Recycled Aggregate Summary Report

By Tuncer B. Edil and Gregory J. Schaertl

The two most common tests used to determine strength parameters for unbound recycled materials are the Static Triaxial Test and the California Bearing Ratio test. The Static Triaxial Test is typically performed in accordance with ASTM D 2850 and AASHTO T 296, although some state DOTs have been known to use their own standards such as CalTRAN.8 The California Bearing Ratio test is typically performed in accordance with ASTM D 1883 or AASHTO T 193. Kuo2 uses the Limerock Bearing Ratio test which is indigenous to the Florida DOT, and is documented as standard FM5-515.

The two most common tests used to determine the stiffness for unbound recycled materials are the resilient modulus test and the free-free resonant column test. The resilient modulus test is typically performed in accordance with AASHTO TP46-94, Strategic Highway Research Program Test Protocol P-46 (SHRP P-46), or National Cooperative Highway Research Program Protocol 1-28A (NCHRP 1-28A). The free-free resonant column test is typically performed according to ASTM D 4015. Permanent deflection is typically performed by use of a cyclic triaxial test. Moisture susceptibility is typically determined by use of the Tube Suction Test. There is no current standard for the use of the test; however Guthrie and Blankenagel use methods as outlined by Scullion and Saarenketo in 1997.5,6,16

Two typical tests used to assess the durability of a material are the LA abrasion test and the freeze-thaw cycling test. The LA abrasion test is typically performed in accordance with ASTM C 131, although other methods are sometimes used by different agencies, such as Australian test method AS 1141.23. The freeze-thaw cycling test is typically performed in accordance with ASTM D 560.

A method that follows ASTM D 6035 for specimen conditioning is used at the University of Wisconsin-Madison11,13 for frost susceptibility. ASTM D 6035 describes a method to determine the freeze-thaw effects on hydraulic conductivity; in the UW procedure, resilient modulus tests are performed to determine the freeze-thaw effects instead of hydraulic conductivity. Test specimens are compacted in molds at the specified moisture content and density. Preliminary testing on specimens instrumented with a thermocouple showed that complete freezing occurred within one day -19˚C. All specimens are retained in their mold and wrapped with plastic sheet in the freezer for at least one day. After freezing, the height and weight are measured and the specimen is allowed to thaw at room temperature. This process is repeated as many freeze-thaw cycles as desired but typically five cycles are used. After the last cycle, specimens are extruded frozen and thawed inside the resilient modulus cell prior to resilient modulus testing.

Summary of Strength and Stiffness Tests
Bejarano et al8 conducted static triaxial tests on one RAP and two different aggregate materials. Individual RAP and aggregate specimens were compacted at OMC and 95% and 100% of maximum wet density (MWD) according to CalTRANS specification CTM 216. Static triaxial tests were conducted at confining pressures of 0, 35, 70 and 105 kPa. After comparing the shear strengths of the RAP and aggregate, it was determined the shear strength calculated for the RAP was comparable in magnitude to shear strengths calculated for the representative aggregate materials. This shear strength correlation was valid at both 95% and 100% MWD and each of the four confining pressures. Bejarano8 also conducted stiffness tests for the three materials according to SHRP test protocol P-46. Of the three tested materials, the RAP had a higher resilient modulus than the two aggregate materials tested at 95% and 100% MWD. When the compaction level was increased from 95% to 100%, the resilient modulus of the RAP and one of the aggregate materials increased. This change in compaction level had no affect on the resilient modulus of the second aggregate material. Lime stabilized RAP specimens cured for seven days had a higher resilient modulus than the non-stabilized material in all cases.

Bennert et a13 conducted a similar test in which the shear strength of pure (100%) RAP and RCA were evaluated against the shear strength of a dense graded aggregate base course (DGABC) typical of the area the recycled materials would be used. Static triaxial test results for the pure samples indicate that the aggregate alone had higher shear strength than either RAP or RCA alone. Stiffness tests were also conducted on blends of the materials used in the study. Specimens were prepared combining the aggregate with RAP and RCA percentages of 100%, 75%, 50%, 25% and 0% (100% aggregate). Contrary to the strength behavior, it was found that as the amount of recycled material in the blend increased, the resilient modulus of the blended material also increased. Pure (100%) specimens of RAP and RCA had higher resilient modulus values than pure specimens of the virgin aggregate.

Guthrie et al6 evaluated the effects of RAP content on the shear strength of base course materials using the California Bearing Ratio test. Two RAP and two aggregate materials (one recycled and one virgin) were acquired for the test. Specimens were prepared at RAP percentages of 100%, 75%, 50%, 25% and 0% (100% aggregate) for each of the permutations of RAP and aggregate samples. The tests found that the shear strength decreased with an increase in RAP content supporting Bennert et al.’s results.

