4%) and achieves an even better flow.
velocity [m/s]

It shows a very stable performance over 9 months warehouse

1000
storage. Wet out time for the warehouse material is about 50
seconds whereas in the climate chamber at 85% rel.
humidity a rise in wet out of up to 100 seconds is observed.
500 The only exception to the extremely good flow of about 290
mm is the F60 value (no flow) at 7 months. This can be
explained by the setting start, which in this case was
0 adjusted to about 60 minutes. By contrast aging in the
0 100 200 300 400 500 600 700 climate chamber at 85% rel. humidity reduces the flow (F30)
time [min] down to no flow after 4 weeks.
Fig. 2c. Ultrasonic graphs of NCC-AB / warehouse storage The setting start at about 60 minutes in the warehouse
aging remains very stable up to 9 months (figure 2b). The
Results and discussion climate chamber (85 and 95% rel. humidity) shows a trend
The results of the warehouse aging up to 9 months are towards shorter setting times during aging, which has not yet
given in table 3 and the impact of aging on the flow been observed in the warehouse testing. With regard to
behaviour (F30) is represented in figure 3. strength no change for both warehouse and climate chamber
The test castables can show some variation in results occurs.
during the testing periods. These are not considered to be ULCC-M-ADS/W shows aging stability, if the above
due to aging unless the specific values are confirmed by later mentioned variations of the test results are not attributed to
results (e.g. from tests performed 2 months later). For aging reactions. The wet out shows a slight increase from 25
example, low flow values of ULCC-M-ADS/W were noted to 50 seconds after 9 months, which is still fast. The flow
after one month, but in the later tests a reasonable flow varies at different test periods, but no trend attributable to
behaviour was once again achieved. One reason for this aging is shown. The same applies for the setting behaviour,
observation may be a slight bag to bag variation, which may where the setting start varies between 100 and 200 minutes.
be more pronounced for the ULCC in comparison to other In the climate chamber at 95% rel. humidity the flow
castables due to the very low cement content. decreases (no F30 after 1 week) and the setting gets shorter.
Also it has been accepted that there can be some The castable becomes unusable after 4 weeks at 85% and
variation of the ultrasonic setting start without attributing 95% rel. humidity.
that as a result of aging unless a trend becomes clear. NCC-AB achieves a wet out time between 50 and 80
However, these variations are not seen as technically critical seconds over the whole test period up to 9 months. In the
unless the workability period for the installation of the climate chamber the wet out remains stable until the castable
castable would become too short or the setting of the becomes unusable after 4 weeks at both, 85 and 95% rel.
castable delayed too much. Only LCC-ADS/W shows a very humidity. The flow shows a slight increase after 1 month but
stable setting start at about 60 +/- 10 minutes. afterwards remains stable at a high level of about 300 mm.
LCC-PACA has a water demand of 5.4% and shows a More severe climate chamber conditions of 85% rel.
very fast wet out at the start. After 9 month warehouse humidity and upwards show a trend towards a decrease in
storage the wet out is achieved at about 90 seconds instead flow, which is more pronounced at 95% rel. humidity. This
probably indicates an aging behaviour which may occur in after drying is comparable to LCC-PACA, but lower when
the warehouse at a later stage. compared to LCC-ADS/W and ULCC-M-ADS/W. The
The setting time of Alphabond bonded castables can be setting start remains stable within a range of 200 to 350
recorded by the ultrasonic method as shown in figure 2c. minutes and shows no clear trend in the warehouse aging. In
This would not be possible with the EXO method [7]. the climate chamber it shows no change at 75% rel.
Compared to both LCC´s (figure 2 a and b) NCC has not humidity, but at 85% at first an increase is seen, followed by
reached the plateau of stable velocity after 700 min, and it a decrease, and at 95% a decrease is observed. Once again,
still keeps rising at 1200 minutes, where the test ends. an aging trend may be indicated, which has not yet been
However, the strength achieved after 24 hours curing and shown in the warehouse aging.

