In order to obtain high strength of concrete, three methods are available in general. The first is to increase the strength of the binder, the second is to select aggregate with high strength, and the third is to improve the bond at the interface of aggregate and binder (Refs. 3.1 and 3.2).
Among them, the most popularly adopted is the first method. This is be- cause of the fact that the binder strength of concrete in the ordinary strength range is smaller than the strength of aggregate, hence the strength of
61
concrete is dictated by that of binder. The increase of binder strength requires the cement and mineral admixtures suitable for high strength, and reduction of water-binder ratio as the most effective means in terms of mix design. This is a well-known fact by the classical name of "water-cement ratio" theory. In addi- tion, to maintain workability of concrete within the practical limit without increasing the unit water content while keeping the low water-binder ratio, that is, without increasing the unit binder content, it is necessary to develop chemical admixtures with high capability of dispersing cement and mineral admixtures.
The increase of binder strength naturally results in producing concrete whose strength is strongly affected by the aggregate strength. Hence the selec- tion of aggregate suitable for high strength concrete becomes an important issue.
Finally, it is an established fact (Ref. 3.3) that the concrete strength depends microscopically on the structure of the transition zone between aggregate and binder. For the strength improvement of the transition zone, not only the reduction of water-binder ratio but also the use of mineral ad- mixtures with ultrafine particle such as silica fume was found to be effective.
Based on these general considerations for high strength of concrete, this subsection presents research accomplishment on the development of cement, chemical and mineral admixtures, and the selection of aggregate, both suitable for high strength, and achievement on the mix proportioning method of high strength concrete.
3.1.1.1. Cement
A series of experiment was carried out in the New RC project with the aim of developing the cement suited for high strength concrete and of developing the quality standard of such cement, leading to test results as summarized below.
Compressive strength of mortar with water-cement ratio of 25 to 65 percent was studied using ordinary, high-early strength, moderate heat, and type B blast furnace slag portland cement. As shown in Fig. 3.1, mortar strength is affected by cement type for water-cement ratio greater than 30 percent, but the difference is small for water-cement ratio of 25 percent. The figure also shows mortar strength for type B fly ash cement and for ordinary portland cement with silica fume, which resulted in lower strength even at the water-cement
ratio of 25 percent.
Ordinary Portland ceaent
High-early-strength Portland ceaent Moderate heat Portland cement Type B blast furnace ceaent Type B fly-ash ceaent OPC with s i l i c a fume Age'-28 days
Fig. 3.1.
ratio.
25 30 35 40 50 later-ceaent ratio (%)
Strength of mortar with various cement types in the range of low water-cement
130 120
1 1 0°
1 90
J e o
70 -
- ^ > - o
°
^ o 9 ldằ1*
i
O
•
I
O > 0
* •
I I I 40 60 80
Percentage of base cement (%)
100
Fig. 3.2. Relationship between base cement percentage in particle size distribution controlled cement and mortar strength.
Setting and compressive strength tests were conducted of mortar with water-cement ratio of 30 percent and sand-cement ratio of 1.4, using ordinary, high-early strength, moderate heat, and type B blast furnace slag portland cement of various makers. High strength could be obtained by any cement, but the correlation between mortar strength and cement strength by JIS (Japanese
Industrial Standard) method was not observed. This indicates that JIS may not be sufficient as a quality standard of cement for high strength concrete.
The fluidity of mortar and concrete using commercially available cement is greatly impaired when the water-cement ratio is low. To increase the fluidity of mortar with low water-cement ratio, particle size distribution controlled cement was manufactured on trial, by replacing part of ordinary portland cement by pulverized matter such as coarse particle portland cement or finely ground limestone. Tests of mortar and concrete were conducted using this particle size distribution controlled cement, and mortar with good fluidity (flow value of 200 mm) was obtained even with water-cement ratio of 20 percent or less. The fluidity of concrete using this cement at the water-cement ratio of 20 percent was also excellent, and as shown in Fig. 3.2, compressive strength of more than 100 MPa was achieved for the new cement with 60 to 80 percent base cement (40 to 20 percent replacement).
