Illinois, Michigan, Indiana, and Ohio. Out of these sources, 15 were construction and
demolition recycling plants. Since these facilities take debris from many sources (e.g.,
sidewalks, pavements, buildings, bridges), the properties of the original concrete are
not known. Furthermore, the RCA from the recycling plants had gradations typical
for road sub-base use in each respective state (including significant fine fraction), and
also contained varying amounts of deleterious substances (e.g., brick, asphalt, wood).
The other source of RCA was a 0.4-scale precast concrete shear wall laboratory
specimen previously tested to failure under reversed-cyclic lateral loading (Smith et
al. 2011). After the testing of the wall, the precast panels were crushed at a mobile
recycling facility, and then sieved and combined into two different gradations (RCAND and RCA-ND1 in Table 1). The reason for crushing a test specimen into RCA
was the availability of detailed information on the mix design as well as the fresh and
hardened properties of the source concrete. In addition, since the RCA from the test
specimen was not mixed with other material (i.e., it contained one concrete mix and
little deleterious substances), it served as a benchmark for the RCA from the other 15
sources.
Table 1. Natural and recycled concrete aggregate properties
Specific Gravity
Aggregate
ID
Type/
Source Location
FA
Concrete Sand
NA-PG Pea Gravel
NA-CL Limestone
RCA-RR 1. Mishawaka, IN
RCA-R 2. Chicago, IL
RCA-W 3. Indianapolis, IN
RCA-P 4. Chicago, IL
RCA-GT 5. South Bend, IN
RCA-L 6. Chicago, IL
RCA-A 7. Cleves, OH
RCA-EL 8. Cincinnati, OH
RCA-EG 9. Elk Grove, IL
RCA-HP 10. Highland Park, MI
RCA-ND 11. Notre Dame, IN
RCA-ND1 11. Notre Dame, IN
RCA-S 12. South Bend, IN
RCA-T 13. Taylor, MI
RCA-SB 14. South Bend, IN
RCA-E 15. Elkhart, IN
RCA-G 16. Goshen, IN
Bulk
SSD App.
Dry
2.59
2.47
2.71
2.31
2.27
2.26
2.33
2.29
2.34
2.32
2.31
2.33
2.28
2.28
2.30
2.18
2.24
2.13
2.14
2.1
2.63
2.55
2.73
2.41
2.37
2.36
2.45
2.41
2.46
2.44
2.43
2.46
2.41
2.41
2.43
2.32
2.38
2.28
2.32
2.29
2.69
2.70
2.76
2.56
2.53
2.52
2.64
2.61
2.67
2.64
2.65
2.68
2.62
2.63
2.66
2.52
2.61
2.49
2.61
2.59
Table 1 shows the type/source location, specific gravity [bulk dry, saturated
surface dry (SSD), and apparent], water absorption (!!" for NA, !!"# for RCA),
residual mortar content (RM), deleterious material content (!!"# ), gradation, and LA
abrasion loss (ASTM 2009a) of the aggregates (with the RCAs listed in order of
increasing absorption). Since significant variations in aggregate size can produce
different fresh and hardened concrete properties (e.g., workability, strength)
3
(Kosmatka et al. 2002), it was necessary to use aggregates with a consistent gradation.
The three target gradations used were ASTM C33 (2009a) No. 8, ASTM C33 No. 7,
and INDOT (2012) No. 8 [corresponding to nominal maximum aggregate sizes of 9.5
mm (3/8 in.), 12.5 mm (1/2 in.), and 19.0 mm (3/4 in.), respectively]. To achieve
these gradations, each coarse aggregate was sieved into individual size fractions and
then recombined as necessary. The details of the specific gravity, absorption, residual
mortar content, and deleterious content testing are described in Knaack and Kurama
(2011). Lastly, the LA abrasion loss, which measures the change in aggregate mass
due to mechanical degradation through tumbling and falling, was determined
according to ASTM C 131 (2009a).
NA Concrete Mix Designs. As listed in Table 2, three target virgin natural aggregate
concrete (NAC) mix designs formed the basis for the RCA concrete mixes in this
research. The original NAC mix (Smith et al. 2011) for the laboratory wall test
specimen (i.e., the
source for RCA- Table 2. Dry weight proportions of NAC target mix designs
Water Cement NA
FA
HRWR
AEA
ND and RCAName
(lb/yd3) (lb/yd3) (lb/yd3) (lb/yd3) (fl oz/yd3) (fl oz/yd3)
ND1) served as NA-PG Target 253
574
1733
1140
48.5
8.1
one of the target NA-CL Target 253
574
1916
1140
48.5
8.1
mixes. This mix
NA-PG-Wet
324
726
1249
1308
21.8
Target
used NA-PG as the
Note: 1 lb/yd3 = 0.5933 kg/m3; 1 fl oz/yd3 = 38.67 mL/m3
coarse aggregate
and was designed with a water-to-cement (w/c) ratio of 0.44 for a target 28-day
strength of 41.4 MPa (6.0 ksi), slump of 12.72.54 cm (51 in.), and air content of
51.5%. From previous data (Smith et al. 2011), the average measured 28-day
strength of the concrete was 44.1 MPa (6.4 ksi). Because the original mix was
designed for precast concrete construction (which requires high early concrete
strength for removal of formwork), ASTM C 150 (2009) Type III portland cement
was used. Keeping all of the volumetric mix proportions the same, a second target
mix design using NA-CL with Type I cement was also determined. For both the NAPG and NA-CL target designs, Sika AEA-14 and Sikament 686 were included as
air-entraining agent (AEA) and high range water reducer (HRWR), respectively.
