- Email: [email protected]
The properties of recycled precast concrete hollow core slabs for use as replacement aggregate in concrete
Waste Materials in Construction G.R. Woolley, J.J.J.M. Goumans and P.J. Wainwright (Editors) 9 2000 Elsevier Science Ltd. All rights reserved.
769
The Properties of Recycled Precast Concrete Hollow Core Slabs for Use as Replacement Aggregate in Concrete B E Marmash and K S Elliott School of Civil Engineering University of Nottingham Nottingham NG7 2RD United Kingdom.
Summary Waste concrete from prestressed precast hollow cored floor units has been recycled for use as replacement aggregate in concrete (RCA). Waste concrete blocks were crushed t o - 1 4 mm using cone, impact and jaw crushers. All produced acceptable physical and mechanical properties, although the impact crusher was best suited in most cases. The water absorption of the RCA was 3 to 4 times greater than natural limestone and river gravel used for control purposes. The fine RCA was at the coarse end of the British Standard limit. Concrete made using zero, 20% and 50% replacement of coarse and fine RCA had increased workability at 20% replacement, but this reduced considerably at 50% replacement. Compressive strength of concrete made with RCA was generally within +5 N/mm 2 of the control value of 62 N/mm 2. RCA from the cone crusher produced the highest strengths. The main conclusion is that concrete made with up to 50% replacement of both coarse and fine RCA appears to be comparable with natural aggregate concrete in terms of workability and compressive strength.
1
Introduction
The precast concrete hollow core floor industry produces a considerable amount of wasted concrete elements, due mainly to the manufacturing processes and in part to natural wastage at the ends of the casting beds. Figure 1 shows how prestressed hollow core units (hcu) are manufactured by extrusion or slip-forming through a machine on long beds, typically 100 m in length x 1.2 m wide. Waste material is made at the beginning and end of each bed, typically 0.3 to 0.4 m 3 per casting. After detensioning the units are cut to length. Waste material is therefore made if the cumulative length of the units does not equal the net cast length. This can be between zero and 0.5 m 3 per casting. The total waste generated in the UK is around 5% of the production. Waste material from hcu is high grade and uncontaminated material. The parent concrete is hard and of compressive strength between 50 to 80 N/mm 2. It is manufactured from Portland cement, and from clean and reliable sources of 10 mm to 14 mm limestone or gravel. The grading of the coarse and fine aggregates is carefully controlled, and together with a
770
Figure 1. End line waste of hollow core slab units water cement ratio of around 0.3 the resulting concrete is of a high density and low porosity. It is also extremely brittle and will fracture into flaky shapes with acute edges. In commercial crushers large amounts of fine aggregate are produced, which cause concern for the reintroduction to hcu manufacture. To date recycled coarse aggregate (RCCA) has been considered at replacement levels of up to 20% while the recycled fine aggregate (RCFA) only up to 10%. A recent project (P.I.T) t~ conducted by the member companies of the UK's Precast Flooring Federation has found that the effects of using 20% of RCCA + 10% of RCFA in precast concrete are small. They indicated that the differences recorded in compressive and flexural strength were mostly less than the variability of the test. Also they reported that shrinkage and creep measurements were more consistently affected but still fairly small and should not contravene any specification requirements. Differences resulting from the crushing method were not investigated. To complement this work, this present research aims to study the properties of recycled concrete from hcus crushed using three different methods - cone, jaw and impact crushers. Commercial hcus, manufactured by Richard Lees Ltd. using the Spiroll extrusion technique, were obtained for this study. The mix content for the parent concrete is given in Table 1. The resulting crushed material was first separated into coarse (10 mm and 14 mm) and fine (< 5 mm) fractions. It was tested for those properties which are important to the reintroduction of RCA into hcu production, namely grading, water absorption, density, shape, ten-per-cent fines value, workability and strength. The recycled aggregates were also tested in concrete mixes in which the aggregates in a 'reference' mix were substituted with varying Table 1: Mix content for the parent concrete (hcus) 14 mm
10 mm
Sand
Cement Class 52.5N
Pozzolan
Water
340 Kg
440 Kg
500 Kg
200 Kg
60 Kg
50 Kg
771 proportions of recycled concrete aggregate. The chosen percentage replacement was 20% and 50% - the former represents a typical limit for RCCA proposed in P.I.T project, and the latter is made deliberately large in order to investigate the sensitivity of the said mechanical and physical properti?s. This paper reviews the determination of mechanical and physical properties of RCA alone. The results are compared for compliance with the relevant British Standards. The paper also presents the results of tests carried out on fresh and hardened concrete made with natural aggregates and RCA.
2
Crushers
To enable discussion on the properties of RCA obtained from the crushing operation, it is first necessary to review this in relation to crushing hardened concrete. It should be noted that most crushing machines were developed for crushing rock into sizes with equivalent diameters of 50 mm or greater. In some cases special modifications to the orifice had to be made to produce RCA of 14 mm size.
2.1
Cone Crusher
Cone crushers are one of the major categories of gyrating crushers, which have developed into being one of the most important types of machine in use in quarrying (2). The principle is shown in Figure 2. It uses a repeated compression action with fixed and moving crushing members. The long stroke and high speed agitate the feed in its passage through the chamber. At the lower extremity of the Crushing chamber the faces of the two crushing member are so shaped that they are parallel for a section, resulting in the larger pieces being assured of having at least one dimension equal to, or less than, the setting, quoted as closedside setting (CSS) when properly fed and certainly all products less than twice the CSS. In the operation to crush hcu, the size of the feed was up to 350 mm and the CSS was set at about 14 mm. The resulting RCA was acceptable in appearance and seems to be slightly towards flaky and elongated in shape. There is no distinguished dust on the coarse particles. The amount of cement paste attached to the virgin aggregate was quite large.
2.2 Jaw Crusher Jaw crushers are also commonly used in quarries. Its principle could be summarized in that the feed is subjected to repeated pressure as it passes downwards and is progressively reduced in size until it is eventually small enough to pass out of the crushing chamber. See Figure 2. The size of machine also affects its speed which decreases as the crusher size increases. The angle between the crushing faces is normally between 19~ and 22 ~ This angle is set to allow the crushing force to be transmitted to very hard materials without a tendency for the feed to rise itself out of the crushing zone and so cause abrasive wear to the liners. The setting is usually measured as a CSS, i.e. when the jaws are at their closest position, and some times as open side setting with the jaws at their greatest distance apart. In the operation to crush hcu, the size of the feed was nominally 25mm and the CSS was set at about 15mm. The resulting RCA was good in appearance and seems to be slightly elongated with no distinguish dust on the coarse particles. The amount of cement paste attached to the virgin aggregate was quite large.
772
Figure 2: Cone Crusher (left), Impact crusher (middle) and Jaw Crusher (right).
2.3 Impact crusher Impact crushing could be described as impact breaking since the feed is fragmented by kinetic energy introduced by a rotating mass (the rotor) which projects the material against a fixed surface causing it to shatter causing further particles size reduction. The process causes the material to break along its natural cleavage planes and this yields a good product shape free from stress. The rotor speed is fundamental to the breaking process where the higher speed the higher reduction factor; the size of the material fed into the crusher to the size of the finished product. In the operation to crush hcu, the size of the feed was nominally 500mm. The opening size was 700mm x 500mm with 1800 RPM rotation. The resulting RCA was less elongated towards rounded shape with acute edges. There was a distinguish dust on the coarse particles. The amount of cement paste attached to the virgin aggregate also was quite large.
