No-Fines Pervious Concrete lar Paving by Richard C. Meininger Results of a laboratory study of nofines pervious concrete for paving are presented. Conclusions are drawn regarding the percent air voids needed for adequate permeability, the optimum water-cement ratio range, and the amounts of compaction and curing required. Recommendations are made regarding appropriate uses for this type of concrete.
o-fines, pervious concrete is being used for paving in situations where it is desired to have rainfall or surface water percolate through the pavement into a permeable base. The elimination of fine aggregate produces concrete in which the coarse aggregate particles are coated with a water-cement paste that bonds them together at their contact points. The fairly large voids left between the coarse aggregate particles allow the concrete to be permeable to water. This concrete is used in Florida to eliminate storm water run-off from parking lots and reduce the need for separate storm water retention ponds in shopping centers and developments. It is particularly useful in areas where local or state regulations require that storm water be retained on the site to recharge the groundwater system with fresh water and to reduce the need for storm sewers.
N
Data and references are available from NRMCANAA, 900 Spring St., Silver Spring, MD 20910.
20
Fig. 1 - No-fines concrete facilitates drainage to a storm water retention basin.
No-fines concrete is also used for paving in greenhouses and nurseries where it is undesirable to have free water on paved surfaces. It permits the surface to appear relatively dry with no streams or puddles of water from irrigation of plants or nursery stock. Use of pervious concrete also allows a parking lot to be built around trees, without cutting off air and moisture to the roots below. In extreme cases, to impound the needed amount of rainfall in heavy storms, parking lots are being designed to store water not only in voids of the pavement or base material, but also on top of the pavement. These parking lots temporarily store an additional 6 in. (15 cm) of rainfall up to the curb line on a completely fIat 101. Their entrance aprons must be humped up enough to retain the design storm water amount and not allow it to run out into the adjacent road or gutter. No fines pervious concrete has been used as an open-graded drainage material in bases under sidewalks and light duty pavement, and also as drainage layers under highway shoulders, to allow water trapped under pavements to flow more rapidly out of the pavement structure. A double-barrelled approach uses no-fines concrete plus a drain pipe to a small storm water retention basin (Fig. 1).
Laboratory research Several research series were conducted in the National Aggregates Association (NAA) - National Ready Mixed Concrete Association (NRMCA) Joint Research Laboratory to develop information concerning proportioning methods as well as methods of measuring the strength and permeability of nofines pervious concrete. Batch sizes ranging from 1 to 3 ft3 (0.028 to 0.085 m') were mixed in rotating drum laboratory mixers. Laboratory stock Type I cement was used with two sizes of stock coarse aggregate: moderately rounded gravel aggregates of % in. (9.5 mm) maximum size (ASTM C 33, No 8 size), and % in. (19 mm) maximum size (ASTM C 33, No. 67 size). Fig. 2 shows the end of a broken 6 x 12 in. (152 x 305 mm) strength cylinder and a 4 x 14 in. (102 x 356 mm) cylinder used in freezing and thawing tests (both cylinders were made from laboratory concrete containing No. 8 size aggregate). The properties of no-fines concrete depend not only on its proportions but also on its compaction. To better understand the effect of compaction on concrete air void content, unit weight, and comKeywords: coarse aggregates; laboratories; no-fines concretes; research; voids.
Concrete International
125
.~.>
120
..... .'~
f'
e: :I:
~ ;::
.
\
BATCH A, w/c
z 110
::>
Il:
w e
~ t. ~'
...
~><~
u
100
Fig. 2 - No-fines pervious concrete strength test and freeze-thaw test cylinders.
• No compaction (just scoop concrete in). • Fill in two layers; tilt and drop cylinder after each. • Two layers; five tamp cornpaction of each with 5 lb circular tamper. • Two layers; five drops of the tamper for each layer. • Two layers; five drops each using a proctor hammer. • Two layers; 15 tamps each layer. • Three layers; 25 drops each with the proctor hammer.