Blankenagel et a15 conducted a study documenting the difference between RCA samples obtained from demolition projects with relatively new RCA samples obtained through batch-plant overruns and haul-backs. The strength of the material was determined immediately after compaction using the California Bearing Ratio test. The demolition RCA and the haul-back RCA had CBR test results of 22% and 55% respectively. Unconfined compressive strength tests conducted on the material were used to determine strength gain over time due to the residual hydration in the RCA. The strength of the demolition material increased 130% and 180% at three and seven days after compaction, respectively. The strength of the haul back material increased 150% to 190% at three and seven days after compaction, respectively. Higher strength gain in the haul back material is most likely due to a greater amount of unreacted cement in the material as well as a finer material gradation. The average seven-day strengths for the demolition and haul-back material were 1,260 kPa and 1,820 kPa, respectively.

Kuo et al2 incorporated the use of the Limerock Bearing Ratio (LBR) in Florida to determine the strength of RCA to be used as potential base course. The overall LBR values for the materials tested were 181.71%, which is higher than the required minimum value of 100%.

Kim et al14 studied the effect of RAP content on the resilient modulus of blended aggregate base course. An in-situ blend of FDR was taken during the reconstruction of an existing road along with pure samples of RAP and aggregate materials. The FDR and several blends of the pure RAP and aggregate base material were tested for material stiffness using the resilient modulus test in accordance with NCHRP 1-28A protocol.

Blended mixtures of the pure materials were prepared at RAP to aggregate ratios (%:%) of 0:100, 25:75, 50:50 and 75:25. The study found that for an increase in RAP content, the resilient modulus of the blended material increased.10 The effects of increased RAP content were more defined when the blends were exposed to higher confining pressures, however specimens also experienced higher permanent deformation at higher confining pressures. Specimens tested at 65% optimum moisture content had higher resilient modulus values when compared to specimens prepared at 100% OMC.

This trend was consistent for all confining pressures. At low confining pressure (~20kPa), specimens with RAP to aggregate ratios of 50% to 50% and specimens consisting of 100% aggregate had resilient modulus values that were approximately equivalent. As the confining pressures increased, the 50:50 and pure RAP blends became stiffer. The 50:50, 100% RAP and in-situ material tested at the corresponding site had similar resilient modulus values.

Nataatmadja et al4 evaluated the resilient modulus of four RCAs. One commercial and three laboratory-produced RCAs were used in the study. The commercial RCA had an estimated compressive strength of 15 MPa, and the three laboratory manufactured RCAs had compressive strengths of 18.5, 49 and 75 MPa. The materials were tested individually and were not blended with any other material, although each material was prepared and mixed as to produce a particle size distribution comparable to typical road aggregate blends. The study found the resilient modulus of each of the RCAs tested was comparable or better (higher) than the typical aggregates used for roadway base course; the resilient modulus seemed to increase with an increase in the compressive strength of the material. An increase in elongated particles also led to a decrease in resilient modulus, as these particles were more prone to degradation after extensive loading. Nataatmadja suggests that RCA with very high compressive strengths are more prone to break into elongated particles during crushing, resulting in a lower resilient modulus than would otherwise be expected. One exception in the test is that the specimen with a high flakiness index produced a lower strength value than would be expected.

Guthrie et al6 used the free-free resonant column test to determine the stiffness of RAP and aggregate blends. At OMC, the stiffness of the material decreased with the addition of 25% RAP, and then increased with the addition of 50%, 75% and 100% RAP. When the material was dried for 72 hours, the trend reversed: the stiffness of the material increased with the addition of 25% RAP and then decreased with the addition of 50%, 75% and 100% RAP. This decrease in stiffness can be attributed to the softening of the asphalt in the RAP during the drying process. Each specimen was then soaked for 24 hours prior to being tested for stiffness a third time. As with the oven-dried specimens, the soaked specimens displayed an increase in stiffness with the addition of 25% RAP followed by a decrease with increased RAP content. However, the soaked materials displayed a 40% to 90% decrease in stiffness when compared to the ovendried materials.

Blankenagel et al5 also used the resonant column test on RCA samples procured from demolition and haul-back sources. During the first 12 hours in 100% relative humidity, the modulus increased 390% for the demolition material and 940% for the haul-back material. Again, a greater amount of unreacted cement in the haul-back material accounts for the larger stiffness. Average seven-days stiffness measurements for the demolition and haul-back materials were 100 MPa and 150 MPa, respectively.

The tests performed at the University of Wisconsin-Madison11,13,15 on two RPMs indicated results in general support of the investigations summarized above. The unsoaked CBR values of RPM varied from 9 to 38 and, as an indicator of strength, were lower than the CBR of aggregates with similar gradation. However, higher resilient moduli11 (257-309 MPa) were measured consistently for RPM compared to different crushed aggregates qualified as base course material.

