CENTER FOR TRANSPORTATION RESEARCH THE UNIVERSITY OF TEXAS AT AUSTIN

Project Summary Report 0-4177-S Project 0-4177: Use of Recycled Asphalt Pavement and Crushed Concrete as Backfill for Mechanically Stabilized Earth Retaining Walls Authors: Ellen Rathje, David Trejo, and Kevin Folliard March 2006 Potential Use of Crushed Concrete and Recycled Asphalt Pavement as Backfill for Mechanically Stabilized Earth Walls Mechanically stabilized earth concrete (CC) and recycled asphalt to MSE walls. The index tests (MSE) walls (Figure 1) have been pavement (RAP) for use as select performed included grain size used throughout the U.S. since backfill for MSE walls. distribution, specific gravity, and the 1970s. The popularity of MSE . The compaction systems is based on their low cost, What We Did... characteristics of CC and RAP aesthetic appeal, simple construc- An extensive laboratory inves- were assessed with standard Tx- tion, and reliability. To ensure tigation was conducted to evaluate DOT laboratory compaction tests, long-term integrity of MSE walls, the potential use of CC and RAP and field tests were performed to select backfills consisting predomi- as backfill in MSE walls. The assess field compaction control of nantly of granular have been laboratory investigation involved CC and RAP. The potential for poor used. However, with increasing a geotechnical evaluation of the performance of CC when the source environmental and sustainabil- materials, as as a corrosion concrete had previously suffered ity concerns, interest in the use evaluation that assessed whether sulfate attack of alkali-silica reac- of recycled materials as backfill these materials will cause excessive tion (ASR) was assessed by a suite for MSE walls has grown. Some corrosions of the metallic reinforce- of expansion tests. of the most commonly available ment embedded in MSE walls back- testing was performed in the labora- recycled materials are crushed fill. For all of these tests, the results tory using a 4-in. diameter triaxial concrete (CC) and recycled asphalt from CC and RAP were compared device, as well as a 20-in. by 20- pavement (RAP), and these materi- with results from a conventional in. large-scale direct shear box. als are being considered for use as fill material (CFM) consisting of tests and backfill in MSE walls in Texas. This crushed limestone. collapse potential tests were also Project Summary Report (PSR) The geotechnical evaluation of performed. The pullout resistance summarizes the results from the CC and RAP involved a suite of of typical steel reinforcing strips Texas Department of Transporta- laboratory and field tests to char- was assessed using pullout tests tion (TxDOT) Project 0-4177, as acterize the important properties performed in the large-scale direct related to the evaluation of crushed of these materials as they pertain shear box. Finally, constant-stress creep tests were performed on RAP using the triaxial device. The corrosion evaluation of CC and RAP consisted of evaluat- ing the performance of plain- and galvanized-steel straps embed- ded in the three different backfill materials with exposure to cyclic water and chloride-ion solution applications. Samples were fabri- cated and exposed to the different environments for approximately 11 months. After the exposure period PROJECT SUMMARY REPORT the samples were evaluated for mean corrosion rates using mass loss data. Corrosion rates of the metallic reinforcement straps were Figure 1. Schematic of MSE wall (Schlosser and Delage 1988) used to estimate the service life of