2.4 Summary of Crushing Operations 9 9 9
3
In visual appearance the RCA produced from the three crushers seems to be relativly the same in apperance but slightly better shape could be seen from impact crusher. The amount of fines obtained from the crushers was cearly high for impact crusher then followed by cone and then jaw. The speed at which the quantity of RCA was produced by the different crushers was convenient. It is qiute high for the imapct crusher while the cone and jaw were relatively the same but slower than the impact.
Physical properties of Recycled Concrete Aggregate (RCA)
3.1 Density and Water Absorption The presence of internal pores in the crushed particles has an influence on the porosity and absorption properties of RCA. These properties have a major effect on the workability and durability of concrete with a low water-cement ratio, especially used in hcu production. They also have an influence on the bond to hydrated cement past as well as the concrete resistance to freezing and thawing, and to a lesser extent carbonation. Water absorption and density were measured (to BS 812, Part 2, 1995 (3)) for the RCA and natural crushed
773 carboniferous limestone (obtained from Tarmac Quarry P r o d u c t s - Retford ). The results given in Tables 2 and 3 are the mean of 2 samples.
3.1.1
Density and Water Absorption of Coarse RCCA
It was expected for the RCCA to have a higher water absorption and lower density than the natural limestone because of the cement mortar attached to it. The results given in Table 2 show that the water absorption for 10 mm and 14 mm RCCA is around four times greater than that of the limestone with similar size. Table 3 shows that the surface saturated density (SSD) is between 4% and 7% lower than that of natural limestone with similar size. Conceming the effects of different crushing methods on water absorption and density, there were no distinctive differences nor trends. This might give an indication to that there is no significant difference on the amount of cement mortar attached to the RCA obtained from the three different crushing machines. In other words, these crushers (if the fed materials are identical) could produce RCCA with approximate close percentages of attached mortar to the aggregate particles. (This could not be considered as a definite conclusion as other methods need to be used to measure accurately the amount of attached mortar to the RCA.)
3.1.2 Density and Water Absorption of Fine RCFA The water absorption tbr RCFA is higher than that for natural river gravel sand. This is also due to the attached mortar. The results given in Table 2 show that the water absorption for RCFA is around three times greater than that of the natural gravel sand of similar maximum size and grading profile. Table 3 shows that the surface saturated density (SSD) is 9% lower than that of natural gravel sand of similar size. There is no significant difference between the values of water absorption for RCFA obtained from the different crushing methods. This also probably limits the influence of the crushing machines on the amount of cement mortar attached to RCFA.
Table 2. Water absorption of RCCA and RCFA Aggregate t~cpe Recycled Recycled Recycled Limestone Gravel sand
Crusher Cone Impact Jaw -
Fine < 5 mm 6.8 % 5.8 % 6.7 % 1.7 %
Coarsel 0 mm 4.6 % 6.0 % 5.3 %
Coarse 14 mm 4.4 % 4.1% 4.9 %
1.3 %
1.1%
-
-
Table 3. Surface Saturated Density of RCCA and RCFA (kg/m 3) , Aggregate type Recycled Recycled Recycled Limestone Gravel sand
Crusher Cone Impact Jaw -
Fine _ 5 mm 2387 2385 2448 2627
Coarsel 0 mm 2416 2461 2426 2641 -
Coarsel4 mm 2434 2484 2439 2646 -
774
3.2
Flakiness Index
A particle is considered to be flaky if its thickness is less than 0.6 times the mean sieve size of the size fraction to which it belongs (BS 812, Part 105.1, 1989(4)). BS 882, 1992 (5) limits the content of the flaky particles to less than 40% for crushed aggregate and not less than 50% for natural gravel. This limitation is recommended in order to avoid entrapped water and air lying beneath flaky aggregate since this could lead to a deteriorating effect on the concrete by affecting its workability. Values for the flakiness index of the RCCA are given in Table 4. The results are the mean of 2 tests. The data are markedly different for each type of crusher, but are much lower than the BS 882 limit. It was found that the impact crusher has the lowest flakiness index, producing about 60% and 40% less flaky recycled aggregate than the cone and jaw crushers, respectively. The result for the cone crusher is slightly worrying, especially in relation to acute edges causing a large reduction in the ten-per-cent fines load.
3.3
Angularity Number
British Standard BS 812, Part 1, 1975 (6) defines the angularity number (AI) as 67 minus the percentage of solid volume in a vessel filled with aggregates in a specified manner. The higher the number the more angular is the aggregate and less able to compact. The method is not popularly used, but nevertheless provides a useful indication of the ability for aggregates to compact. The size of coarse aggregate used in these tests was 10-14 mm. The results are the mean of 2 tests. The results given in Table 4 indicate that the shape of the RCA produced by the cone and impact crushers was bordering near the end of the acceptable range, i.e. AI = 9 to 11, whilst the natural limestone and jaw crushed RCA was within the desirable range of AI = 3 to 6. According to Kaplan (7) there is an inverse correlation between AI and the compaction factor (CF), a result which is confirmed in Figure 6 where the CF for the cone crushed RCA replacements is considerable lower than for all other cases. (The implications on mix design, in terms of the required quantity of mixing water to achieve a given compaction for different RCA replacements, is presently under investigation.)
Table 4: Flakiness Index & Angularity Number for RCCA Derived from Different Crushers RCCA obtained from
Impact
Jaw
Cone Limestone BS Limits
9%
15%
21%
7%
_ 40%
9
6
11
3
*
Ten Percent Fine Value %
170
160
110
150
Aggregate Impact Value KN
23
24
25
20
Flakiness Index Angularity Number
9 No BS data, but generally considered to be in the range 0 - 11.
775
3.4 Grading of RCA and Natural Aggregate Grading was carried out according to BS 812:Part 103:1985. The grading of the RCA were compared with BS 882, 1992 as well as with natural coarse and fine aggregates. The grading curves are shown in Figure 3. All RCCA grading complied with BS 882, 1992 for single sized aggregate, although the gradings for the jaw and cone crushed RCA lie closer to the BS limits than the impact crushed RCA and the natural aggregate. However it would possible to easily adjust the crushing and sieving operation to produce better grading if time had permitted. The gradings are extremely concentrated owing to the previous separations. Figure 4 presents the gradings for RCFA and river gravel within the limits of the BS 882, 1992 coarse category. It was found that the grading of the RCFA complied with these limits with one exception, viz. jaw crushed RCA passing the 2.36 mm sieve. The RCFA does not comply with the medium category. As expected the natural river gravel complies with the medium category. The RCFA has a rather coarse grading with only about 20% passing the 600ktm sieve, as opposed to the more usual figure of 30% to 35% in quarried sands. This has a significant effect of mix design as the proportion of fine aggregate required to maintain constant workability would need to be considerably increased, typically by about 20%. The proportion of fines passing the 300~tm sieve is more than desirable, although the dust was removed from the RCFA during the screening process during crushing.