105
) l. 2. 3. 4. 5. 6. 7. 8.
pressive strength, concrete batches were mixed at two different watercement ratios (0.31 and 0.34) and 6 x 12 in. (152 x 305 mm) cylinders were then made using eight different procedures:
/
// /
= 0.34
~v
~
./.ATCH
1 B, W/C
=
0.31
7 8 METHODS OF COMPACTION (AND 28-DAY COMPRESSIVE STRENGTH OF CYLINDER FROM BATCH A, BATCH B) NO COMPACTION, CONCRETE SCOOPED INTO MOLO 2 LAYERS, TILT ANO OROP MOLO 2 LAYERS, 5 TAMPS EACH OF THE TAMPER (1355, 975 psi) 2 LAYERS, 5 OROPS EACH OF THE TAMPER (1340 psi, 1050 psi) 2 LAYERS, 5 DROPS EACH, PROCTOR HAW1ER (1360 psi, 1100 psi) 2 LAYERS, 15 TAMPS EACH OF THE TAMPER (1550 psi, 1395 psi) 3 LAYERS, 25 DROPS EACH, PROCTOR HAMMER (1945 psi, 1540 psi) ASTM C 31 ROOOING PROCEOURE, 3 LAYERS (2475 psi, 2095 psi)
Fig. 3 - Cylinder unit weights and strengths for eight different compaction methods.
• C 31 rodding compaction; 25 strokes on each of 3 layers. The resulting cylinder unit weights and strengths are shown in Fig. 3. Unit weight ranged from a low of about 105 lh/ft ' (1682 kg/ m') with no compaction up to about 120 lb /ft" (1922 kg z'rn') with the ASTM C 31 rodding procedure. The cylinders with no compaction were unsatisfactory because they contained large voids and discontinuities, so three levels representing different amounts of cornpaction that might be obtained in paving were chosen for future work: Light (5 tamp compaction of each of two layers) Unit weight range: 107 to 111 Ib/ft3 (1714 to 1778 kg/rn ')
b .t~ Fig. 4 - Bottom surface of percolation specimens with open channels (flow). August 1988
/)
115
8¡,¡
.)
-:
Compressive strength range: 980 to 1360 psi (6.8 to 9.4 MPa) Medium (15 tamps on each of two layers) Unit weight range: 111 to 1141b/ft3 (1778 to 1826 kg/rri') Compressive strength range: 1400 to 1550 psi (9.6 to 10.7 MPa) Heavy (C 31 rodding compaction [3 layers]) Unit weight range: 120 to 122 Ib/ft3 (1922 to 1954 kg/ru') Compressive strength range: 2100 to 2480 psi (14.5 to 17.1 MPa) The tamper used for the 5 and 15 tamp compaction level was made from pipe fittings. It weighs 5.0 lb
4'~~"~·"~;"~~~~¡__JL~ZL:~~...
LUL~:':.
Fig. 5 - Bottom surface of percolation specimens with blocked channels (no flow). 21
3000
Pervious
Concrete
continued .~
2500
Ul
o.
:2
E-<
t!)
z
""
6, NO. 8 COARSE AGGREGATE
2000
P:: E-< U)
""
~ 30~~--~~----+------+------4------4--4
:>
3::
H U) U)
~ ~
e,
~ 25r------+------~~~~~----~----~~ u ¡:¡:;
u ><
ril
Cl
1500
""
P::
z
~ 8 Z ril ~Z
o
1000
~
co N
20
500
o
MIXER HOLDBACK DRY
u
¡:¡:;
15
H
.:¡;
.25
.30
.30
.50
.45 .40 .35 WATER-CEMENT RATIO
Fig. 6 - Relationship between w/c and air content for two compaction levels.
(2.3 kg) and has a vertical handle with a horizontal circular metal, 4in. (l02-mm) diameter tamping head.