Addition of fly ash increased the modulus of RPM (at least a factor of 6, which is less than for a similarly stabilized natural aggregate), and the modulus increased as the fly ash content was increased. Modulus also increased with curing time, with the rate of increase being largest between seven and 28 days of curing. The moduli of RPM stabilized with fly ash were independent of bulk stress and could be described by a constant modulus.

Summary of Moisture Susceptibility Tests
In the tube suction test, a specimen is oven dried for 72 hours before being allowed to soak in a shallow water bath for 10 days. Over the course of the soaking period, unbound water within the material rises through the aggregate matrix and collects at the surface. The dielectric value at the surface of the material increases with an increase in the amount of unbound water permeating the specimen, and thereby provides an estimate of the materials susceptibility to moisture permeation.

Guthrie et al6 used the tube suction test to determine the effect of RAP content on the moisture susceptibility of RAP/aggregate blends. It was found the moisture susceptibility of the material increased as RAP was added to the mixture. However, tests were only conducted with the addition of 25% and 50% RAP. Materials with RAP contents above 75% were classified as non-moisture-susceptible and were not tested. Overall, the dry density of the blended material decreased as RAP content increased.

Blankenagel et al5 used the tube suction test on demolition and haul-back RCA to help determine the moisture susceptibility characteristics of the material. The moisture susceptibility of the demolition material was classified as “good”, with a dialectic value of 6.4 and a gravimetric water content of 10.6%. The moisture susceptibility of the haul-back material was classified as “marginal”, with a dialectic value of 15 and a gravimetric water content of 2%.

Summary of Durability Tests
Blankenagel et al5 incorporated the LA Abrasion and freeze-thaw cycling test into his study comparing demolition and haul-back materials. Results of the LA Abrasion tests indicated that the demolition and haul-back materials experienced average material losses of 31% and 18%, respectively. The primary cause of the degradation was thought to be the stripping of cement paste from the aggregate. This degradation caused an increase in fines that affected each of the two RCAs differently. The demolition material was initially low in fines content, and an increase in degradation fines would lead to an increase in MDD. The haul-back material was initially high in fines content, and the addition of degradation fines would decrease the structural stability and increase the moisture susceptibility of the material.

Nataatmadja et al4 attempted to use the LA abrasion test to determine the relative hardness of the four RCAs. Commercial RCA had a lower hardness than laboratory manufactured RCAs, even though commercial RCA had the lowest (estimated) compressive strength. The relative hardness between the laboratory-manufactured RCAs could not be differentiated by the LA Abrasion Test method, most likely due to test severity.

Blankenagel et al5 used freeze-thaw cycling to measure the durability of the demolition and haul-back RCMs. Freeze thaw testing was performed after seven days of curing. Specimens were submerged for four hours, frozen (-29˚ C) for 24 hours and thawed (+20˚C) for 24 hours. Stiffness was measured after each freezing period and after each thawing period. The demolition RCM experienced a 30% stiffness loss within the first two cycles and thereafter stabilized at a stiffness of 70 MPa. The haul-back RCM experience a 90% stiffness loss over the first nine cycles and thereafter stabilized at a stiffness of 30 MPa. Unconfined compressive strength tests for the materials after freeze-thaw testing indicated strength losses of 52% and 28% for the demolition and haul-back material, respectively.

Freeze-thaw cycling tests performed at the University of Wisconsin-Madison showed that there was a small effect on resilient modulus (less than 15%) for RPM and also for natural aggregate with or without fly ash, with no consistent effect for materials stabilized with fly ash.

Summary of Permanent Deflection Tests
Bennert et al3 studied the effect of recycled material content on the permanent deflection experienced by base course materials. Specimens were created from blends of aggregate with either RAP or RCA. For cyclic loads of 100,000 cycles, specimens blended with RCA were found to have the lowest amount of permanent deformation, and specimens blended with RAP had the highest amount of permanent deformation. RPMs tested at the University of Wisconsin-Madison exhibited smaller plastic strains during resilient modulus testing than base course aggregate, i.e., the opposite of the resilient modulus trend. However, other data show that plastic strains for RPM may be higher or lower than those of conventional base aggregates, depending on the type of aggregate used. Plastic strains for RPM stabilized with fly ash were smaller than the plastic strains of the RPM alone.

Conclusions
Several important findings were noted in the course of this literature review. Kim et al14 compared the compaction properties of specimens prepared by typical proctor methods with specimens prepared with a gyratory compactor and found that the OMC and MDD of the specimens compacted via gyratory compactor were found to more closely correlate with field density measurements. Kim also found that at low confining pressures, pure aggregate and 50%:50% blends of RAP and aggregate had an equivalent stiffness, but at high confining pressures the 50%:50% blends had a higher stiffness than the pure aggregate. Bennert et al3 found that pure specimens of RAP and RCA had higher resilient moduli than pure virgin aggregate specimens. Bennert also found that specimens of pure aggregate had higher shear strength than pure RAP or RCA specimens. This trend is supported in a study by Guthrie et al6 in which RAP/aggregate blends showed a decrease in shear strength as RAP content increased.