4
Mechanical Properties of RCA
It is well known that it is difficult, and often meaningless, to test the compressive strength of individual particles of aggregate, and that the most common method is to compact aggregates in bulk or use other indirect methods such as the ten percent fines value (TFV) test. Because the RCA in this project was obtained from a parent concrete of known strength, it was considered unnecessary to measure the compressive strength of the RCA. However, because of the varied shape and uncertain effects of the angularity or flakiness of the RCA, a TFV test was carried out. An aggregate impact value (AIV) test was also carried out for completeness. The TFV test was carried out according to BS 812, Part 111, 1990 (8) and AIV test according to BS 812, Part 110, 1990 (9). The tests were carried out on 14 mm coarse RCA size. The results are the mean of 2 tests.
4.1
Ten Percent Fines Value (TFV) of RCCA
Considering the origin of the RCCA, which is identical, this test showed some differences between the crushing machines as shown in Table 4. BS 882:1992 limits the minimum values for TFV when the aggregate used for (a) heavy duty floor 150 kN, (b) wearing surface 100 kN, (c) other uses 50 kN. Based on the assumption that the proportion of RCCA is not likely to exceed 50% replacement, this would qualify the RCCA obtained from impact and jaw crusher to be used in any type of concrete, and exclude the cone crushed RCCA from heavy duty floors. However, it is not known what overall effect mixing RCA with natural aggregates would have on wear.
776
_y, _ _
,
,
103 t
75
75
,~ t
~) I i
-,x..ESl~ . x.,BSI~
oe
~
9,x- BSMEv
-~-InmI
I
i
--~Jaw -m-Ore
25~
9,x-,BSMn
o~
.-~-I~
'
--~JaN e~ o~ ~
25-
w
--~s -.-~-Li~
o • o
~
0,
Se.e236
Se~5
9euelO
9e~14
~]eue5
8e~10
~
14
8eu~2D
Figure 3" 10 mm, 14 mm Single Size Grading Curves for Natural Limestone and RCCA
120 100
-.~-
BS Min
- ~- BS Max
--e--
Im
~ J a w
pact
~ C o n e
--e--Gravel
1| ..
-"
80 oo
.,~ . . . . . . . . . . . .
Sand
60
o
~
~
o
o ~ ~
40 20 0
Sieve
~
I
0.3
Sieve
0.6
Sieve1.18
Sieve 2.36
Figure 4: C-Limit Grading Curves for RCFA and Natural River Gravel
4.2 Aggregate Impact Value (AIV) of RCCA The results shown in Table 4 reveal that the values are almost identical for the RCCA that derived from three different crushing methods, around 24%. The natural limestone value is 20 %. BS 882:1983 limits the maximum values for AIV when the aggregate used for (a) heavy duty floor 25%, (b) wearing surface 30%, and (c) other uses 45%. As with the TFV, assuming a maximum RCCA replacement of 50% all aggregates are suitable for all the above conditions.
777
5
Fresh Concrete Properties
Specimens were manufactured in the laboratory to determine the workability and compressive cube strength of concrete made from natural (called the 'control' mix) and partially replaced RCA. In all cases the following basic mix proportions were used: Ordinary Portland cement type 52.5 N 405 kg/m 3 14 mm Coarse aggregate (SSD) 670 kg/m 3 10 mm Coarse aggregate (SSD) 445 kg/m 3 Fine aggregate (SSD) 710 kg/m 3 Water 170 kg/m 3 Mixing complied with BS 1881, Part 125, 1986 (l~ viz dry mixing for 30 second; 1/3 water added; mixing 2 minutes; standing 10 minutes; cement added; mixing 30 second; remaining water added; mixing 2 minutes. Workability tests were carried out within 5 minutes after mixing. Concrete cubes of 100 mm dimension were made, cured and tested according to BS1881, Part 103, 1991 using a vibrating table to compact the concrete. The replacement proportions of RCA were chosen as 20% and 50%. The former represents a typical limit for RCCA proposed in P.I.T project. The latter was chosen because this is thought to be an extremity worthy of consideration in studying the sensitivity of concrete made with such high replacements, especially for the fine aggregate. The replacement was made for (i) coarse aggregate alone, (ii) fine aggregate alone, and (iii) coarse and fine mixed.
5.1 Workability The methods used were slump according to BS 1881, Part 102, 1991 (ll) and compacting factor (CF) according to BS 1881, Part 103, 1991 (12). Although the slump method is crude, it is an easily understood measurement of workability and therefore used here. The CF method gives a better understanding of workability. The results shown in Figures 5 and 6 are the mean of 3 samples.
5.1.1 Slump Measurement The results are shown in Figure 5. The target slump for the control mix was 60 mm. The general trend is an increase in slump as the replacement RCA increases up to 20%, which is followed by a decrease at 50% replacement. The exception is in the case of RCCA replacement where the slump increases at replacement levels greater than 20%. The effect of the different crushing methods on the coarse and fine RCA is confusing and contradictory. Changes in the slump for RCFA are greatest of all, especially for the cone and jaw crushing methods. RCFA crushed in this manner have greater water absorption (Table 2) so one would have expected a reduction in slump, which is not seen until the replacement is 50%. The effect of grading, in particular the low fraction passing the 600 ~tm sieve, would suggest a reduction in workability as the replacement RCFA increases for a fixed ratio of fine-coarse aggregate. For the RCCA, it is the impact crushing method which sees the greatest change. This confirms the results from Table 4 where the RCCA obtained from the cone crusher gave the greatest angularity and thereibre its effect on workability would be greater as more bleed water might be retained. However, it contradicts the water absorption result (Table 2) where the impact crushed RCCA had the greatest absorption value, suggesting a reduction in slump. It appears that there are many factors influencing these results.
778
150
..................................................................................................................................
125 E
>
4
9 ;""~;
1 O0
9 ;;;
75
~
'" -" "--,At,,
- -
-..
" " " -" ~ " , . . . .
:
S
............
" - - .
.
* ~ ~
E
"~-
50 ~Cone J
25
RCCA al. J a w R C C A -.-~--Impact RCCA - HI- C o n e R C F A - i- Jaw RCFA
"~'~ ~ " ~ .~'.~
- O- I m p a c t R C F A ~ .Cone RCCA & RCFA ~ .Jaw R C C A & R C F A ~ .Impact RCCA & RCFA
"'-
.....=
. "'-
:
.
o j-.................................................................................... + ................................................................................................................................................... o % 20 % 50 %
Figure 5 9 Slump values for concrete with different replacement percentages) of RCA obtained from different crushers 1.000 ..
-O-
.....~176176
~ ~176176176176176176
0.975
0.950
[L
~Cone RCCA "-- Jaw R C C A +Impact RCCA
0.925
'L
- II- Cone R C F A
0.900
- ,L- Jaw R C F A - O - Impact RCFA --m. ,Cone R C C A & R C F A 9--A- ,Jaw R C C A & RCFA - - ~ ,Impact R C C A & RCFA
0.875
0.850
!