Test methods
unit weight of the concrete was measured in a 0.25 fe (0.007 m ') unit weight bucket using the 5 tamp two layer compaction. Unit weights were also determined on cylinders made by both the light compaction (5 tamps, 2 layers) and heavy compaction (C 31 rodding, 3 layers), and on flexural beams made at two compaction levels. Compressive strength was measured on 6 x 12 in. (152 x 305 mm) cylinders made using the light cornpaction and heavy compaction.
w/c
Water (lb/yd')
0.51 0.47 0.43 0.39 0.35 0.31 0.27
440 430 430 425 415 410 395
224 203 184 165 145 125 106
22
.40
.45
.50
Fig. 7 - Relationship between 28-day compressive strength and water-cement ratio.
Flexural strength was measured on 6 x 6 x 36 in. (152 x 152 x 914 mm) beams using two breaks per beam with third-point loading (ASTM C 78). Beams were prepared to simulate the same two compaction levels used on cylinders. For the light compaction they were prepared in one layer using the 5 lb (2.3 kg) tamper, and for the heavy cornpaction using two layers and the ASTM C 31 rodding procedure. Air void content of the concrete was calculated gravimetrically from the unit weight data determined from a11 types of specimens (unit weight bucket, 6 x 12 in. [152 x 305 mm] cylinders, and flexural beams). The rate at which water could percolate through the no-fines per-
Table 1 - Detailed data for light compaction tests (5 tamp compaction; No. 8 coarse aggregate; aggregate·cement ratio = 6) Cement (lb/ydJ)
.35
WATER-CEMENT RATIO
Coarse aggregate
Air
(lb/yd')
(OJo)
Strength (psi)
Percolation (in./min)
2640 2575 2570 2550 2520 2430 2370
22 23 25 27 29 32 33
1350 1370 1500 1400 1250 1010 870
5 4 10 30 40 51 59
vious concrete was measured in a percolation test rig consisting of a 6 x 12 in. (152 x 305 mm) plastic cylinder mold with the bottom cut out placed over a cured half-height (6 x 6 in. [152 x 152 mm]) cylinder made from the test concrete. Mastic was placed on the sides of the 6 x 6 in. (152 x 152 mm) cylinder to prevent water flowing out the sides, and then the plastic cylinder mold was slid down about 3 in. (76 mm) over the percolation cylinder. Water was run through the percolation cylinder for a few minutes to condition it, and then the water level aboye the cylinder was raised to more than 5 in. (127 mm) aboye the cylinder and the hose shut off. The percolation rate was measured by timing the drop in the water SUfface from a point 5 in. (127 mm) aboye the top surface of the cylinder to a point 1 in. (25 mm) aboye the cylinder. This was converted to inches of rainfa11 transmitted per minute by dividing 240 by the number of seconds it took the water level to drop 4 in. (102 mm). Fig. 4 (flow) and Fig. 5 (no flow) shows the bottom surface of sorne of the percolation specimens. The voids between the coarse aggregate particles of specimens with a higher paste content can become blocked so that the channels in the specimen Concrete International
Table 2 - Cernent content ranges for strength test results shown in Fig. 7 (No. 67 coarse aggregate) a/c
w/c
Cernent (lb/ydJ)
4 6 10
0.25 to 0.49 0.25 to 0.49 0.27 to 0.51
600 to 680 400 to 450 250 to 270
100
80 e
'e
<,
e
60
•..< Ul
Z
;;
8E-<
o
Table 3 values
~1>:
1>:
Ul
40
2000
E-<
¡:iI
e,
¡..¡
~
20
5-Tarnp
O 10 20
o
U
b/bo 070 Sand
E-<
o
H
Effective b/bo
3000
o.
Ul Ul ¡:.¡
IW.
1000
1>:
C 31
67 C. AGG
A/C = 6
p..
W/C = 0.38
~
o
(.)
NO.8 0.92 0.84 0.78
No. 67 0.92 0.85 0.78
No.8 0.99 0.93 0.85
No. 67 0.99 0.93 0.86
are not continuous in sorne cases, causing a no flow condition.