Project Summary Report 0-4177-S –  – MSE wall structures. In addition, the upon wetting is not a concern for CC was less than the average mass loss of research team attempted to correlate if the materials are compacted at water the same reinforcement embedded in the short-term solution test results with the contents that represent saturation levels CFM (crushed limestone) when exposed long-term test results such that, when of 70 to 80 percent. to a non-chloride solution environment. necessary, other backfill materials Pullout testing: Pullout tests were For galvanized-steel reinforcement, the could be assessed over shorter durations performed in the 20-in. by 20-in. shear CC and the CFM exhibited similar mass ( ~ 1 month). box using steel ribbed reinforcement loss values when exposed to a non-chlo- embedded in crushed concrete. Mea- ride solution environment. Although What We Found... sured values of pullout force were used the comparative performance of the to evaluate F*, the pullout resistance metallic material embedded in the CC Crushed Concrete (CC) factor or -bearing factor, which and CFM backfill materials exhibited Gradation and Compaction: The is used to predict the ultimate pullout similar results for the chloride solution gradation of crushed concrete provided resistance of reinforcement for MSE exposure, average corrosion rates were by commercial producers in Texas meets wall design. The measured F* values at very high for both cases. No correlation the TxDOT Item 423 Type B backfill different confining pressures were all between the short- and longer-term test- gradation specification, which is the greater than those predicted by the de- ing was observed. The service life times gradation generally designated for per- sign procedures from the Federal High- of galvanized reinforcement embedded manent MSE walls. Crushed concrete way Administration (FHWA), indicating in CC and CFM, based on the measured exhibited compaction characteristics that the traditional predictive equations average corrosion rates from the longer- similar to CFM. for F* can be used for CC. term tests, are shown in Figure 2 and Durability: Expansion of compacted Corrosion testing: Various backfill indicate that the service life for CC is CC samples was monitored over a pe- material characteristics were assessed. longer than for CFM. The pH and resistivity values are com- riod of 70 to 100 days under various Recycled Asphalt Pavement (RAP) detrimental conditions. The samples monly used to indicate the corrosion included commercial CC and CC potential of backfill materials. It was Gradation and Compaction: The gra- derived from concrete that had previ- determined that the CC used in this in- dation of RAP provided by commercial ously suffered ASR or sulfate attack. vestigation does not meet the current pH producers in Texas meets the TxDOT The expansion of most samples was and resistivity TxDOT specifications for Item 423 Type B backfill gradation negligible except for the samples that MSE wall backfill. However, the results specification, which is the gradation had experienced sulfate attack. These showed that the average mass loss (and generally designated for permanent samples experienced up to 4 percent average corrosion rates) of the plain- MSE walls. RAP exhibited adequate volumetric expansion over 70 days steel reinforcement embedded in CC compaction, although it could not hold when exposed to water. Strength: The results from consoli- dated-drained shear strength tests indi- Galvanized Reinforcement in CFM (No chloride exposure) Galvanized Reinforcement in CC (No chloride exposure) cate that crushed concrete has strength 100 Galvanized Reinforcement in RAP (No chloride exposure) characteristics comparable to those of conventional fill materials. Using the combined results from triaxial testing 80 and large-scale direct shear testing, the derived effective shear strength param- eters for CC were: c′=9 psi, ′=46°. ϕ The CFM, which also met TxDOT back- 60 % reduction that results in fill gradation specifications, displayed yielding of reinforcement very similar shear strength parameters (c′=10 psi, ′=46°). 40 ϕ Drainage properties: Falling-head, rising-tail hydraulic conductivity tests were performed on CC specimens in 20 a triaxial apparatus. The hydraulic

conductivity of CC ranged from 10-4 to (%) Thickness Reinforcement in Reduction MSE 10-5 cm/s over confining pressures of 5 to 50 psi. The hydraulic conductivity 0 of the CFM was close to 10-3 cm/s. The 0 20 40 60 80 100 low hydraulic conductivity of CC is a Time (years) concern. Collapse Potential: The collapse Figure 2. Estimated service life times for galvanized MSE strips embedded in potential tests showed that collapse CFM, CC, and RAP using average corrosion rates.

Project Summary Report 0-4177-S –  – 35 duced service life values for MSE walls. ı'3 = 20 psi As with the CC testing, no correlation 0.80 0.70 0.66 30 0.64 was observed between the data from 0.85 D = 0.88 0.62 the shorter- and longer-term corrosion test programs, and the shorter-term 25 testing is not recommended for assess- ing the corrosion activity of metallic 20 materials embedded in different back- fill materials. The service life time of galvanized reinforcement embedded in 15 0.58 RAP was shorter than for CC and CFM

