Research and experience have shown thatpervious concrete mixtures proportioned to have 15 to 25% air voidcontents should have sufficient infiltration rates to limit storm watersurface runoff and adequate strength to avoid raveling.1 Untilrecently, however, there were no U.S. standards for the verification ofair void content in fresh concrete or infiltration rates for in-placeconcrete. To help producers, contractors, and owners verify that theirpavement projects will perform as needed, ASTM Committee C09.49,Pervious Concrete, has recently introduced Standard C1688, “StandardTest Method for Density and Void Content of Pervious Concrete”2 andStandard C1701, “Standard Test Method for Infiltration Rate of PerviousConcrete.”3 These standards were used as part of the quality assuranceprogram for the construction of a parking lot at the MetropolitanCommunity College (MCC) in Omaha, NE. Using test placements to developa compaction-density relationship, appropriate mixture properties couldbe defined without guesswork. Workability tests and unit weight testsper ASTM C1688 were used to screen loads to ensure that we placed onlyworkable concrete that could be consolidated to achieve a target airvoid content.
UNIT WEIGHT AND AIR VOID CONTENT
Pervious concrete typically comprises a zero slump mixture with littleto no fine aggregate and uniformly graded coarse aggregate. Theworkability of such mixtures can be highly sensitive to variations inmoisture content and compaction effort, leading to large variations inthe final void contents for a given pavement project. By mixing trialbatches for the contractor to use in test placements of pavement, theproducer can obtain unit weight data per ASTM C1688 and air voidcontent Vair (in %) from cylindrical samples according to the procedurein Reference 4. Vair is given by
V^sub air^ = [1 – (W^sub D^ – W^sub S^)/(gamma^sub W^ . V^sub T^)] x 100 (1)
where WD is the weight of the oven-dried sample, WS is the submergedweight of the sample (after tapping to release trapped air), gW is theunit weight of water, and VT is the calculated volume of the sampleusing its measured diameter and length.
For the MCC project,six mixtures were prepared and samples were produced per ASTM C1688during placement of the preliminary test panels (Fig. 1). Unit weightand air void content for each mixture were measured and plotted, and alinear regression analysis was used to determine the relationshipbetween air void content and unit weight (Fig. 2). As one might expect,there is a linear relationship between void content and unit weight ofpervious concrete mixtures, with a maximum unit weight (about 150lb/ft3 [2400 kg/m3]) associated with zero air void content.
It must be noted that the ASTM C1688 procedure (filling a 0.25 ft3 [7L] cylindrical container in two lifts, with each lift consolidatedusing 20 blows from a standard Proctor hammer) will not produce thesame air void content as would be produced in pavement. Our preliminaryfield determination for cores removed from the test panels indicatedthat a mixture with an air void content of 12% and unit weight of 133.5lb/ft3 (2140 kg/m3) when tested per ASTM C1688 would have an in-placeair void content (found per Eq. (1) using core sample data) of 17.5%.This in-place value was specified for the project.
QUALITY ASSURANCE PROGRAM
The owner recognized pervious concrete as a new product and thus madeit very clear that, regardless if the product was successful or not,”we need to know why.” The team was therefore expected to implementprocedures within a set quality control program, including:
* Aggregate moisture tests conducted by the concrete producer before batching operations;
* Unit weight tests per ASTM C1688 conducted at the batch plant by the producer and at the job site on every load of concrete;
* Inverted slump cone tests (described in the following section) conducted at the job site by the owner’s testing agency;
* Estimated unit weight test (described in the following section) conducted on site by the owner’s testing agency;
* Unit weight tests (five total) using 4 in. (100 mm) diameter corestaken from the hardened pavement and tested using the proceduredescribed in Reference 4 by the owner’s testing agency; and
* Permeability tests (six total) per ASTM C1701, taken at the corelocations (prior to coring) by the owner’s testing agency (Fig. 3).
UNIQUE TEST PROCEDURES
Inverted slump cone test
The inverted slump cone test is qualitative, but it allows a rapidevaluation of workability. The procedure involves resting the smallopening of a slump cone against a smooth, hard surface. The cone isthen filled with fresh concrete in one lift, with no consolidation.Excess concrete is struck off, level with the large end of the cone,and the cone is then lifted. The fresh concrete is observed as it flowsout of the cone. If the bulk of the concrete remains in the cone andcan only be discharged by vigorous shaking of the cone, the mixturewill be unworkable. Figure 4 shows two different mixtures afterdischarge. The concrete in Fig. 4(a) was discharged after tapping ofthe cone-the batch was remediated by increasing the water content. Theconcrete in Fig. 4(b) flowed freely from the cone and was approved forplacement.