The third target mix (NA-PG-Wet) contained NA-PG coarse aggregate,
ASTM C 150 Type III portland cement, Sika AEA-14 air entraining agent, and no
water reducer. The mix was designed with a w/c ratio of 0.45 for a target 28-day
strength of 31.0 MPa (4.5 ksi) and slump of approximately 17.8 cm (7 in.). While the
w/c ratio of this mix and the other two target mixes were nearly equal, the volumetric
ratio of fresh mortar to coarse material for the third mix was greater (2.23 compared
to 1.38), resulting in a much wetter mix.
RCA Concrete Mix Designs. Three different aggregate replacement methods are
explored in this paper as the direct weight replacement method (DWR), equivalent
mortar replacement (EMR), and direct volume replacement method (DVR). A
detailed comparison of these replacement methods is given by Knaack and Kurama
(2011). By replacing the NA in the mix designs above, a total of 144 RCA mixes
were made. The aggregate replacement ratio, ! was used to quantify the amount of
4
NA replaced by RCA as given by Equations 1, 2, and 3 for the DWR, EMR, and
DVR methods, respectively.
!"#
!"#
! = 1 !!"
/!!"
(1)
!"#
!"#
! = 1 !!" /!!"
(2)
!"#
!"#
! = 1 !!" /!!"
(3)
!"#
!"#
where, !!"
= weight of NA in DWR mix; !!"
= weight of NA in NAC mix;
!"#
!"#
!!"
= volume of NA in EMR mix; !!"
= volume of NA in DVR mix; and
!"#
!!" = volume of NA in NAC mix. Sample RCA mix designs utilizing the DWR,
EMR, and DVR methods are shown in Table 3, with the footnotes at the bottom of
the table listing all of the remaining RCA mixes studied.
Table 3. Dry weight proportions of sample RCA-ND mix designs (with NA-PG target)
Water
Cement
NA
RCA
FA
HRWR
AEA
!
(kg/m3)
(kg/m3)
(kg/m3)
(mL/m3) (mL/m3)
(%) (kg/m3) (kg/m3)
21
150
341
812
216
657
1876
313
41
150
341
607
421
640
1876
313
DWR1,2
61
150
341
401
627
623
1876
313
81
150
341
195
833
606
1876
313
100
150
341
1028
590
1876
313
4
142
322
987
110
640
1775
298
9
132
301
940
235
597
1659
279
EMR3,4,5
14
122
276
887
380
549
1520
255
20
109
248
824
549
492
1365
228
21
150
341
815
203
676
1876
313
41
150
341
606
404
676
1876
313
DVR1,6,7,8
61
150
341
400
600
676
1876
313
81
150
341
198
787
676
1876
313
100
150
341
971
676
1876
313
Note: 1 lb/yd3 = 0.5933 kg/m3; 1 fl oz/yd3 = 38.67 mL/m3
1
Similar mixes also made at ! = 10, 30, 50, 70, and 90%.
2
Mixes repeated with RCA-SB at ! = 10, 20, 30, 40, 50, 60, 80, and 100%.
3
Mixes repeated with RCA-SB at ! = 5, 10, 15, and 20%.
4
Using NA-CL target, similar mixes made with RCA-G at ! = 20, 40%, and with RCA-RR at ! = 20, 40, 60%.
5
Using NA-PG-Wet target, similar mixes made with RCA-ND at ! = 10, 20, 30, 36, 40, 50%.
6
Mixes also repeated with RCA-SB at ! = 10, 20, 40, 50, 60, 80, 100%.
7
Using NA-CL target, similar mixes made with RCA-RR, RCA-R, RCA-W, RCA-P, RCA-GT, RCA-L, RCA-A,
RCA-EL, RCA-EG, RCA-HP, RCA-S, RCA-T, RCA-E, RCA-G, RCA-ND1 at ! = 20, 40, 60, 80, 100%.
8
Using NA-PG-Wet target, similar mixes made with RCA-ND at ! = 10, 20, 30, 36, 40, 50, 60, 70, 80, 90, 100%.