Quality standards for cement to be used for concrete between 36 MPa and 60 MPa were developed, which will be explained in Chapter 8.
3.1.1.2. Aggregate
The relationship between the quality of high strength concrete and the quality of aggregate was studied experimentally, to establish method for selection of aggregate suitable for high strength concrete. Major findings were as follows.
Assuming that concrete is a two-element system of matrix (mortar) and coarse aggregate, and that mortar is another two-element system of matrix (cement paste) and fine aggregate, strength variation of concrete and mortar was studied by varying the amount of aggregate from various places while keeping the matrix quality constant. Figure 3.3 shows results for concrete. For both water-cement ratio of 25 percent and 35 percent, concrete was made using four different kinds of coarse aggregate, which are marked O, T, K and D. Compression tests were made at the age of 28 days. With the increase of unit coarse aggregate content of K or D, compressive strength decreased almost linearly, while it remained more or less constant with the increase of good quality aggregate such as O or T. Thus it is clear that coarse aggregate with inferior quality affects the strength of high strength concrete remarkably.
Figure 3.4 shows a similar results as above for mortar. Using nine different kinds of sand of varying sand-cement ratio while keeping the water-cement ratio constant at 25 percent, mortar strength was tested at ages of 7 or 28 days. The compressive strength showed tendency to decrease as sand
Unit coarse aggregate content (!/•*)
00
80 0
1 200 400 600
i i i
S \ . ^ " & • Hortar
N X oo
Nof-O A T
^ v v K
O D
100 °
80
200 400 600
60
40
(b) W/C = 35 X (a) f/C = 25 %
Fig. 3.3. Relationship between unit coarse aggregate content and compressive strength of concrete using various kinds of coarse aggregate (O, T, K and D).
Si 1 2 0
1 1 0 1 0 0 90
•M SO
I 80
to
ô S 70
B3 D
| 90 80 70 60
Sand-ceoent r a t i o
0.5 1.0 1.5 2.0
T 1 r
W/C=25%
^ ^ Age^ 28 days
O A 1 © A 3 O B i D O F - ô A 2 â A 4 • C v E
NA
Fig. 3.4. Relationship between sand-cement ratio and compressive strength of mortar using various kinds of sand ( A 1 ~ F ) .
8r~ W = 1 6 0 k g / m ' , Drying p e r i o d s months
Sfc...
20 22.5 25 25N 30 20 22.5 25 25N 30 20 22.5 25 25N 30 w/B (%)
Andesite Limestone Hard Sandstone Pig. 3.5. Effect of kinds of coarse aggregate on t h e drying shrinkage of concrete.
content increases for all kinds of sand used, but the decreasing trend was more conspicuous for some sand, for example with marks D and E.
A study into the effect of aggregate size, shape, and unit coarse aggregate content on the compressive strength was conducted. No effect was found of aggregate size and aggregate content on the concrete strength, but angular shape was found to be advantageous for high strength.
Crushed hard sandstone, limestone, and andesite aggregates with BS (British Standard) crushing value of 15 to 20 were used for high strength con- crete of 100 to 120 MPa compressive strength, to investigate Young's modulus at 28-day age and drying shrinkage at 6-month age. Limestone concrete showed higher Young's modulus of about 50 GPa compared to about 40 GPa of hard sandstone or andesite concrete. Drying shrinkage was also smaller for lime- stone concrete as shown in Fig. 3.5, which illustrates shrinkage strain after 6 months of drying period for concrete using three kinds of coarse aggregate and water-cement ratio ranging from 20 to 30 percent while keeping the unit water content of 160 kg/m3 constant. All concrete except for 25N used the cement with 15 percent replacement by silica fume for the binder.
High strength concrete with 120 MPa strength can be made by using selected aggregate, both coarse and fine, and the fluidity can be improved by using fine aggregate with adjusted fineness, i.e. by removing very fine com- ponent from the fine aggregate.