Estimated in-place unit weight
In thisprocedure, a 0.25 ft3 (7 L) cylindrical container is filled with freshconcrete in one lift, with no consolidation. Excess concrete is struckoff, level with the top of the container. The net weight of theconcrete is determined and the unit weight of the test sample iscalculated. The resulting value is multiplied by a compaction factor,which is based on observations that typical consolidation methods leadto a 1 in. (25 mm) reduction in thickness relative to the initialplacement depth. Thus, for the 6 in. (150 mm) thick pavement requiredon this project, the compaction factor was 7 in./6 in. = 1.17.Estimated unit weight values were correlated with specific regions ofthe in-place pavement.
The pervious concrete pavement was placed by directly discharging theconcrete from mixer trucks onto an aggregate base. Concrete was rakedinto place and consolidated and finished using a hydraulicroller-screed operating directly on top of side forms. As per ACI522.1-08, the concrete was covered with a polyethylene sheetimmediately after finishing.5
The 5650 ft2 (525 m2) pavedarea required 110 yd3 (84 m3) of concrete, which was delivered in 14truckloads. Most of the placement was completed in 2 days, during whichthe average ambient temperature was 65[degrees]F (18.5[degrees]C) andthe relative humidity was 70%.
Inverted slump testing showedthat the first truck was not workable and additional water was addeduntil the concrete had about 12% air void content as measured per ASTMC1688. The second truck had too much water added at the concrete plantand was held until the concrete had about 12% air void content per ASTMC1688. The water content for the third truck was acceptable, soconcrete from this truck was placed while the second load was beingheld. Tests of subsequent loads indicated they also had acceptablewater contents.
Pavement sections were installed with noreports of consolidation or finishing problems. Workers with previousexperience with pervious concrete pavements reported, however, that themixtures would have been considered too “wet” if evaluated by visualinspection only.
Fresh and hardened properties
ACI522.1-08 Section 188.8.131.52 requires that the unit weight of freshconcrete is within +-5 lb/ft3 (+-80 kg/m3) of the specified fresh unitweight. ACI 522.1-08 Section 184.108.40.206.1.b requires that the unit weightof the hardened concrete is within +-5% of the approved hardened unitweight measured in test pan
As indicated previously, thespecified in-place air void content was 17.5%. Extending ACI 522.1-08in-place density requirements to air void content, the allowable rangewould be from 12.5 to 22.5%. Air void contents measured using coresranged from 13.4 to 21.6%- well within the allowable range. Acomparison between estimated and in-place void contents is shown inTable 1. Even though the estimated in-place unit weight test is highlyoperator dependant, the mean of the test results was within 3% of themean of the values measured using cores (Fig. 5).
Table 2compares the air void contents of the fresh concrete (measured usingASTM C1688) and hardened concrete (measured using core samples). Forall five cores, the void contents measured per ASTM C1688 were lowerthan the void contents found using the core samples.
Figure6 shows the general relationship between void contents, as determinedper ASTM C1688, and permeability, as determined per ASTM C1701.Permeability tests were not performed directly on the cores, as thatASTM standard is under development. Because the same equipment andmethods were used to consolidate all pavement sections, Fig. 6 impliesthat initial workability, which influences compaction, also influenceshardened permeability. The largest infiltration rate measured per ASTMC1701 was 2016 in./hour (51,200 mm/hour) and the lowest was only 62in./hour (1600 mm/hour). While our observation of an exponentialincrease in permeability with increased void content is consistent withobservations made by others, the multi-operator reproducibility of thetest method is under evaluation.1,6 INDICATIONS
Our work for the MCC pavement project (Fig. 7) indicates that:
* Air void and unit weight tests per ASTM C1688 can be used to predict in-place air void content;
* The inverted slump cone test is a good predictor of mixtureworkability and provides a rapid method for culling mixtures that willhave unacceptably low unit weights; and
* A requirement thatthe in-place unit weight is within +-5% of the specified unit weight(as per ACI 522.1-08) is appropriate and achievable.
Forworkable mixtures that passed the inverted slump cone test, estimatedin-place unit weights correlated well with measured in- place air voidcontents. Mixtures that met the specified unit weight of 133.5 lb/ft3(2140 kg/m3) were very workable, although they might have beenconsidered too “wet” if evaluated by visual inspection only.