Reactive Powder Concrete Introduction Reactive Powder Concrete (RPC) is a developing composite material that will allow the concrete industry to optimize material use, generate economic benefits, and build structures that are strong, durable, and sensitive to environment. comparison of the physical, mechanical, and durability properties properties of RPC and !PC (!igh Performance Concrete) shows that RPC possesses better strength (both compressive and fle"ural) and lower permeability compared to !PC. #his page reviews the available literature on RPC, and also presents the results of laboratory investigations investigations comparing RPC with w ith !PC. $pecific benefits and potential applications applications of RPC have also been described. !igh%Performance Concrete (!PC) is not &ust a simple mi"ture of cement, water, and aggregates. It contains mineral components and chemical admi"tures having very specific characteristics, which give specific properties to the concrete. #he development of !PC results from the materialization of a new science of concrete, a new science of admi"tures and the use of advanced scientific e'uipments to monitor concrete microstructure. !PC has achieved the ma"imum compressive strength in its e"isting form of microstructure. !owever, at such a level of strength, the coarse aggregate becomes the weaest lin in concrete. In order to increase the compressive strength of concrete even further, the only way is to rem ove the coarse aggregate. #his philosophy philosophy has been employed in Reactive Powder Concrete (RPC) . Reactive Powder Concrete (RPC) was developed in *rance in the early ++s and the world-s first Reactive Powder Concrete structure, the $herbrooe ridge in Canada, was erected in /uly ++0. Reactive Powder Concrete (RPC) is an ultra high%strength high%strength and high ductility cementitious composite with advanced mechanical and physical properties. It consists of a special concrete where the microstructure is optimized by precise gradation of all particles in the mi" to yield ma"imum density. density. It uses e"tensively the pozzolanic properties of highly refined silica fume and optimization optimization of the Portland cement chemistry to produce the highest strength hydrates. #he concept of reactive powder concrete was first developed by P. Richard and 1. Cheyrezy and RPC was first produced in the early ++s by researchers at ouygues- laboratory in *rance 2. field application of RPC was done on the Pedestrian3ieway Pedestrian3ieway ridge in the city of $herbrooe, 4uebec, Canada5. RPC was nominated for the +++ 6ova wards from the Construction Innovation *orum.
RPC has been used successfully for isolation and containment of nuclear wastes in 7urope due to its e"cellent impermeability8. #he re'uirements for !PC used for the nuclear waste containment structures of Indian 6uclear Power Plants are normal compressive strength, moderate 7 value, uniform density, good worability, and high durability 9. #here is a need to evaluate RPC regarding its strength and durability to suggest its use for nuclear waste containment structures in Indian conte"t. Composition of Reactive Powder Concrete RPC is composed of very fine powders (cement, sand, 'uartz powder and silica fume), steel fibres (optional) and superplasticizer. #he superplasticizer, used at its optimal dosage, decreases the water to cement ratio (w3c) while improving the worability of the concrete. very dense matri" is achieved by optimizing the granular pacing of the dry fine powders. #his compactness gives RPC ultra%high strength and durability :. Reactive Powder Concretes have compressive strengths ranging from 2 1Pa to ; 1Pa. Richard and Cheyrezy indicate the following principles for developing RPC< . 7limination of coarse aggregates for enhancement of homogeneity 2. =tilization of the pozzolanic properties of silica fume 5. >ptimization of the granular mi"ture for the enhancement of compacted density 8. #he optimal usage of superplasticizer to reduce w3c and improve worability 5. pplication of pressure (before and during setting) to improve compaction 6. Post%set heat%treatment for the enhancement of the microstructure 7. ddition of small%sized steel fibres to improve ductility #able lists salient properties of RPC, along with suggestions on how to achieve them. #able 2 describes the different ingredients of RPC and their selection parameters. #he mi"ture design of RPC primarily involves the creation of a dense granular seleton. >ptimization of the granular mi"ture can be achieved either by the use of pacing models0 or by particle size distribution software, such as ?I$ ; @developed by 7lem $ 1aterialsA. *or RPC mi"ture design an e"perimental method has been preferred thus far. #able 5 presents various mi"ture proportions for RPC obtained from available literature ,5,+,. #able < Properties of RPC enhancing its homogeneity and strength Property of RPC
Bescription
Coarse Reduction in aggregates are
Recommended alues
#ypes of failure eliminated
1a"imum size of fine sand is : 1echanical,
aggregate size
replaced by fine sand, with a reduction in the size of the coarsest aggregate by a factor of about 9.