Name
For the DVR and DWR methods, the water reducer was batched
proportionally to the amount of cement (per HRWR manufacturers specifications) so
as to maintain a consistent composition for the fresh mortar. For the EMR mixes, this
approach resulted in a decrease in water reducer (since the amount of cement
decreases as ! increases in EMR mixes). To explore this effect, the maximum
amount of water reducer recommended by the manufacturer was used in the EMR
mixes based on the NA-CL target while the water reducer was varied proportionally
to the amount of cement in the EMR mixes based on the NA-PG target.
Aggregate Preparation. Following dry sieving, the recycled and virgin natural
coarse aggregates were washed with water over a No. 8 (2.36 mm) sieve to further
5
remove excess fines. All coarse and fine aggregates were then dried in an oven at
110 C (230 F) for at least 24 hours. To ensure that the aggregates did not absorb
mix water (in turn pulling water from the cement paste and leading to decreased
workability and cement hydration), the material was removed from the oven, batched
according to the dry weights in Tables 2 and 3, blended (as necessary) according to
the gradations in Table 1, and then soaked in water for a period of 18 to 24 hours.
After soaking, the excess water was decanted from the aggregates, which were then
weighed to determine the amounts of absorbed and residual water. Using the
absorption values from Table 1, the amount of residual water beyond the saturated
surface dry (SSD) condition of the aggregates was subtracted from the required mix
water for each concrete batch.
Mixing. Following aggregate preparation, each concrete batch was mixed in a
rotating drum mixer [with a wet material capacity of 0.02 m3 (0.80 ft3)] according to
ASTM C 192 (ASTM 2009b). The fresh concrete was immediately removed from the
mixer and placed in a non-absorbent metal pan for slump testing and cylinder
molding. In removing the fresh concrete from the mixer, some mortar adhered to the
inside of the drum. This mortar loss was accounted for by adding an estimated
amount of mortar left inside the drum to the original mix design. To determine this
estimated amount, three batches each were made for the NA-PG and NA-CL target
mixes. The average amount of mortar loss for each set of three batches was
determined by subtracting the total material removed from the mixer from the total
weight of the material placed in the mixer. The amounts of fine aggregate, water,
cementitious material, and liquid admixtures were adjusted proportionally to account
for this loss in all subsequent mixes.
Concrete Slump. To determine the workability of each mix, a variation of the ASTM
C 143 (2009a) slump test was conducted. The ASTM test requires a slump cone with
a concrete volume of 5663 cm3 (0.20 ft3). However, because of the relatively small
amount of RCA available from each source, the volume of each batch was limited to
3398 cm3 (0.12 ft3) to make four 7.6 x 15.2 cm (3 x 6 in.) cylinders. To allow for
slump measurement, the dimensions of the ASTM slump cone were scaled by half,
thus reducing the required concrete volume to 721 cm3 (0.03 ft3). The correlation
between the full slump cone and the scaled mini slump cone is described in Knaack
and Kurama (2011) and is given by Equation 4 as:
!"## !"#$% = 2.13 !"#" !"#$%
(4)
Curing. Following slump testing, four 7.6 x 15.2 cm (3 x 6 in.) cylinders were made
from each batch. Three cylinders were used to determine the compressive strength
and stiffness of the concrete, while the fourth cylinder was left unbroken for future
reference and analysis. The cylinders were made using plastic disposable molds,
which were capped and stored at room conditions for approximately one day. The day
after casting, the cylinders were removed from the molds and were cured for 27 days
in water at room temperature, bringing the total curing duration to 28 days. In
accordance with ASTM C 511 (2009a), high-calcium hydrated lime (calcium
hydroxide) was added to the curing water to prevent the leaching of lime from the
concrete.
Testing. After curing, the specimens were hand dried using paper towels and tested
in a moist state according to ASTM C 39 (2009a). Each cylinder was capped using
unbonded steel caps with rubber bearing pads to help maintain an even load
distribution during testing. A servo-controlled hydraulic universal testing machine
was used to apply a constant stress rate of 241 kPa/s (35 psi/s) and the concrete axial
strain was measured using an Epsilon 3542RA rock averaging extensometer with
5.1 cm gauge length. So as to not damage the extensometer, each test was briefly
paused (for less than 10 s) to remove the sensor at a stress of 20.7 to 37.9 MPa (2.5 to
5.5 ksi), depending on the expected strength of the concrete, after which the loading
continued at the same stress rate until specimen failure.
The compressive strength was determined from the
peak stress, !!" reached in each test, and the secant
modulus, !!"# was determined from the slope between
two points on the measured stress-strain curve, first
point at a strain of !!! = 0.00005 and the second point
second at a stress of !!! = 0.40!!" .
RESULTS
(a)
(b)
(c)
Figure 1. RCA concrete
mini slump: (a) DWR
mixes; (b) EMR mixes;
and (c) DVR mixes
(a)
(b)
10
11