In general, RPM seems to show a better response than natural aggregate for similar gradation and compaction in tests that induce relatively smaller strains such as resilient modulus tests than tests that induce large strains such as triaxial compression or CBR tests.

Tuncer B. Edil is a professor and Greg Schaertl is a graduate student at the University of Wisconsin-Madison. Dr. Edil can be reached at This e-mail address is being protected from spambots. You need JavaScript enabled to view it .

References
1    Poon, C. S., Qiao, X. C. and Chan, D. X. (2006). “The Cause and Influence of Self-Cementing Properties of Fine Recycled Concrete Aggregates on the Properties of Unbound Sub-Base”, Waste Management, Vol. 26, No. 10, pp. 1166-1172
2    Kuo S. S., Mahgoub, H. S. and Nazef, A. (2002). “Investigation of Recycled Concrete Made with Limestone Aggregate for a Base Course in Flexible Pavement”, Geomaterials, No. 1787, pp. 99-108
3    Bennert, T., Papp Jr, W. J., Maher, A. and Gucunski, N. (2000). “Utilization of Construction and Demolition Debris Under Traffic-Type Loading in Base and Subbase Applications”, Transportation Research Record, No. 1714, pp. 33-39
4    Nataatmadja, A. and Tan, Y. L. (2001) “Resilient Response of Concrete Road Aggregates”, Journal of Transportation Engineering, Vol. 127, No. 5, pp 450-453
5    Blankenagel, B. J. and Guthrie, W. S. (2006). “Laboratory Characterization of Recycled Concrete for Use as Pavement Base Material”, Geomaterials, No. 1952, pp. 21-27
6    Guthrie, W. S., Cooley, D. and Eggett, D. L. (2007). “Effects of Reclaimed Asphalt Pavement on Mechanical Properties of Base Materials”, Transportation Research Record, No. 2006, pp. 44-52
7    “User Guidelines for Byproducts and Secondary Use Materials in Pavement Construction.” FHWA Report FHWA-RD-97-148, Federal Highway Adminis-tration, McLean, Virginia (2008).
8    Bejarano, M.O., Harvey, J. T., Lane, L. (2003). “In-Situ Recycling of Asphalt Concrete as Base Material in California”, Proceedings of the 82nd Annual Meeting, Transportation Research Board, Washington D.C. CD-Rom, 22 pp.
9    Saeed, A. (2008). “Performance-Related Tests of Recycled Aggregates for Use in Unbound Pavement Layers”, NCHRP Report 598, Transportation Research Board, Washington, D.C., 53 pp.
10    Li, L., Benson, C. H., Edil, T. B., Hatipoglu, B., and Tastan, O. (2007). “Evaluation of Recycled Asphalt Pavement Material Stabilized with Fly Ash”, ASCE Geotechnical Special Publication, CD-Rom, 10 pp.
11    Carmargo, F., Wen, H., Edil, T. B., and Son, Y. H. (2009). “Laboratory Evaluation of Sustainable Materials at MnRoad”, Proceedings of the 88th Annual Meeting, Paper No. 09-3160, National Research Council, Washington D.C., CD-ROM.
12    Wen, H., and Edil, T. B. (2009). “Sustainable Reconstruction of Highways with In-Situ Reclamation of Materials Stabilized for Heavier Loads”, BCR2A Conference, Champaign, Illinois
13    Wen, H., Baugh, J., and Edil, T.B. (2007). “Use of Cementitious High Carbon Fly Ash to Stabilize Recycled Pavement Materials as a Pavement Base Material”, Proceedings of the 86th Annual Meeting, Paper No. 07-2051, National Research Council, Washington D.C., CD-ROM.
14    Kim, W., Labuz, J. F. and Dai, S. (2007). “Resilient Modulus of Base Course Containing Recycled Asphalt Pavement”, Transportation Research Record, No. 2006, pp. 27-35
15    Wen, H., Warner, J., and Edil, T. B., (2008). “Laboratory Comparison of Crushed Aggregate and Recycled Pavement Material with and without High Carbon Fly Ash”, Proceedings of the 87th Annual Meeting, Paper No. 08-3075, National Research Council, Washington D.C., CD-ROM
16    Scullion, T. and Saarenekto,
T. (1997). “Using Suction and Dielectric Measurements as Perfor-mance Indicators for Aggregate
Base Materials”, Transportation Research Record, No. 1577, pp.
37-4

| More...