Chemical E #hermo% mechanical
Dm
Improved mechanical Bisturbance 7nhanced Foung-s modulus properties of the of the mechanical values in 9 GPa H paste by the mechanical properties 09 Gpa range addition of silica stress field. fume Reduction in aggregate to matri" ratio
?imitation of sand content
olume of the paste is at least y any 2 greater than e"ternal the voids inde" of source (e.g., non%compacted formwor). sand.
#able 2< $election Parameters for RPC components Components
$election Parameters
*unction
Particle $ize
#ypes
$and
Good hardness 9 Dm Give Readily to strength, available and : Dm ggregate low cost.
6atural, Crushed
Cement
C5 $ < :J inding C2$ < 22J material, Dm C5 < 5.;J Production to C8*< 0.8. of primary Dm (optimum) hydrates
>PC, 1edium fineness
4uartz Powder
fineness
$ilica fume
ery less 'uantity of impurities
1a". reactivity 9 Dm during to Crystalline heat% 29 Dm treating *illing the . Dm Procured voids, to from 7nhance Dm *errosilicon rheology, industry
Production of secondary hydrates
$teel fibres
Good aspect ratio
(highly refined)
? < 5 H 29 mm Improve K < .9 ductility H .2 mm
?ess $uperplasticizer retarding Reduce w3c characteristic
L
$traight
Polyacrylate based
#able 5< RPC mi"ture designs from literature P. Richard and 1. $. . . $. 5 + Cheyrezy ouygues 1atte $ta'uet @++9A 6on fibred Portland Cement
.29 .25 .29 .25
.528
.529
.528
$and
.
.
.
.
.825
.85
.85
4uartz Powder
%%
.5+
%%
.5+
.2+:
.5
.5
$uperplasticizer.:.+.:.+
.20
.;
.2
.09 .09
.2:;
.209
.2;
.9 .0 .0 .+
.2;2
.2
.25
%%
%%
%%
+NC
+NC
+NC
Mater Compacting pressure
%%
%%
%%
%%
%%
29 mm *ibred *ibred fibres
$teel fibre
2 mm fibres
@+++A @2A
$ilica fume
@++0A
%%
!eat treatment 2NC +NC 2NC +NC temperature
#he ma&or parameter that decides the 'uality of the mi"ture is its water demand ('uantity of water for minimum flow of concrete). In fact, the voids inde" of the mi"ture is related to the sum of water demand and entrapped air. fter selecting a mi"ture design according to minimum water demand, optimum water content is analyzed using the parameter relative density (d 3d$). !ere d and d $ represent the density of the concrete and the compacted density of the mi"ture (no water or air) respectively. Relative density indicates the level of pacing of the concrete and its
ma"imum value is one. *or RPC, the mi"ture design should be such that the pacing density is ma"imized. 1icrostructure enhancement of RPC is done by heat curing. !eat curing is performed by simply heating (normally at +OC) the concrete at normal pressure after it has set properly. #his considerably accelerates the pozzolanic reaction, while modifying the microstructure of the hydrates that have formed . Pre%setting pressurization has also been suggested as a means of achieving high strength . #he high strength of RPC maes it highly brittle. $teel fibres are generally added to RPC to enhance its ductility. $traight steel fibres used typically are about 5 mm long, with a diameter of .9 mm. #he fibres are introduced into the mi"ture at a ratio of between .9 and 5 by volume . #he cost%effective optimal dosage is e'uivalent to a ratio of 2 by volume, or about 99 g3m 5. 1echanical Performance and Burability of RPC #he RPC family includes two types of concrete, designated RPC 2 and RPC ;, which offer interesting implicational possibilities in different areas. 1echanical properties for the two types of RPC are given in #able 8. #he high fle"ural strength of RPC is due to the addition of steel fibres. #able 9 shows typical mechanical properties of RPC compared to a conventional !PC of compressive strength ; 1Pa . s fracture toughness, which is a measure of energy absorbed per unit volume of material to fracture, is higher for RPC, it e"hibits high ductility. part from their e"ceptional mechanical properties, RPCs have an ultra%dense microstructure, giving advantageous waterproofing and durability characteristics. #hese materials can therefore be used for industrial and nuclear waste storage facilities. RPC has ultra%high durability characteristics resulting from its e"tremely low porosity, low permeability, limited shrinage and increased corrosion resistance. In comparison to !PC, there is no penetration of li'uid and3or gas through RPC 8. #he characteristics of RPC given in #able :, enable its use in chemically aggressive environments and where physical wear greatly limits the life of other concretes 2. #able 8< Comparison of RPC 2 and RPC ;
Pre%setting pressurization !eat%treating
RPC 2
RPC ;
6one