3.1.1.3. Chemical Admixtures
Various commercially available as well as newly developed chemical admix- tures, generally known as air-entraining and high-range water-reducing agents,
were compared in a series of unified tests. Concrete with four grades of com- pressive strength were considered. They were 40 MPa at water-cement ratio of 40 percent, 60 MPa at water-cement ratio of 30 percent, 80 MPa at water- binder ratio of 25 percent of both plain concrete and concrete mixed with silica fume or ground granulated blast furnace slag, and 100 MPa at water-binder ratio of 22 percent of concrete mixed with silica fume or ground granulated blast furnace slag. Items such as relationship between unit water content and admixture addition ratio to achieve the target slump or air content, time varia- tion of slump, setting time, compressive strength, drying shrinkage, and freeze- thaw resistance, were studied.
As an example, the case of 60 MPa concrete at water-cement ratio of 30 percent is illustrated below. Figure 3.6 shows change with time of slump of concrete using various brands of chemical admixtures. Unit water content of 165 kg/m3 was kept content, and air content was in the range of 3 to 4 percent. Some brands, e.g. marks A and G, showed larger slump loss with time than other brands. Figure 3.7 shows range of setting time of concrete with unit water content of 165 kg/m3 and 150 kg/m3 using the same ten brands of chemical admixtures as above. Some brands showed very long setting time, particularly when the unit water content was low. The drying shrinkage strain of concrete using certain brands of admixture was also found to be longer.
L_J 1 i I i 0 15 30 60 90
Time (min)
Fig. 3.6. Change with time of concrete slump using various brands of air-entraining and high-range water-reducing agents.
1 2 5 0
1000
•S 750
500
250
Z l W = 1 6 5 k g / m3
H i W = 150kg/m3
W/C=30%
_L
D
_L
1
EH 151
I ri
S A B C D E F G
Brand of admixtures
H I
Fig. 3.7. Setting time of concrete using various brands of air-entraining and high-range water-reducing agents.
120
100 80
60
40
20
W/C=3Q%
• W=165kg/m\ Air=3~4X M W=165kg/m\ Air=2 % W& W=150kg/m3, Air=3~4 X H I W=150kg/m3, *ir=2 *
I d e n o t e s r a n g e of max and min
T
[W
4 if"
I ll
rfi i
i
28 91
Age (days)
Fig. 3.8. Compressive strength of high strength concrete using various brands of air- entraining and high-range water-reducing agents.
Nevertheless, compressive strength of concrete was satisfactory for all brands of admixtures. Figure 3.8 shows compressive strength for four different combination of unit water content and air content at five different ages. Water- cement ratio of 30 percent was kept constant, aiming at compressive strength of 60 MPa. As can be seen in the figure, the target strength was more than satisfied at the age of 28 days. It was even cleared at 7 days in this test.
The cases of higher strength concrete indicated the significance of air en- trainment on the freeze-thaw resistance. Figure 3.9 is the results of freezing and thawing test of 80 MPa concrete with water-cement ratio of 25 percent, indicating the relationship of spacing factor and durability factor, which is the relative value of dynamic modulus of elasticity at the end of freezing and thawing test. For plain concrete without mineral admixture with air of 3 to 4 percent and plain concrete with ground granulated blast furnace slag, no reduction of durability factor was observed. However, plain concrete with low air content and plain concrete with silica fume showed inferior durability.
Based on these unified tests, quality standard and usage guideline for chemical admixtures for 60 MPa high strength concrete were developed. Test data for higher strength concrete were not compiled into practical form as above at the present stage, but they are believed to throw some light into the future advancement of the concrete research.
Mineral admixture air content O ; not mixed ^ 2 %
• : not mixed 3~4 % Q : silica fume <J2 % A ; blast furnace slag ^ 2 %
• Vôằ L<c 4 ^ ^ cg^ ^ Ê-
D Q
D a
O O D
a
° o
J I I I I I I I I I I I I I l_
0.1 0.3 0.5 0.7 0.9 1.1 1.3 1.5 Spacing factor (mm)
Fig. 3.9. Durability factor and spacing factor of concrete using various brands of air- entraining and high-range water-reducing agents.