Tab. 3. Castable properties from start status up to 9 months warehouse storage

LCC-PACA Start 1 month 3 months 5 months 7 months 9 months
water demand % 5,4 5,4 5,4 5,4 5,4 5,4 / 5,8
wet out time sec 15 40 40 15 60 90 / 90
no flow /
10 min mm 272 255 271 270 228
218
no flow /
self flow 30 min mm 242 228 253 233 198
211
no flow /
60 min mm no flow no flow no flow no flow no flow
175
ultrasonic setting start min 52 96 106 72 164 102 / n.d.
20°C / 24h MPa 1 n.d. 1 n.d. 1 n.d.
CMoR 110°C / 24h MPa 7 n.d. 8 n.d. 8 n.d.
800°C / 5h MPa 5 n.d. 7 n.d. 6 n.d.
LCC-ADS/W
water demand % 4,9 4,9 4,9 4,9 4,9 4,9
wet out time sec 50 60 60 30 40 40
10 min mm 286 307 305 300 299 296
self flow 30 min mm 288 300 293 291 292 282
60 min mm 285 298 260 272 no flow 253
ultrasonic setting start min 62 62 54 58 64 62
20°C / 24h MPa 4 n.d. 5 n.d. 4 n.d.
CMoR 110°C / 24h MPa 12 n.d. 10 n.d. 16 n.d.
800°C / 5h MPa 8 n.d. 12 n.d. 11 n.d.
ULCC-M-ADS/W
water demand % 4,8 4,8 4,8 4,8 4,8 4,8
wet out time sec 25 25 30 35 70 45
10 min mm 238 170 185 204 203 211
self flow 30 min mm 248 146 183 217 204 217
60 min mm 244 155 183 206 211 208
ultrasonic setting start min 134 92 174 130 198 124
20°C / 24h MPa 4 n.d. 4 n.d. 4 n.d.
CMoR 110°C / 24h MPa 15 n.d. 13 n.d. 13 n.d.
800°C / 5h MPa 16 n.d. 18 n.d. 20 n.d.
NCC-AB
water demand % 4,9 4,9 4,9 4,9 4,9 4,9
wet out time sec 50 50 80 50 70 80
10 min mm 270 315 315 310 299 294
self flow 30 min mm 250 302 302 283 297 294
60 min mm 220 296 285 263 293 279
ultrasonic setting start min 206 304 214 340 296 252
20°C / 24h MPa 1 n.d. 1 n.d. 1 n.d.
CMoR 110°C / 24h MPa 8 n.d. 8 n.d. 14 n.d.
800°C / 5h MPa 3 n.d. 5 n.d. 4 n.d.
 [5] Mathieu, A.; Bayoux, J.P.; Falaschi, J.P.; Vialle, M.:
400
Ageing of aluminous cement in LCC, UNITECR
Proceedings Vol. 2, Kyoto 1995, 165-172
300 [6] Sugimoto, M.; Hosokawa, K.; Sugiyama, K.;
Furukawa, K.: Deterioration of Low Cement Castables
F 30 [mm]

during Storage, Taikabutsu Overseas Vol. 19, 1999,

200
No. 2, 25-30
[7] Kockegey-Lorenz, R.; Büchel, G.; Buhr, A.; Aroni,
100 LCC-PACA J.M.; Racher, R.P.: Improved Workability of Calcia
LCC-ADS/W free Alumina binder Alphabond for Non-Cement
ULCC-M-ADS/W
NCC-AB Castables, 47th International Colloquium on
0
Refractories, Aachen 2004, 67-71
0 1 2 3 4 5 6 7 8 9
storing / months

Fig. 3. Impact of warehouse aging on flow at 30 minutes

Summary
The comparison between the climate chamber conditions
at 35°C and the warehouse aging shows, that 75% rel.
humidity is not strong enough to indicate aging trends within
4 weeks. However, 95% rel. humidity is very severe and can
show strong effects even after 1 week. Therefore 85% rel.
humidity seems the most suitable condition in order to
accelerate the aging process and indicate aging trends for the
castables.
All castables except LCC-PACA have not reached their
shelf life after 9 months storing in the warehouse. The shelf
life of the LCC with dispersing aluminas ADS/W is higher
when compared to the additive system polyacrylate + citric
acid. This becomes unusable after 2 weeks in the climate
chamber at 85% rel. humidity compared to 4 weeks for
LCC-ADS/W.
The NCC with Alphabond 300 and dispersing aluminas
ADS/W shows an aging resistance which is comparable to
the LCC-ADS/W. This is an important result, because it
shows, that Alphabond bonded castables do not require the
special packaging of the pure Alphabond binder to achieve a
reasonable shelf life.
The ULCC with dispersing aluminas M-ADS/W also
shows a high aging resistance, even under demanding
conditions in the climate chamber at 85 and 95% rel.
humidity. But the variation of test results between these
tests, where new bags were always used, is higher when
compared to the other castables.
Future evaluations will show if longer aged warehouse
material will show the same aging trend as under the
demanding conditions exhibited in the climate chamber.
No investigations have been done with regard to aging
mechanism, during this project.

REFERENCES
[1] Almatis Calcium Aluminate Cements global product
data sheet GP/005/R05/1005
[2] Almatis Alphabond 300 global product data sheet
GP/015/R03/1005
[3] Parr, C.; Revais, C.; Bunt, N.; Duncan, J. : Ageing of a
Low Cement Castable, UNITECR Proceedings Vol. 1,
New Orleans 1997, 81-90
[4] Krebs, R.: Innovationen bei verflüssigten
Feuerbetonen, presentation at the refractory and
chimney congress 2003 of GARC (German
Association of Refractories and Chimneys) in
Düsseldorf/Germany, June 16, 2003