I 0.0
20 %
50 %
Figure 6: Compacting factor values for concrete with different replacement percentages of RCA obtained from different crushers
5.1.2 Compacting Factor (CF) The results are shown in figure 6. There was no target CF for the control mix although published literature would suggest that for an aggregate/cement ratio of 6 and a slump of 60 mm, the CF should be about 0.96. The general trend is similar to the slump results. The exception is that there is very little change in the jaw and impact crushed RCCA. The influence of the crushing method appears to be more consistent than in the slump results - the CF for cone crushed RCA is considerably lower in all cases. This is as expected from the angularity tests (Table 4) where the RCCA was found to be angular. The effect of replacement RCFA on the CF factor is as expected, i.e. a reduction in CF from 0.975 to 0.93 (average), owing to increase water absorption and reduced density. The effect of grading, in particular the low fraction passing the 600 I.tm sieve, would suggest a reduction in CF as the replacement RCFA increases. A change in CF from 0.975 to 0.93 is
779 680
--
650
-Z[
,
~ E
620
,-
590
...... "...-
-:-
r i..
560
9
~ 9
~
9
o
530
500
0%
I 20 %
50 %
Figure 7 928 days Compressive strength of concrete with different replacements of RCA derived from different crushers (notations as Figure 6) quite considerable and could be interpreted as a three fold increase in the mixing air content. The results for the RCCA are more encouraging with little change in the CF, with the exception of the cone crushed RCCA which we have already noted as being rather angular. The implications for compacting concrete are therefore less onerous (but still important). The effect of particle size is not conclusive owing to the small variations in grading (Figure 3).
6
Strength of Concrete with Natural Aggregates and RCA
The strength of concrete reported in this paper is compressive cube strength, according to BS 1881, Part 116, 1991 (13). The target strength for the control mix reached about 60 N/mm2; results are shown in Figure 7. In all cases of different types and proportions of replacement RCA the mix using RCA from the cone crusher achieved the highest compressive strength. This contradicts all previous expectations that the cone crusher gives the poorest performance in terms of aggregate properties (e.g. high water absorption, low density, high flakiness) and workability (e.g. lowest CF). It is therefore clear that the compaction of the concrete has an important effect on the final product.
7
Discussion
Different crushing methods have an influence on the properties of recycled concrete aggregate. Physical properties, specifically shape and texture, appears to be effected mostly and showed some variance related to each type of crusher. It was found that one crusher performed well in some properties and shows some disadvantages in others.
780 All crushers produced RCA with acceptable strength and shape (See "Fable 4). However, the impact crusher appears to be the most suitable overall by producing RCA with better shape and strength, then followed by the jaw and then cone crusher. There is no distinctive influence of the crushing methods on the water absorption and density for recycled concrete aggregate (Table 2 and 3). However, concrete made with RCA produced by the cone crusher achieved the greatest compressive cube strengths- several of which exceeded the control mix using natural limestone and river gravel sand. Comparing the coarse RCCA with natural limestone aggregate, the RCA showed similar (or even better) properties especially for shape and strength. However the RCCA will absorb about 4 to 6 times more water than the natural limestone aggregate. All RCCA appear to have a reasonable grading comparing with the British Standard and is similar to that for natural limestone aggregate. However, it should be noted that the impact crusher has one major disadvantage, which is that it has a large reduction factor (from the feed to the output) and consequently it produces large amount of fine aggregate than coarse aggregate. This agrees with Boesman's findings (14). The fine RCFA was considerably coarser than the natural river gravel, and technically did not comply with the BS coarse category limits, failing at the 2.36 mm sieve size only. The effect of a smaller fraction of RCFA below 600 ~tm may have a significant effect on the desired mix proportions to keep workability constant. In spite of the generally poor characteristics of the RCFA in terms of porosity and grading, the effect on the workability and strength of the resulting concrete was, in 8 out of 9 cases, not deleterious, even at 50% replacement.
8
Conclusion
Waste concrete from mature prestressed concrete hollow cored floor units, produced by proprietary extrusion techniques, and having compressive strengths of about 60 N/mm 2 have been recycled for use as replacement aggregate in high strength concrete. The waste concrete was crushed to -14 mm using three different types of crushers - the cone, impact and jaw crushers. The recycled material was separated into fractions of 14 mm, 10 mm a n d - 5 mm, and tested for physical and mechanical properties relevant to use in concrete. Concrete was then made using zero, 20% and 50% replacement of recycled coarse (RCCA), recycled fine (RCFA) and mixed (RCCA+RCFA) aggregates. The control mix was made using natural limestone coarse and river gravel fine aggregates. The concrete was tested for slump, compaction factor and compressive cube strength. The following may be concluded: 9
9 9
All three crushers produced acceptable shape and strength of RCCA. Some properties are competitive to that of natural limestone aggregate except for water absorption, which is 4 times greater. RCFA was much coarser than river gravel and just complied with the British Standard coarse grading limits. Its water absorption is 3 times greater than the gravel. The impact crusher performed best with regard to most aggregate properties, e.g. flakiness, strength and water absorption, but has a disadvantage in producing a large amount of fine-to-coarse RCA.
781 9
Concerning shape and strength, RCA showed similar properties, and in some cases better, than the conventional limestone aggregate. 9 The slump value of fresh concrete made with RCA varied widely depending on the percentage and type of replacement, and the type of crusher, a fact which may be linked to the angularity of the RCCA. 9 The compaction factor of fresh concrete made with RCA was more consistent and logical, and showed the problems encountered with using angular RCCA produced by the cone crusher. 9 Compressive strength of concrete made with RCA were generally within +5 N/mm 2 of the control value of 62 N/mm 2.
References: 1. Building Research Establishment, UK., Protocol for the Use of Reclaimed Product in Precast Concrete: Second Draft 29 March 1999. 2. Quarrying Reference : 9 Laboratory based jaw crusher, School of Civil Engineering, Nottingham University 9 10/07 Impact crusher, Webfell Ground Engineering ltd, Normanton Industrial Estate, Wakefield, UK. 9 Private owned adapted cone crusher, Mr Nigel Gill, 3. British Standards Institution, London, BS 812, Part 2, 1995. 4. British Standards Institution, London, BS 812, Part 105.1, 1989. 5. British Standards Institution, London, BS 882, 1992. 6. British Standards Institution, London, BS 812, Part 1, 1975. 7. Kaplan, M. F., Flexural and compressive strength of concrete as affected by the properties of coarse aggregate, J Amer. Concr. Inst., 55, 1959, pp 1193-208. 8. British Standards Institution, London, BS 812, Part 111, 1990 9. British Standards Institution, London, BS 812, Part 110, 1990 10. British Standards Institution, London, BS 1881, Part 125, 1986 11. British Standards Institution, London, BS 1881, Part 102, 1991. 12. British Standards Institution, London, BS. 1881, Part 103, 1991. 13. British Standards Institution, London, BS 1881, Part 116, 1991. 14. Boesman, B., Crushing and separating techniques for demolition material. EDA/Rilem Demo-recycling conference. Re-Use of Concrete and Brick Materials, European Demotion Association, Den Haag, The Netherlands. Proc. Vol. 2.1985
Acknowledgement. The authors wish to thank the Altajir World of Islam Trust and University of Nottingham Overseas Research Fund for financial support of this project. They are grateful to the technicians in the Civil Engineering Laboratory at Nottingham University, to Richard Lees Ltd. for the supply of materials, and to Webfell Ltd. and to Mr. Nigel Gill for the use of their crushing machines.