Water-cernent ratio To investigate the effect of water-cernent ratio, a series of batches were rnixed with ratios varying frorn 0.51 down to 0.27. In this series a fixed aggregate-cernent ratio (a/c = 6) was used with the No. 8 coarse aggregate. Here basically relatively fixed cernent and aggregate contents were used, and as the water content of the batches increased, the water-cernent paste occupied more of the voids in the coarse aggregate, thus lowering the air void content. Fig. 6 shows the gravirnetric air content of the light and rnediurn cornpaction batches calculated frorn the unit weight data. As water-cernent ratio increased the air void content decreased in a linear relationship, with the light cornpaction having air voids about 2 percent higher than the rnediurn cornpaction. Observation of the consistency of the water-cernent paste and how the concrete handled in the mixer indicated that water-cernent ratios in the 0.35 to 0.45 range are best for efficient coating of the aggregate. Low water-cernent ratios
August 1988
10
20
AIR CONTENT,
30
40
PERCENT
10
20
30
PERCENT SAND
Fig. 8 - Minimum air void content of 15% is needed for flow.
Fig. 9 - Adding 10 to 20% sand increases compressive strength.
tended to cause balling and sticking of the concrete in the mixer, resulting in substantial hold-back of concrete when the drum was tilted for discharge. High water-cernent ratios gave a thin paste that could run off the aggregate during placernent resulting in increased variability and blockage of water flow channels. Table 1 shows detailed data for the light cornpaction. As the air content increased frorn 22 percent to 33 percent, the rate of percolation increased from about 5 to 50 in./rnin (127 to 1270 rnrn/rnin). The cornpressive strength appeared to be optirnurn in the mid-range of watercernent ratios. At high water-cernent ratio the strength was lower, and at very low water-cernent ratios (where the air content is higher) strengths becarne very low, because the past volurne was greatly reduced and did not bind the aggregate particles together as well.
The strength data frorn these mixtures is shown in Fig. 7, along with the strength data for those using No. 8 size coarse aggregate. Again, the trend is to have better strengths in the rnid-range of watercernent ratios for both sizes of coarse aggregate. The traditional water-cernent ratio law does not hold for these mixtures beca use of the large differences in air-void content and the difficulty of handling and cornpacting mixtures with very low water-cernent ratios. The percolation rate becornes very low (or no flow was observed) when the air void content of the specirnens become as low as 15 percent (Fig. 8). It appears that void content values need to be 15 percent or more to assure flow.
Test results for No. 67 coarse aggregate A series of batches were rnixed using the % in. (19 mm) rnaxirnurn size coarse aggregate (No. 67) and three cernent content ranges governed by the aggregate-cement and water-cernent ratios used (Table 2).
Effect of adding sand The strength of no-fines concrete irnproves when a srnall arnount of sand is added (10 to 20 percent as a percentage of the total aggregate weight) (Fig. 9). As sand is added to the mixture it tends to fill the voids, reducing the air content frorn 26 percent to 22 and 17 percent in this case, and raising the compres sive strength frorn about 1500 psi (10.3 MPa) to about 2500 psi (17.2 MPa).
23
Pervious
Concrete
continued 1.0 NO. 8 C. AGG. 40
0.9
30
0.8
E-< Z >il U
tt: >il
o..
~
E-< Z >il E-< Z
o .c .c
<,
0.7
o u 20
5-TAMP
o:; H
,c¡;
0.6
10
NO. 8 C. AGG. 0.5
O O
10
20
30
40
50
o
PASTE CONTENT, PERCENT
Fig. 10 - Relationship between paste and air contents for two levels of compaction and three sand contents.
Proportioning
relationships
A series of batches of concrete was mixed, all using the same midrange w/c of 0.38, and two size of coarse aggregate - No. 8 and No. 67. For each aggregate size, sand percentages were varied from O percent up to 50 percent. At sand contents over about 30 percent the concrete became more normal in consistency and did not have the content of larger voids necessary to allow water to flow through.
Proportianing
20
30
40
50
PASTE CONTENT, %
Fig. 11 - Relationship between paste content and b/b, for two levels of compaction and three sand contents.