9 1Pa
2 to +OC 29 to 8OC
Compressive strength (using 'uartz sand)
0 to 25 1Pa
8+ to :; 1Pa
Compressive strength (using steel aggregate)
%%
:9 to ; 1Pa
*le"ural strength
5 to : 1Pa
89 to 8 1Pa
#able 9< Comparison of !PC (; 1Pa) and RPC 2 + Property
!PC (; 1Pa)
Compressive strength
; 1Pa
2 1Pa
*le"ural strength
0 1Pa
8 1Pa
1odulus of 7lasticity
8 GPa
: GPa
*racture #oughness
Q /3m
5SQ /3m
RPC 2
#able :< Burability of RPC Compared to !PC brasive Mear
2.9 times lower
Mater bsorption
0 times lower
Rate of Corrosion
; times lower
Chloride ions diffusion
29 times lower
?imitations of RPC In a typical RPC mi"ture design, the least costly components of conventional concrete are basically eliminated or replaced by more e"pensive elements. #he fine sand used in RPC becomes e'uivalent to the coarse aggregate of conventional concrete, the Portland cement plays the role of the fine aggregate and the silica fume that of the cement. #he mineral component optimization alone results in a substantial increase in cost over and above that of conventional concrete (9 to times higher than !PC). RPC should be used in areas where substantial weight savings can be realized and where some of the remarable characteristics of the material can be fully utilized2. >wing to its high durability, RPC can even replace steel in compression members where durability issues are at stae (e.g. in marine condition). $ince RPC is in its developing stage, the long%term properties are not nown. 7"perimental study at II# 1adras 1aterials =sed #he materials used for the study, their I$ specifications and properties have been presented in #able 0. 1i"ture Besign of RPC and !PC
•
•
•
•
•
Considerable numbers of trial mi"tures were prepared to obtain good RPC and !PC mi"ture proportions. Particle size optimization software, ?I$ ; @developed by 7lem $ 1aterialsA was used for the preparation of RPC and !PC trial mi"tures. arious mi"ture proportions obtained from the available literature w ere also studied. #he selection of best mi"ture proportions was on the basis of good worability and ideal mi"ing time. *inalized mi"ture proportions of RPC and !PC are shown in #able ;. #able 0< 1aterials used in the study and their properties $l. 6o.
$ample
$pecific Gravity
Particle size range
Cement, >PC, 95% grade @I$. 22:+ H +;0A
5.9
5 Dm H 0.9 Dm
2
1icro $ilica @$#1 C28 H +0bA
2.2
9.5 Dm H .; Dm
5
4uartz Powder
2.0
9.5 Dm H .5 Dm
8
$tandard sand, grade% @I$. :9 H ++A
2.:9
2.5: mm H .: mm
9
$tandard sand, grade%2 @I$. :9 H ++A
2.:9
.: mm H .5 mm
:
$tandard sand, grade%5 @I$. :9 H ++A
2.:9
.9 mm H .9 mm
0
$teel fibres (5 mm) @$#1 ;2 H +:A
0.
length< 5 mm E dia< .8 mm
;
$teel fibres (5: mm) @$#1 ;2 H +:A
0.
length< 5: mm E dia< .9 mm
+
2 mm ggregate @I$. 5;5 H +0A
2.0;
29 mm H mm
mm ggregate @I$. 5;5 H +0A
2.0;
2.9 mm H 8.09 mm
River $and @I$. 5;5 H +0A
2.:
2.5: mm H .9 mm
#able ;< 1i"ture Proportions of RPC and !PC
1aterials
1i"ture Proportions RPC RPC%*S !PC !PC%*SS
Cement
.
.
.
.
$ilica fume
.29
.29
.2
.2
4uartz powder
.5
.5
%
%
$tandard sand grade 2
.+
.+
%
%
$tandard sand grade 5
.9;
.9;
%
%
River $and
%
%
2.8
2.8
2 mm aggregate
%
%
.8
.8
mm aggregate
%
%
.9
.9
5 mm steel fibres
%
.2
%
%
5: mm steel fibres
%
%
%
.2
.5
.25
.25
.29
.8
.8
dmi"ture (Polyacrylate based) .5 Mater S *ibre RPC
.29
SS *ibre !PC
Morability and density were recorded for the fresh concrete mi"tures. $ome RPC specimens were heat cured by heating in a water bath at +OC after setting until the time of testing. $pecimens of RPC and !PC were also cured in water at room temperature. #he performance of RPC and !PC was monitored over time with respect to the following parameters< Compressive $trength (as per I$ 9: 5 on 9 cm cubes for RPC, cm cubes for !PC), *le"ural $trength (as per I$ 9: on 8 " 8 " : cm prisms for RPC, " " 9 cm beams for !PC), Mater bsorption (on 9 cm cubes for both RPC and !PC), 6on destructive water permeability test using Germann Instruments (on 9 cm cubes for both RPC and !PC), Resistance to Chloride ions Penetration test (on discs of diameter cm and length 9 cm as per $#1 C 228). Results *resh concrete properties #he worability of RPC mi"tures (with and without fibres), measured using the mortar flow table test as per $#1 C+ 9, was in the range of 2 H 8. >n the other hand, the worability of !PC mi"tures (with and without fibres), measured using the slump test as per $#1 C25 :, was in the range of 2 H 9 mm. #he density of fresh RPC and !PC mi"tures was found to be in the range of 29 H 2:9 g3m5.