769
The Properties of Recycled Precast Concrete Hollow Core Slabs for Use as Replacement Aggregate in Concrete B E Marmash and K S Elliott School of Civil Engineering University of Nottingham Nottingham NG7 2RD United Kingdom.
Summary Waste concrete from prestressed precast hollow cored floor units has been recycled for use as replacement aggregate in concrete (RCA). Waste concrete blocks were crushed t o - 1 4 mm using cone, impact and jaw crushers. All produced acceptable physical and mechanical properties, although the impact crusher was best suited in most cases. The water absorption of the RCA was 3 to 4 times greater than natural limestone and river gravel used for control purposes. The fine RCA was at the coarse end of the British Standard limit. Concrete made using zero, 20% and 50% replacement of coarse and fine RCA had increased workability at 20% replacement, but this reduced considerably at 50% replacement. Compressive strength of concrete made with RCA was generally within +5 N/mm 2 of the control value of 62 N/mm 2. RCA from the cone crusher produced the highest strengths. The main conclusion is that concrete made with up to 50% replacement of both coarse and fine RCA appears to be comparable with natural aggregate concrete in terms of workability and compressive strength.
1
Introduction
The precast concrete hollow core floor industry produces a considerable amount of wasted concrete elements, due mainly to the manufacturing processes and in part to natural wastage at the ends of the casting beds. Figure 1 shows how prestressed hollow core units (hcu) are manufactured by extrusion or slip-forming through a machine on long beds, typically 100 m in length x 1.2 m wide. Waste material is made at the beginning and end of each bed, typically 0.3 to 0.4 m 3 per casting. After detensioning the units are cut to length. Waste material is therefore made if the cumulative length of the units does not equal the net cast length. This can be between zero and 0.5 m 3 per casting. The total waste generated in the UK is around 5% of the production. Waste material from hcu is high grade and uncontaminated material. The parent concrete is hard and of compressive strength between 50 to 80 N/mm 2. It is manufactured from Portland cement, and from clean and reliable sources of 10 mm to 14 mm limestone or gravel. The grading of the coarse and fine aggregates is carefully controlled, and together with a
770
Figure 1. End line waste of hollow core slab units water cement ratio of around 0.3 the resulting concrete is of a high density and low porosity. It is also extremely brittle and will fracture into flaky shapes with acute edges. In commercial crushers large amounts of fine aggregate are produced, which cause concern for the reintroduction to hcu manufacture. To date recycled coarse aggregate (RCCA) has been considered at replacement levels of up to 20% while the recycled fine aggregate (RCFA) only up to 10%. A recent project (P.I.T) t~ conducted by the member companies of the UK's Precast Flooring Federation has found that the effects of using 20% of RCCA + 10% of RCFA in precast concrete are small. They indicated that the differences recorded in compressive and flexural strength were mostly less than the variability of the test. Also they reported that shrinkage and creep measurements were more consistently affected but still fairly small and should not contravene any specification requirements. Differences resulting from the crushing method were not investigated. To complement this work, this present research aims to study the properties of recycled concrete from hcus crushed using three different methods - cone, jaw and impact crushers. Commercial hcus, manufactured by Richard Lees Ltd. using the Spiroll extrusion technique, were obtained for this study. The mix content for the parent concrete is given in Table 1. The resulting crushed material was first separated into coarse (10 mm and 14 mm) and fine (< 5 mm) fractions. It was tested for those properties which are important to the reintroduction of RCA into hcu production, namely grading, water absorption, density, shape, ten-per-cent fines value, workability and strength. The recycled aggregates were also tested in concrete mixes in which the aggregates in a 'reference' mix were substituted with varying Table 1: Mix content for the parent concrete (hcus) 14 mm
10 mm
Sand
Cement Class 52.5N
Pozzolan
Water
340 Kg
440 Kg
500 Kg
200 Kg
60 Kg
50 Kg
771 proportions of recycled concrete aggregate. The chosen percentage replacement was 20% and 50% - the former represents a typical limit for RCCA proposed in P.I.T project, and the latter is made deliberately large in order to investigate the sensitivity of the said mechanical and physical properti?s. This paper reviews the determination of mechanical and physical properties of RCA alone. The results are compared for compliance with the relevant British Standards. The paper also presents the results of tests carried out on fresh and hardened concrete made with natural aggregates and RCA.
2
Crushers
To enable discussion on the properties of RCA obtained from the crushing operation, it is first necessary to review this in relation to crushing hardened concrete. It should be noted that most crushing machines were developed for crushing rock into sizes with equivalent diameters of 50 mm or greater. In some cases special modifications to the orifice had to be made to produce RCA of 14 mm size.
2.1
Cone Crusher
Cone crushers are one of the major categories of gyrating crushers, which have developed into being one of the most important types of machine in use in quarrying (2). The principle is shown in Figure 2. It uses a repeated compression action with fixed and moving crushing members. The long stroke and high speed agitate the feed in its passage through the chamber. At the lower extremity of the Crushing chamber the faces of the two crushing member are so shaped that they are parallel for a section, resulting in the larger pieces being assured of having at least one dimension equal to, or less than, the setting, quoted as closedside setting (CSS) when properly fed and certainly all products less than twice the CSS. In the operation to crush hcu, the size of the feed was up to 350 mm and the CSS was set at about 14 mm. The resulting RCA was acceptable in appearance and seems to be slightly towards flaky and elongated in shape. There is no distinguished dust on the coarse particles. The amount of cement paste attached to the virgin aggregate was quite large.
2.2 Jaw Crusher Jaw crushers are also commonly used in quarries. Its principle could be summarized in that the feed is subjected to repeated pressure as it passes downwards and is progressively reduced in size until it is eventually small enough to pass out of the crushing chamber. See Figure 2. The size of machine also affects its speed which decreases as the crusher size increases. The angle between the crushing faces is normally between 19~ and 22 ~ This angle is set to allow the crushing force to be transmitted to very hard materials without a tendency for the feed to rise itself out of the crushing zone and so cause abrasive wear to the liners. The setting is usually measured as a CSS, i.e. when the jaws are at their closest position, and some times as open side setting with the jaws at their greatest distance apart. In the operation to crush hcu, the size of the feed was nominally 25mm and the CSS was set at about 15mm. The resulting RCA was good in appearance and seems to be slightly elongated with no distinguish dust on the coarse particles. The amount of cement paste attached to the virgin aggregate was quite large.
772
Figure 2: Cone Crusher (left), Impact crusher (middle) and Jaw Crusher (right).
2.3 Impact crusher Impact crushing could be described as impact breaking since the feed is fragmented by kinetic energy introduced by a rotating mass (the rotor) which projects the material against a fixed surface causing it to shatter causing further particles size reduction. The process causes the material to break along its natural cleavage planes and this yields a good product shape free from stress. The rotor speed is fundamental to the breaking process where the higher speed the higher reduction factor; the size of the material fed into the crusher to the size of the finished product. In the operation to crush hcu, the size of the feed was nominally 500mm. The opening size was 700mm x 500mm with 1800 RPM rotation. The resulting RCA was less elongated towards rounded shape with acute edges. There was a distinguish dust on the coarse particles. The amount of cement paste attached to the virgin aggregate also was quite large.