An example of the data for the Vs in. (9.5 mm) aggregate (No. 8) is shown in Fig. 10 for light cornpaction (5 tamp; solid lines) and heavy compaction (C 31; dashed lines). For each condition sand contents of O, 10, and 20 percent are shown. The family of curves for the % in. (19 mm) aggregate (No. 67) are very similar. These curves could be used to proportion concrete within the desired air content range of about 15 to 22 pecent (enough voids to allow water flow but not reduce
strength to an unacceptable level). For example, one could pick a series of batches along the 20 percent air line at a w/c = 0.38. In increasing order of paste content the trial batches would be for C 31 compaction: 13OJo paste volume; 20% sand 15% paste volume; 10% sand 17% paste volume; 0% sand for 5 tamp compaction: 18% paste volume; 20% sand 20% paste volume; 10OJo sand 22% paste volume; 0% sand
procedure
1. Required data on coarse aggregate: SSD specific gravity; absorption; dry-rodded unit weight 2. Select a mid-rangc w!c (0.33 to 0.45) 3. Select a trial b/b: (Table 3) 4. Calculate batch weight of coarse aggregate (b/bo) (dry-rodded unit weight) (batch volume); correct to SSD 5. Calculate absolute volume of coarse aggregate 6. Select target air void content 10-15% - little or no flow; good strength 15-20% - permeable; fair strength
24
10
20-30% - highly permeable; poor strength 7. Calculate absolute volume of air 8. Calculate absolute volume of sand, (if any) 9. Calculate sand batch weight (if any) 10. Calculate absolute volume of water-cement paste 11. Calculate batch weight of cement 12. Calculate batch weight of water 13. Mix trial batch and determine: unit weight; air void content; yield; strength; etc. for compaction level desired (or several compaction levels) 14. Adjust batch weights
Concrete International
6000
COMPACTION -- 5 TAMP
50
~
5000
:>
~E-<
oex:
o
o.
40
::C E-<
t!l
o o
~
ti)
30
~
~
z H H
:>
3000
~ ~ O
2000
H ti) ti)
::E Z
4000
Z cr; E-<
oex:
oex: cr;
PERCENT SAND O, 10, 20
fJJ
~cr;
.....:¡
-- C 31
• .-1
cr; P<
20
ti) Q
U
H
O
:> E-<
z
10
w
1000
NO. 8 C. AGG.
U cr; W
P<
o o
10
20
30
40
50
5
PASTE CONTENT, PERCENT
Fig. 12 - Relationship between paste content and VMA for two levels of compaction and three sand contents.
Another approach to proportioning that appears to work well is the use of the b rb ; concept as ernployed in A'C] 211.1 for proportioning normal weight concrete. Here the ratio b/b; compares the amount of coarse aggregate in a unit volume of concrete with the amount of coarse aggregate in a like volume of dry rodded coarse aggregate (ASTM e 29 Test Method). Fig. 11 shows calculated b/b; values for the series of batches with No. 8 coarse aggregate at various paste contents, for 2 levels of compaction and O, 10, and 20 percent sand contents. For practical pastecontents aboye about 10 percent the curves are relatively flat, indicating that b/b; (amount of coarse aggregate in the concrete) is relatively constant and not affected by the past content. However, compaction level and sand content do affect the amount of coarse aggregate in the concrete. The b/b; curves for the % in. (19 mm) (No. 67) coarse aggregate are very similar to those for the Vs in. (9.5 mm) (No. 8) size. Table 3 shows the effective b/b, values for the series of concrete batches. The b/b; concept, in using the dry-rodded unit weight of coarse aggregate, automatically
August 1988
10
15
20
25
30
AIR CONTENT, PERCENT
Fig. 13 - Relationship between gravimetric air content and 28-day compressive strength of cylinders.
compensates for the effect of different coarse aggregate particle shape, grading, and specific gravity. Therefore, these values should be usable for trial batches with any normal weight coarse aggregate. They can be used to estimate the amount of coarse aggregate per cubic yard, if the level of compaction in the resulting construction can be selected accurately. The same data can be plotted in a different way (Fig. 12) to show voids in the mineral aggregate (VMA) in the same way that asphalt technicians look at the voids between the aggregate skeleton in an asphaltic concrete mixture. In that context VMA is the voids in the mixture of coarse aggregate and fine aggregate (i f any is used). However, I feel the b/b; approach is easier to use.