100
! - 3 60
1 .
3.1.1.4. Mineral Admixtures
Mineral admixtures for high strength concrete are to replace a part of cement and form a part of binder. Admixtures such as silica fume, fly ash fume, ground granulated blast furnace slag, and etringite type special admixture were con- sidered. Fly ash fume is obtained by processing fly ash at high temperature, thereby evaporating silicon dioxide whose boiling temperature is relatively low among substances in the fly ash, and then coagulating it at the lowered tem- perature for collection. Etringite was used to be known as cement bacillus, but the etringite type special admixture is a kind of mineral admixture mainly con- sisting of Type 2 anhydrous gypsum, with the aim of growing hardened binder body with fine structure by utilizing the growth of needle-shaped crystal of etringite (formed by the reaction of aluminate in the cement and gypsum).
Followings are major findings of unified tests for mineral admixtures.
Fluidity and compressive strength of cement paste, mortar and concrete were tested using silica fume or fly ash fume whose specific surface area was modified to range of 260 000 to 700 000 cm2/g. It was found that replacement of 10 to 15 percent of silica fume or fly ash fume lead to the maximum compressive strength. Greater specific surface area of fly ash fume resulted in the increase of strength.
Workability, strength development and freeze-thaw resistance of mortar and concrete were studied using ground granulated blast furnace slag with specific surface area of 6000, 8000 and 10 000 cm2/g. The strength development was slow at low temperature, but strength was improved when the specific surface area was greater.
Strength development of concrete with low water-binder ratio was mea- sured where the binder consisted of three components of cement, ground granulated blast furnace slag, and silica fume or fly ash fume. The concrete with the three components showed greater increase of strength at long term than two component concrete.
Properties of concrete with etringite type special admixture were investi- gated, and it was shown that increase of compressive strength of about 15 MPa was obtained by adding this admixture. Figure 3.10 shows the effect of curing condition on this kind of concrete, in terms of compressive strength at 28 and 91 days under four different curing conditions, i.e. exposed to air after 2, 4, 7, or 28 days of wet curing either in the form or under standard curing condi- tion. Strength of cylinders under standard curing condition is shown in all four groups as a common reference. An exception in this figure is the leftmost group
130 120 I 110
S ioo -
90
80 - 70
0 S t r i p at 2d. then standard-cured 0 Strip at 2d tin, sealed
Strip at 2d then air-cured
I Strip at 2d standard-cured 2d than air-cured
Strip at 4d then air-cured
1(3 Strip at 2d standard-cured 5d than aircured fU Strip at 7d than air-cured
2 Strip at 2d standard-cured 25dthen air-cured
• Strip at 28d than ail-cured
(w/ad. ) (w/out ad.) (w/ad. ) (w/out ad. ) (w/ad.) (ằ/out ad. ) (w/ad.) (w/out ad.)
Fig. 3.10. Effect of curing condition on t h e compressive strength of concrete using etringite type special admixture.
where strength of sealed cylinders is shown, which was happened to be simi- lar to that under standard curing. In each group strengths with and without etringite type admixture are compared. For all four curing conditions, strength increase due to addition of the admixture is clearly seen. Furthermore, this admixture improves the strength development in the concrete exposed in the air. Concrete with this admixture revealed strength comparable or even better strength compared to standard curing even after 4 or 7 days of wet curing condition.
These mineral admixtures are very important for high strength concrete, especially in excess of 60 MPa, and the individual special features must be carefully considered in their practical use.
3.1.1.5. Mix Design
Aiming at developing specification for mix design of high strength concrete of 60 to 80 MPa specified strength, procedure to determine water-cement ratio or
water-binder ratio, unit water content, unit bulk volume of coarse aggregate, and dosage of chemical admixtures was studied to achieve the required average (proportioning) strength, air content and slump or slump flow. Major findings are summarized below.