Compressive strength #he compressive strength analysis throughout the study shows that RPC has higher compressive strength than !PC, as shown in *ig. . Compressive strength at early ages is also very high for RPC. Compressive strength is one of the factors lined with the durability of a material. In the conte"t of nuclear waste containment materials, the compressive strength of RPC is higher than re'uired.
*ig < Compressive strength of RPC and !PC he ma"imum compressive strength of RPC obtained from this study is as high as 2 1Pa, while the ma"imum strength obtained for !PC is 09 1Pa. #he incorporation of fibres and use of heat curing was seen to enhance the compressive strength of RPC by 5 H 9. #he incorporation of fibres did not affect the compressive strength of !PC significantly. *le"ural strength Plain RPC was found to possess marginally higher fle"ural strength than !PC. #able + clearly e"plains the variation in fle"ural strength of RPC and !PC with the addition of steel fibres. !ere the increase of fle"ural strength of RPC with the addition of fibres is higher than that of !PC. #able +< *le"ural strength (as per I$ 9:) at 2; days (1Pa) RPC
RPC%*
!PC
!PC%*
6CS
!MCSS
6CS
!MCSS
6CS
6CS
2
;
22
;
S6ormal Curing
SS!ot Mater Curing
s per literature5, RPC 2 should have an appro"imate fle"ural strength of 8 1Pa. #he reason for low fle"ural strength obtained in this study could be that the fibres used (5 mm) were long. *ibre reinforced RPC (with appropriate fibres) has the potential to be used in structures without any additional steel reinforcement. #his cost reduction in reinforcement can compensate the increase in the cost by the elimination of coarse aggregates in RPC to a little e"tent.
Mater absorption *ig. 2 presents a comparison of water absorption of RPC and !PC. common trend of decrease in the water absorption with age is seen here both for RPC and !PC. #he percentage of water absorption of RPC, however, is very low compared to that of !PC. #his 'uality of RPC is one among the desired properties of nuclear waste containment materials.
*ig. 2< Mater absorption of RPC and !PC
#he incorporation of fibres and the use of heat curing is seen to marginally increase the water absorption. #he presence of fibres possibly leads to the creation of channels at the interface between the fibre and paste that promote the uptae of water. !eat curing , on the other hand, leads to the development of a more open microstructure (compared to normal curing) that could result in an increased absorption. Mater permeability
#he non%destructive assessment of water permeability using the Germann Instruments e'uipment actually only measures the surface permeability, and not the bul permeability lie in conventional test methods. comparison of the surface water permeability of RPC and !PC is shown in *ig. 5. It can be seen from the data that water permeability decreases with age for all mi"tures. 2;th day water permeability of RPC is negligible when compared to that of !PC (almost 0 times lower). s in the case of water absorption, the use of fibres increases the surface permeability of both types of concrete.