Comprehensive example A series of batches were mixed at two compaction levels (light [5 tamp compaction] and heavy [e 31 cornpaction]) and with two sizes of coarse aggregate (No. 8 and No. 67). All were mixed using a watercement ratio of 0.39. The principalpurpose of this series was to better
define the fIow Ino flow boundary for air void contents in the lOto 25 percent range, using a series of mixtures with O, 10, and 20 percent sand. Water percolation rate was measured on specimens made from each mixture, and both compressive and flexural strength was determined. Fig. 13 shows the relationship of compressive strength to air void content as calculated gr avirnetrically from the unit weight of the cylinders. All test conditions are included in this data. Fig. 14 shows flexural strength data versus the air content calculated from the unit weight of each beam specimen. These mixtures are highly dependent on the void content. Fig. 15 shows the relationship of flexural strength to compressive strength for this series and for other fairly low strength regular concrete data from the NAA-NRMeA Joint Research Laboratory. The heavy compaction data (black circles) appear to be more in line with previous laboratory data, while the light compaction (open circles) show a higher flexural strength than might be expected for the corresponding compressive strength. Sorne of the explanation for this
25
700
D D
600
Ul
a.
UJ
:i
o.
i E-< o
E-<
:z: ¡,:¡
400
e
E-< Ul
e
400
E-<
~p
>< ¡,:¡
300
,_:¡
rx..
. .. . . D
a
• ·a
l' •
e e
,_:¡
Ul
... .. . .. o
300
X W H tLo
-:
200
200
5
10
15 AIR CONTENT,
100 ~----~----~----~----~----~----~ 2000 3000 4000 1000 20 PERCENT
25
appears to be related to a difference of about 2.5 percent air content of the concrete as compacted in the beam mold versus that in the cylinder mold using light compaction. The air content was higher in the cylinders than that in the beams, probably due to the more confined shape of the cylinders and greater probability of friction and confinement in the center porion of a beam moldo
Limited freezing and thawing tests Freezing and thawing specimens (4 x 14 in. [102 x 356 mm] cylinders with gage studs in the ends) were molded from no-fines concrete with No. 8 coarse aggregate containing: Cement 497 lb/yd' (295 kg/rn') Water 194 lb/yd' (115 kg/rri') Gravel 26001b/yd3 (1543 kg/rri') w/c 0.39 No admixtures were used. The mix characteristics were: Air void content: 21 percent. Unit weight: 121 lb/yd' (72 kg/rn') Compressive strength: 1910 psi (13.2 MPa)
5000
6000
30 COMPRESSIVE
Fig. 14 - Relationship between gravimetric air content and flexural strength of beams.
26
..
¡:>:;
W p:;
~::J
e
(él
Z
H
..
500
/f..
.
•.-1
500 ...,
STRENGTH,
psi
Fig. 15 - Relationship between flexural strength and compressive strength.