2.4 Summary of Crushing Operations 9 9 9
3
In visual appearance the RCA produced from the three crushers seems to be relativly the same in apperance but slightly better shape could be seen from impact crusher. The amount of fines obtained from the crushers was cearly high for impact crusher then followed by cone and then jaw. The speed at which the quantity of RCA was produced by the different crushers was convenient. It is qiute high for the imapct crusher while the cone and jaw were relatively the same but slower than the impact.
Physical properties of Recycled Concrete Aggregate (RCA)
3.1 Density and Water Absorption The presence of internal pores in the crushed particles has an influence on the porosity and absorption properties of RCA. These properties have a major effect on the workability and durability of concrete with a low water-cement ratio, especially used in hcu production. They also have an influence on the bond to hydrated cement past as well as the concrete resistance to freezing and thawing, and to a lesser extent carbonation. Water absorption and density were measured (to BS 812, Part 2, 1995 (3)) for the RCA and natural crushed
773 carboniferous limestone (obtained from Tarmac Quarry P r o d u c t s - Retford ). The results given in Tables 2 and 3 are the mean of 2 samples.
3.1.1
Density and Water Absorption of Coarse RCCA
It was expected for the RCCA to have a higher water absorption and lower density than the natural limestone because of the cement mortar attached to it. The results given in Table 2 show that the water absorption for 10 mm and 14 mm RCCA is around four times greater than that of the limestone with similar size. Table 3 shows that the surface saturated density (SSD) is between 4% and 7% lower than that of natural limestone with similar size. Conceming the effects of different crushing methods on water absorption and density, there were no distinctive differences nor trends. This might give an indication to that there is no significant difference on the amount of cement mortar attached to the RCA obtained from the three different crushing machines. In other words, these crushers (if the fed materials are identical) could produce RCCA with approximate close percentages of attached mortar to the aggregate particles. (This could not be considered as a definite conclusion as other methods need to be used to measure accurately the amount of attached mortar to the RCA.)
3.1.2 Density and Water Absorption of Fine RCFA The water absorption tbr RCFA is higher than that for natural river gravel sand. This is also due to the attached mortar. The results given in Table 2 show that the water absorption for RCFA is around three times greater than that of the natural gravel sand of similar maximum size and grading profile. Table 3 shows that the surface saturated density (SSD) is 9% lower than that of natural gravel sand of similar size. There is no significant difference between the values of water absorption for RCFA obtained from the different crushing methods. This also probably limits the influence of the crushing machines on the amount of cement mortar attached to RCFA.
Table 2. Water absorption of RCCA and RCFA Aggregate t~cpe Recycled Recycled Recycled Limestone Gravel sand
Crusher Cone Impact Jaw -
Fine < 5 mm 6.8 % 5.8 % 6.7 % 1.7 %
Coarsel 0 mm 4.6 % 6.0 % 5.3 %
Coarse 14 mm 4.4 % 4.1% 4.9 %
1.3 %
1.1%
-
-
Table 3. Surface Saturated Density of RCCA and RCFA (kg/m 3) , Aggregate type Recycled Recycled Recycled Limestone Gravel sand
Crusher Cone Impact Jaw -
Fine _ 5 mm 2387 2385 2448 2627
Coarsel 0 mm 2416 2461 2426 2641 -
Coarsel4 mm 2434 2484 2439 2646 -
774
3.2
Flakiness Index
A particle is considered to be flaky if its thickness is less than 0.6 times the mean sieve size of the size fraction to which it belongs (BS 812, Part 105.1, 1989(4)). BS 882, 1992 (5) limits the content of the flaky particles to less than 40% for crushed aggregate and not less than 50% for natural gravel. This limitation is recommended in order to avoid entrapped water and air lying beneath flaky aggregate since this could lead to a deteriorating effect on the concrete by affecting its workability. Values for the flakiness index of the RCCA are given in Table 4. The results are the mean of 2 tests. The data are markedly different for each type of crusher, but are much lower than the BS 882 limit. It was found that the impact crusher has the lowest flakiness index, producing about 60% and 40% less flaky recycled aggregate than the cone and jaw crushers, respectively. The result for the cone crusher is slightly worrying, especially in relation to acute edges causing a large reduction in the ten-per-cent fines load.
3.3
Angularity Number
British Standard BS 812, Part 1, 1975 (6) defines the angularity number (AI) as 67 minus the percentage of solid volume in a vessel filled with aggregates in a specified manner. The higher the number the more angular is the aggregate and less able to compact. The method is not popularly used, but nevertheless provides a useful indication of the ability for aggregates to compact. The size of coarse aggregate used in these tests was 10-14 mm. The results are the mean of 2 tests. The results given in Table 4 indicate that the shape of the RCA produced by the cone and impact crushers was bordering near the end of the acceptable range, i.e. AI = 9 to 11, whilst the natural limestone and jaw crushed RCA was within the desirable range of AI = 3 to 6. According to Kaplan (7) there is an inverse correlation between AI and the compaction factor (CF), a result which is confirmed in Figure 6 where the CF for the cone crushed RCA replacements is considerable lower than for all other cases. (The implications on mix design, in terms of the required quantity of mixing water to achieve a given compaction for different RCA replacements, is presently under investigation.)
Table 4: Flakiness Index & Angularity Number for RCCA Derived from Different Crushers RCCA obtained from
Impact
Jaw
Cone Limestone BS Limits
9%
15%
21%
7%
_ 40%
9
6
11
3
*
Ten Percent Fine Value %
170
160
110
150
Aggregate Impact Value KN
23
24
25
20
Flakiness Index Angularity Number
9 No BS data, but generally considered to be in the range 0 - 11.
775
3.4 Grading of RCA and Natural Aggregate Grading was carried out according to BS 812:Part 103:1985. The grading of the RCA were compared with BS 882, 1992 as well as with natural coarse and fine aggregates. The grading curves are shown in Figure 3. All RCCA grading complied with BS 882, 1992 for single sized aggregate, although the gradings for the jaw and cone crushed RCA lie closer to the BS limits than the impact crushed RCA and the natural aggregate. However it would possible to easily adjust the crushing and sieving operation to produce better grading if time had permitted. The gradings are extremely concentrated owing to the previous separations. Figure 4 presents the gradings for RCFA and river gravel within the limits of the BS 882, 1992 coarse category. It was found that the grading of the RCFA complied with these limits with one exception, viz. jaw crushed RCA passing the 2.36 mm sieve. The RCFA does not comply with the medium category. As expected the natural river gravel complies with the medium category. The RCFA has a rather coarse grading with only about 20% passing the 600ktm sieve, as opposed to the more usual figure of 30% to 35% in quarried sands. This has a significant effect of mix design as the proportion of fine aggregate required to maintain constant workability would need to be considerably increased, typically by about 20%. The proportion of fines passing the 300~tm sieve is more than desirable, although the dust was removed from the RCFA during the screening process during crushing.