*ig. 5< $urface Mater Permeability of RPC and !PC Resistance to chloride ion penetration Results of rapid chloride permeability test conducted after 2; days of curing are presented in #able . Bata indicate that penetration of chloride increases when heat curing is done in concrete. #otal charge passed for normal%cured RPC is negligible compared to the other mi"tures. 7ven though heat%cured RPC shows a higher value than normal%cured RPC, in absolute terms, it is still e"tremely low or even negligible ( Coulombs). #his property of RPC enhances its suitability for use in nuclear waste containment structures. #he data also indicate that addition of steel fibres leads to an increase in the permeability, possibly due to increase in conductance of the concrete. #he !PC mi"tures also showed very low permeability, although higher compared to RPC. #able < Rapid Chloride Permeability #est (as per $#1 C 22) RPC
RPC with
!PC
fibres 6CS
!MCSS 6CS !MCSS 6CS !MCS
Cumulative 8 Charge passed (less than in Coulombs )
+8
8
8
29
;9
$#1 C22 ery ery ery ery 6egligible 6egligible classification low low low low S6ormal Curing
SS!ot Mater Curing
$ummary Reactive Powder Concrete (RPC) is an emerging technology that lends a new dimension to the term Thigh performance concrete-. It has immense potential in construction due to its superior mechanical and durability properties compared to conventional high performance concrete, and could even replace steel in some applications. #he development of RPC is based on the application of some basic principles to achieve enhanced homogeneity, very good worability, high compaction, improved microstructure, and high ductility. RPC has an ultra%dense microstructure, giving advantageous waterproofing and durability characteristics. It could, therefore, be a suitable choice for industrial and nuclear waste storage facilities. laboratory investigation comparing RPC and !PC led to the following conclusions< •
•
•
•
ma"imum compressive strength of +; 1Pa was obtained. #his is in the RPC 2 range (09 1Pa H 229 1Pa). #he ma"imum fle"ural strength of RPC obtained was 22 1Pa, lower than the values 'uoted in literature (U 8 1Pa). possible reason for this could be the higher length of fibres used in this study. comparison of the measurements of the physical, mechanical, and durability properties of RPC and !PC shows that RPC possesses better strength (both compressive and fle"ural) and lower permeability compared to !PC. #he e"tremely low levels of water and chloride ion permeability indicate the potential of RPC as a good material for storage of nuclear waste. !owever, RPC needs to be studied with respect to its resistance to the penetration of heavy metals and other to"ic wastes emanating from nuclear plants (such as Cesium 50 ion in alaline medium) to 'ualify for use in nuclear waste containment structures.
References
Richard P, and Cheyrezy 1, VComposition of Reactive Powder ConcreteW, Cement and Concrete Research, ol. 29, 6o.0, (++9), pp. 9 H 9. 2. itcin P.C, VCements of yesterday and today Concrete of tomorrowW, Cement and Concrete Research, ol. 5, (2), pp 58+ % 59+. 3. lais P. F, and Couture 1, VPrecast, Prestressed Pedestrian ridge % Morld-s first reactive powder concrete structureW, PCI /ournal, ol. 88, (+++), pp. : % 0. 4. Bauriac C, V$pecial Concrete may give steel stiff competition, uilding with CincreteW, #he $eattle Baily /ournal of Commerce, 1ay +, ++0. 5. asu P.C, VPerformance Re'uirements of !PC for Indian 6PP $tructuresW, #he Indian Concrete /ournal, $ep. +++, pp. 95+ H 98:. 6. onneau >, ernet C, 1oranville 1, and itcin P. C, VCharacterization of the granular pacing and percolation threshold of reactive powder concreteW, Cement and Concrete Research, ol. 5 (2) pp. ;: H ;:0. 7. Goltermann P, /ohansen , and Palbol ?, VPacing of ggregates< n lternative #ool to Betermine the >ptimal ggregate 1i"W, CI 1aterials /ournal, $ep%>ct. ++0, pp. 859 H 885. 8. 7lem $ website H http<33www.silicafume.net3 9. 1atte and 1oranville 1, VBurability of Reactive Powder Composites< Influence of $ilica *ume on the leaching properties of very low water3binder pastesW, Cement and Concrete Composites, 2 (+++) pp. % +. 10. $ta'uet $, and 7spion , VInfluence of Cement and $ilica *ume #ype on Compressive $trength of Reactive Powder ConcreteW, : th International $ymposium on !PC, =niversity of russels, elgium, (2), pp. H 8. 11. icley /. , and 1itchell B, V $tate%of%the%rt Review of !igh Performance Concrete $tructures uilt in Canada< ++%2W, (2), pp. +: H 2. 12. !BR 7ngineering Mebsite on Reactive Powder Concrete, ?ast modified 6ov. +++, http<33www.hdrinc.com3engineering3engres.htm 13. Indian $tandard Besignation I$ 9:%+9+, V1ethods of #est for $trength of Concrete,W I$, 6ew Belhi, 22. 14. $#1 $tandard Besignation C22%+0, V$tandard #est 1ethod for 7lectrical Indication of Concrete-s bility to Resist Chloride Ion Penetration,W $#1, Pennsylvania, 2. 15. $#1 $tandard Besignation C+%++, V$tandard #est 1ethod for Compressive $trength of !ydraulic Cement 1ortars,W $#1, Pennsylvania, 2. :. $#1 $tandard Besignation C85%, V$tandard #est 1ethod for $lump of !ydraulic Cement Concrete,W $#1, Pennsylvania, 2. 1.