Flexural strength: 320 psi (2.2 MPa) Percolation rate: 7.3 in./min (185 mm/min) The specimens were cured in a standard moist room for 30 days, at which time half of the specimens were subjected to ASTM C 666 Procedure A (freezing and thawing in water) and half to Procedure B (freezing in air and thawing in water). AH of the specimens failed fairly quickly in both freezing exposures, indicating that the voids in the concrete became saturated and the water was not able to drain out quickly enough to prevent freezing damage in the rapid (5 cyc1es per day) freeze-thaw exposure. The cylinders appeared to crack and tend to split lengthwise, indicating a build up of pressure due to freezing of water in larger internal voids. It was not the type of failure where material sloughs off the outside surface. The rapid freezing from all directions may have driven water to the interior of the specimens, and when the internal water froze there was no avenue of pressure relief. For a situation where freezing is slower, and from one direction, there may be more opportunity for water to drain out of the no-fines paving material. Caution needs to be exercised when using a product such as this
where it might beco me saturated prior to a hard freeze. Sorne concrete producers are using air-entrainment in the paste; this may improve durability, but it may affect the permeability characteristics as well. Caution must also be exercised in using pervious concrete in exposures where sulfates or acids may be involved since the perrneability of the product would allow such aggressive solutions to penetrate and attack the interior of the concrete.
Example parking lots Fig. 16 shows a good job where no- fines pervious concrete has been used to advantage. Raveling can occur when there is insufficient hardened paste to hold the top coarse aggregate, when the top aggregate pieces are not correctly seated into the concrete, and when poor curing allows the cement paste to dry before sufficient hydration has taken place. Fig. 17 shows a hand screed finish on a parking lot. No additional compaction and seating of the coarse aggregate was accomplished, and that coupled with poor curing caused the parking lot to have a rough, raveling surface one year after construction. Concrete International
Fig. 16 - A quality job using no-fines pervious concrete.
Fig. 17 - Hand screed finish of no-fines permeable concrete paving provides inadequate compaction and seating of aggregate.
Fig. 18 shows the surface and a construction joint of a new parking lot where extra effort was applied in properly seating the top layer of coarse aggregate and in curing the concrete. The screed used to strike off the concrete had a rounded edge which tended to compact the top surface and a manually operated steel lawn roller was run over the surface just behind the screeding operation to properly embed the aggregate. Immediately following that the concrete was covered with sheet plastic to insure proper curing of the concrete.
porous concrete can dry out very rapidly if not quickly covered with plastic sheeting. Curing is vital to the continued hydration, and resistance to abrasion, of the top surface. The level of compaction must be considered in the design of the mixture. Too much compaction can reduce the air voids to below 15 percent and plug the flow channels. Too little compaction will leave the structure with very high air voids resulting in low strength and a raveling surface. Compact test specimens to the same density as will be obtained in the field. It may take sorne experimenting to obtain comparable compaction in the field and laboratory. The CSA Canadian Standard s have sorne information on how this can be done. No-fines pervious concrete is a viable option for automobile parking lots in warm climate areas. There is concern that this lower strength concrete will not stand up well where frequent truck or bus traffic may be involved. Regular normal-weight concrete should be used for bus or truck lanes in parking lots and also in areas with frequent abrasion or turning maneuverso The use of no-fines concrete in surface courses should be confined to automobile parking areas or other light duty uses.
Conclusions It appears that at least 15 percent air void content is required to obtain the needed percolation in nofines concrete. A water-cement ratio in the range of 0.35 to 0.45 does a better job of coating the coarse aggregate without causing too much balling in the mixer or , at the opposite extreme, being so wet that the paste tends to run off the aggregateo Construction methods are critical to proper performance. Sorne compaction is needed during placement and the coarse aggregate on the top surface needs to be properly seated to reduce ravelling of the surface. Curing is very important since the August 1988
,.\~
b·l'
,.
,
..
.1:....•
11~·~.:~
~~:ii~~ ..'.....' ~ .~ '..
,t
'r... ·
1... L~,,\~u. Fig. 18 - Proper compaction and curing gives a tight surface. ACI member Rich· ard C. Meininger is Vice President of Research of the National Ready Mixed Concrete Association and the National Aggregates Association, Silver Spring, Maryland. These associations sponsor the NAA-NRMCA Joint Research Laboratory in College Park, Maryland, where this research was conducted. He is a member of the ACI Technical Activities Committee, and Committees 211, Proportioning Mixtures, 221, Aggregates, and 226, Fly Ash, other Pozzolans, and Slag. Mr. Meininger was a recipient of the ACI Construction Practice Award in 1984. 27