4
Mechanical Properties of RCA
It is well known that it is difficult, and often meaningless, to test the compressive strength of individual particles of aggregate, and that the most common method is to compact aggregates in bulk or use other indirect methods such as the ten percent fines value (TFV) test. Because the RCA in this project was obtained from a parent concrete of known strength, it was considered unnecessary to measure the compressive strength of the RCA. However, because of the varied shape and uncertain effects of the angularity or flakiness of the RCA, a TFV test was carried out. An aggregate impact value (AIV) test was also carried out for completeness. The TFV test was carried out according to BS 812, Part 111, 1990 (8) and AIV test according to BS 812, Part 110, 1990 (9). The tests were carried out on 14 mm coarse RCA size. The results are the mean of 2 tests.
4.1
Ten Percent Fines Value (TFV) of RCCA
Considering the origin of the RCCA, which is identical, this test showed some differences between the crushing machines as shown in Table 4. BS 882:1992 limits the minimum values for TFV when the aggregate used for (a) heavy duty floor 150 kN, (b) wearing surface 100 kN, (c) other uses 50 kN. Based on the assumption that the proportion of RCCA is not likely to exceed 50% replacement, this would qualify the RCCA obtained from impact and jaw crusher to be used in any type of concrete, and exclude the cone crushed RCCA from heavy duty floors. However, it is not known what overall effect mixing RCA with natural aggregates would have on wear.
776
_y, _ _
,
,
103 t
75
75
,~ t
~) I i
-,x..ESl~ . x.,BSI~
oe
~
9,x- BSMEv
-~-InmI
I
i
--~Jaw -m-Ore
25~
9,x-,BSMn
o~
.-~-I~
'
--~JaN e~ o~ ~
25-
w
--~s -.-~-Li~
o • o
~
0,
Se.e236
Se~5
9euelO
9e~14
~]eue5
8e~10
~
14
8eu~2D
Figure 3" 10 mm, 14 mm Single Size Grading Curves for Natural Limestone and RCCA
120 100
-.~-
BS Min
- ~- BS Max
--e--
Im
~ J a w
pact
~ C o n e
--e--Gravel
1| ..
-"
80 oo
.,~ . . . . . . . . . . . .
Sand
60
o
~
~
o
o ~ ~
40 20 0
Sieve
~
I
0.3
Sieve
0.6
Sieve1.18
Sieve 2.36
Figure 4: C-Limit Grading Curves for RCFA and Natural River Gravel
4.2 Aggregate Impact Value (AIV) of RCCA The results shown in Table 4 reveal that the values are almost identical for the RCCA that derived from three different crushing methods, around 24%. The natural limestone value is 20 %. BS 882:1983 limits the maximum values for AIV when the aggregate used for (a) heavy duty floor 25%, (b) wearing surface 30%, and (c) other uses 45%. As with the TFV, assuming a maximum RCCA replacement of 50% all aggregates are suitable for all the above conditions.
777
5
Fresh Concrete Properties
Specimens were manufactured in the laboratory to determine the workability and compressive cube strength of concrete made from natural (called the 'control' mix) and partially replaced RCA. In all cases the following basic mix proportions were used: Ordinary Portland cement type 52.5 N 405 kg/m 3 14 mm Coarse aggregate (SSD) 670 kg/m 3 10 mm Coarse aggregate (SSD) 445 kg/m 3 Fine aggregate (SSD) 710 kg/m 3 Water 170 kg/m 3 Mixing complied with BS 1881, Part 125, 1986 (l~ viz dry mixing for 30 second; 1/3 water added; mixing 2 minutes; standing 10 minutes; cement added; mixing 30 second; remaining water added; mixing 2 minutes. Workability tests were carried out within 5 minutes after mixing. Concrete cubes of 100 mm dimension were made, cured and tested according to BS1881, Part 103, 1991 using a vibrating table to compact the concrete. The replacement proportions of RCA were chosen as 20% and 50%. The former represents a typical limit for RCCA proposed in P.I.T project. The latter was chosen because this is thought to be an extremity worthy of consideration in studying the sensitivity of concrete made with such high replacements, especially for the fine aggregate. The replacement was made for (i) coarse aggregate alone, (ii) fine aggregate alone, and (iii) coarse and fine mixed.
5.1 Workability The methods used were slump according to BS 1881, Part 102, 1991 (ll) and compacting factor (CF) according to BS 1881, Part 103, 1991 (12). Although the slump method is crude, it is an easily understood measurement of workability and therefore used here. The CF method gives a better understanding of workability. The results shown in Figures 5 and 6 are the mean of 3 samples.
5.1.1 Slump Measurement The results are shown in Figure 5. The target slump for the control mix was 60 mm. The general trend is an increase in slump as the replacement RCA increases up to 20%, which is followed by a decrease at 50% replacement. The exception is in the case of RCCA replacement where the slump increases at replacement levels greater than 20%. The effect of the different crushing methods on the coarse and fine RCA is confusing and contradictory. Changes in the slump for RCFA are greatest of all, especially for the cone and jaw crushing methods. RCFA crushed in this manner have greater water absorption (Table 2) so one would have expected a reduction in slump, which is not seen until the replacement is 50%. The effect of grading, in particular the low fraction passing the 600 ~tm sieve, would suggest a reduction in workability as the replacement RCFA increases for a fixed ratio of fine-coarse aggregate. For the RCCA, it is the impact crushing method which sees the greatest change. This confirms the results from Table 4 where the RCCA obtained from the cone crusher gave the greatest angularity and thereibre its effect on workability would be greater as more bleed water might be retained. However, it contradicts the water absorption result (Table 2) where the impact crushed RCCA had the greatest absorption value, suggesting a reduction in slump. It appears that there are many factors influencing these results.
778
150
..................................................................................................................................
125 E
>
4
9 ;""~;
1 O0
9 ;;;
75
~
'" -" "--,At,,
- -
-..
" " " -" ~ " , . . . .
:
S
............
" - - .
.
* ~ ~
E
"~-
50 ~Cone J
25
RCCA al. J a w R C C A -.-~--Impact RCCA - HI- C o n e R C F A - i- Jaw RCFA
"~'~ ~ " ~ .~'.~
- O- I m p a c t R C F A ~ .Cone RCCA & RCFA ~ .Jaw R C C A & R C F A ~ .Impact RCCA & RCFA
"'-
.....=
. "'-
:
.
o j-.................................................................................... + ................................................................................................................................................... o % 20 % 50 %
Figure 5 9 Slump values for concrete with different replacement percentages) of RCA obtained from different crushers 1.000 ..
-O-
.....~176176
~ ~176176176176176176
0.975
0.950
[L
~Cone RCCA "-- Jaw R C C A +Impact RCCA
0.925
'L
- II- Cone R C F A
0.900
- ,L- Jaw R C F A - O - Impact RCFA --m. ,Cone R C C A & R C F A 9--A- ,Jaw R C C A & RCFA - - ~ ,Impact R C C A & RCFA
0.875
0.850
!