In order to find relationship between required average strength and water- cement ratio or water-binder ratio, and relationship between unit water con- tent or dosage of chemical admixture and workability, tests were made on the various concrete properties in the range of water-binder ratio of 15 to 40 percent and unit water content of 145 to 175 kg/m3, using air-entraining and high-range water-reducing agent, silica fume and ground granulated blast furnace slag 8000. For the same slump or slump flow of fresh concrete, it was found that flow speed of concrete was faster, hence the workability was better, when silica fume was used or when unit water content was increased. Further- more, as shown in Fig. 3.11, compressive strength at ages of 7, 28 and 91 days increased in proportion to binder-water ratio in the range of water-binder ratio of 25 percent or more, but for lower water-cement ratio compressive strength did not increase with the increase of binder-water ratio. As shown in the figure, concrete with ordinary portland cement (OPC) showed the strength at 28 days of about 100 MPa, and concrete with silica fume replacement of 15 percent (OPC + SF) and concrete with ground granulated blast furnace slag 8000 re- placement of 30 percent (OPC -I- BS) showed the strength at 28 days of about 120 MPa, both more or less constant for different binder-water ratio above 4.
140 120
S 1 "
I 80
•B 60 40
j£^
xfe*
< ? /
. , — &
\
o* /
/
/ a
OPC OPC+SF 0PC+8S Age=7 days Age-28 days Age=91 days
I _1_ I I
40
4 5
Binder-water r a t i o J I L I
30 25 22.5 20 18.2 16.7
Water-binder ratio (X)
15
Fig. 3.11. Relationship between water-binder ratio and compressive strength.
Relationship between concrete mix and various properties was studied of
60 MPa concrete with water-cement ratio of 30 percent and slump 21 cm, and of 80 MPa concrete with water-binder ratio of 25 percent and slump 25 cm. Figure 3.12 indicates setting time of 60 MPa concrete with the same
1000 -
800
600
S 400
200
W7C=30%
slunp=2] ca
D" n
m final setting I |~1 initial setting ]
j right after sizing Q " ^ ] 90 iin, after ailing I ; aortar
J I I 140 160 180
Unit water content ( k g / m3)
200
Fig. 3.12. Relationship between unit water content and setting time.
1 0 0 95 90 85 80 75
(2-S-^(1.9%>
V//C = 30%, sluằp=21 en standard curing 28 days
J l . 5 % ) l(1.2%)
>(1.0%)
Numbers in parentheses indicate dosage of admixture
1 1 I I I I I L_
130 140 150 160 170 180 190 200 210
Unit water content ( k g / m * )
Fig. 3.13. Relationship between unit water content and compressive strength.
1 4 0 -
130
~ 1 2 0
1 1 0
Water-binder ratio:25% Age:28 days
ja a .
1 0 0
90
W=160kg/m3
W=200kg/m3
I _L _L _L _L _L
0 10 20 30 40 50 60 70 Replacement ratio by mineral admixture (%)
Fig. 3.14. Relationship between replacement ratio of O P C by mineral admixture and com- pressive strength.
water-cement ratio but different unit water content, showing longer setting time for smaller unit water content. To keep the water-cement ratio constant while reducing unit water content, one must reduce the paste content, and to keep the slump constant one has to use large amount of chemical admixture leading to higher viscosity. This is the reason for longer setting time for smaller unit water content in the figure. On the other hand, as shown in Fig. 3.13, concrete with larger unit water content showed smaller compressive strength even under constant water-cement ratio.
Figure 3.14 is for 80 MPa concrete where silica fume or ground granulated blast furnace slag 8000 is used as mineral admixture. This figure shows the compressive strength for different replacement ratio of mineral admixture, and it can be seen that silica fume replacement of 10 percent and blast furnace slag replacement of 30 to 50 percent resulted in maximum strength, for any of the unit water content considered.
Appropriate unit bulk volume of coarse aggregate can be determined corresponding to the slump or slump-flow value referring to the mix design for unified tests of chemical admixtures and other available sources.
Based on the above-mentioned studies, a general procedure of mix calcu- lation is organized as shown by a flow chart in Fig. 3.15. Detail of proposed