I 0.0
20 %
50 %
Figure 6: Compacting factor values for concrete with different replacement percentages of RCA obtained from different crushers
5.1.2 Compacting Factor (CF) The results are shown in figure 6. There was no target CF for the control mix although published literature would suggest that for an aggregate/cement ratio of 6 and a slump of 60 mm, the CF should be about 0.96. The general trend is similar to the slump results. The exception is that there is very little change in the jaw and impact crushed RCCA. The influence of the crushing method appears to be more consistent than in the slump results - the CF for cone crushed RCA is considerably lower in all cases. This is as expected from the angularity tests (Table 4) where the RCCA was found to be angular. The effect of replacement RCFA on the CF factor is as expected, i.e. a reduction in CF from 0.975 to 0.93 (average), owing to increase water absorption and reduced density. The effect of grading, in particular the low fraction passing the 600 I.tm sieve, would suggest a reduction in CF as the replacement RCFA increases. A change in CF from 0.975 to 0.93 is
779 680
--
650
-Z[
,
~ E
620
,-
590
...... "...-
-:-
r i..
560
9
~ 9
~
9
o
530
500
0%
I 20 %
50 %
Figure 7 928 days Compressive strength of concrete with different replacements of RCA derived from different crushers (notations as Figure 6) quite considerable and could be interpreted as a three fold increase in the mixing air content. The results for the RCCA are more encouraging with little change in the CF, with the exception of the cone crushed RCCA which we have already noted as being rather angular. The implications for compacting concrete are therefore less onerous (but still important). The effect of particle size is not conclusive owing to the small variations in grading (Figure 3).
6
Strength of Concrete with Natural Aggregates and RCA
The strength of concrete reported in this paper is compressive cube strength, according to BS 1881, Part 116, 1991 (13). The target strength for the control mix reached about 60 N/mm2; results are shown in Figure 7. In all cases of different types and proportions of replacement RCA the mix using RCA from the cone crusher achieved the highest compressive strength. This contradicts all previous expectations that the cone crusher gives the poorest performance in terms of aggregate properties (e.g. high water absorption, low density, high flakiness) and workability (e.g. lowest CF). It is therefore clear that the compaction of the concrete has an important effect on the final product.
7
Discussion
Different crushing methods have an influence on the properties of recycled concrete aggregate. Physical properties, specifically shape and texture, appears to be effected mostly and showed some variance related to each type of crusher. It was found that one crusher performed well in some properties and shows some disadvantages in others.
780 All crushers produced RCA with acceptable strength and shape (See "Fable 4). However, the impact crusher appears to be the most suitable overall by producing RCA with better shape and strength, then followed by the jaw and then cone crusher. There is no distinctive influence of the crushing methods on the water absorption and density for recycled concrete aggregate (Table 2 and 3). However, concrete made with RCA produced by the cone crusher achieved the greatest compressive cube strengths- several of which exceeded the control mix using natural limestone and river gravel sand. Comparing the coarse RCCA with natural limestone aggregate, the RCA showed similar (or even better) properties especially for shape and strength. However the RCCA will absorb about 4 to 6 times more water than the natural limestone aggregate. All RCCA appear to have a reasonable grading comparing with the British Standard and is similar to that for natural limestone aggregate. However, it should be noted that the impact crusher has one major disadvantage, which is that it has a large reduction factor (from the feed to the output) and consequently it produces large amount of fine aggregate than coarse aggregate. This agrees with Boesman's findings (14). The fine RCFA was considerably coarser than the natural river gravel, and technically did not comply with the BS coarse category limits, failing at the 2.36 mm sieve size only. The effect of a smaller fraction of RCFA below 600 ~tm may have a significant effect on the desired mix proportions to keep workability constant. In spite of the generally poor characteristics of the RCFA in terms of porosity and grading, the effect on the workability and strength of the resulting concrete was, in 8 out of 9 cases, not deleterious, even at 50% replacement.
8
Conclusion
Waste concrete from mature prestressed concrete hollow cored floor units, produced by proprietary extrusion techniques, and having compressive strengths of about 60 N/mm 2 have been recycled for use as replacement aggregate in high strength concrete. The waste concrete was crushed to -14 mm using three different types of crushers - the cone, impact and jaw crushers. The recycled material was separated into fractions of 14 mm, 10 mm a n d - 5 mm, and tested for physical and mechanical properties relevant to use in concrete. Concrete was then made using zero, 20% and 50% replacement of recycled coarse (RCCA), recycled fine (RCFA) and mixed (RCCA+RCFA) aggregates. The control mix was made using natural limestone coarse and river gravel fine aggregates. The concrete was tested for slump, compaction factor and compressive cube strength. The following may be concluded: 9
9 9
All three crushers produced acceptable shape and strength of RCCA. Some properties are competitive to that of natural limestone aggregate except for water absorption, which is 4 times greater. RCFA was much coarser than river gravel and just complied with the British Standard coarse grading limits. Its water absorption is 3 times greater than the gravel. The impact crusher performed best with regard to most aggregate properties, e.g. flakiness, strength and water absorption, but has a disadvantage in producing a large amount of fine-to-coarse RCA.
781 9
Concerning shape and strength, RCA showed similar properties, and in some cases better, than the conventional limestone aggregate. 9 The slump value of fresh concrete made with RCA varied widely depending on the percentage and type of replacement, and the type of crusher, a fact which may be linked to the angularity of the RCCA. 9 The compaction factor of fresh concrete made with RCA was more consistent and logical, and showed the problems encountered with using angular RCCA produced by the cone crusher. 9 Compressive strength of concrete made with RCA were generally within +5 N/mm 2 of the control value of 62 N/mm 2.
References: 1. Building Research Establishment, UK., Protocol for the Use of Reclaimed Product in Precast Concrete: Second Draft 29 March 1999. 2. Quarrying Reference : 9 Laboratory based jaw crusher, School of Civil Engineering, Nottingham University 9 10/07 Impact crusher, Webfell Ground Engineering ltd, Normanton Industrial Estate, Wakefield, UK. 9 Private owned adapted cone crusher, Mr Nigel Gill, 3. British Standards Institution, London, BS 812, Part 2, 1995. 4. British Standards Institution, London, BS 812, Part 105.1, 1989. 5. British Standards Institution, London, BS 882, 1992. 6. British Standards Institution, London, BS 812, Part 1, 1975. 7. Kaplan, M. F., Flexural and compressive strength of concrete as affected by the properties of coarse aggregate, J Amer. Concr. Inst., 55, 1959, pp 1193-208. 8. British Standards Institution, London, BS 812, Part 111, 1990 9. British Standards Institution, London, BS 812, Part 110, 1990 10. British Standards Institution, London, BS 1881, Part 125, 1986 11. British Standards Institution, London, BS 1881, Part 102, 1991. 12. British Standards Institution, London, BS. 1881, Part 103, 1991. 13. British Standards Institution, London, BS 1881, Part 116, 1991. 14. Boesman, B., Crushing and separating techniques for demolition material. EDA/Rilem Demo-recycling conference. Re-Use of Concrete and Brick Materials, European Demotion Association, Den Haag, The Netherlands. Proc. Vol. 2.1985
Acknowledgement. The authors wish to thank the Altajir World of Islam Trust and University of Nottingham Overseas Research Fund for financial support of this project. They are grateful to the technicians in the Civil Engineering Laboratory at Nottingham University, to Richard Lees Ltd. for the supply of materials, and to Webfell Ltd. and to Mr. Nigel Gill for the use of their crushing machines.