Axial Strain (%) Strain Axial (Figure 2). 10 0.50 The Researchers 5 0.40 Recommend... The results from the geotechnical 0 and corrosion studies of crushed concrete 0.1 1 10 100 1000 10000 100000 (CC) indicate that this material displays Time (min) adequate gradation and compaction characteristics, shear strength, pullout Figure 3. Axial strains developed over time at different resistance, and corrosion performance, deviatoric stress levels ( D ) in creep tests. although it displays low hydraulic con- significant water because of the bitumen box using steel ribbed reinforcement ductivity and may experience significant coating on the particles. embedded in RAP. The force-controlled expansion if the source concrete suffered Strength: The results from strain- pullout tests experienced significant sulfate attack. Based on these results, controlled consolidated-drained triaxial creep deformations, with the deforma- CC is recommended for use as backfill tests indicate that RAP has acceptable, tion limit of 0.75 in. being reached for MSE walls. However, the following although somewhat marginal, strength before shear failure along the rein- issues must be considered: characteristics comparable to those forcement- interface. Nonetheless, 1. MSE walls with crushed concrete of CFM. The derived effective shear derived F* values at the deformation backfill should include adequate strength parameters for RAP were: limit were similar to those predicted drains and high permittivity filter c′=8 psi, ′=37°. Because these tests ϕ by design procedures, except at larger fabrics behind the wall to avoid were strain controlled, they do not confining pressures. drainage problems. include the effects of creep. The large- Creep testing: A series of con- 2. Concrete structures that have scale direct shear tests, which are force stant stress, deviatoric creep tests suffered sulfate attack cannot be controlled, could not be successfully were performed under drained condi- crushed and used as backfill in performed on RAP because of the creep tions in a triaxial apparatus. Tests per- MSE walls. deformations. formed at larger deviatoric stress levels Drainage properties: Falling-head, ( D = ( ' - ' )/( ' - ' ) ) expe- 3. pH and Resistivity specifications σ 1 σ 3 σ 1 σ 3 ult rising-tail hydraulic conductivity tests rienced significant strains and creep for MSE wall backfill materials were performed on RAP specimens in rupture within 1 week of testing (Figure should be waived for crushed con- a triaxial apparatus. The hydraulic con- 3). These results indicate that the creep crete. ductivity of RAP ranged from 0.5x10-3 potential in RAP is significant, and The results from the geotechnical to 4x10-3 cm/s over confining pressures its creep behavior is similar to that of and corrosion studies of recycled as- of 5 to 50 psi. These values indicate that clays under undrained conditions. Ad- phalt pavement (RAP) indicate that this RAP is a free-draining material. ditionally, creep rupture is a concern material displays adequate gradation, Collapse Potential: The collapse and the creep potential of RAP appears strength, and hydraulic conductivity potential tests showed that collapse to be most severe at smaller confining properties. However, RAP displays a upon wetting for RAP is larger than pressures. significant potential for creep deforma- for CC and CFM. This finding stems Corrosion testing: The longer-term tions, and these creep deformations may from the fact that RAP cannot hold corrosion test results using RAP indi- lead to excessive deformation in a MSE significant amounts of water, and thus it cate that the corrosion of MSE metallic wall. Additionally, corrosion testing in- is compacted at lower saturation levels. reinforcement (both galvanized and dicated that RAP caused more corrosion Nonetheless, the collapse potential was plain) embedded in this backfill mate- than either CC or CFM. Based on these only slight and not a concern for RAP. rial results in higher average corrosion results, RAP is not recommended for use Pullout testing: Pullout tests were rates than in the CFM specimens. These as backfill for MSE walls. performed in the 20-in. by 20-in. shear higher corrosion rates can lead to re-

Project Summary Report 0-4177-S –  – For More Details... Research Supervisor: Ellen Rathje, Ph.D., P.E., (512) 232-3683 email: [email protected] TxDOT Project Director: Marcus Galvan, P.E., Bridge Division, (512) 416-2224 email: [email protected] TxDOT Research Engineer: Tom Yarbrough, P.E., Research and Technology Implementation Office, (512) 465-7403 email: [email protected]

The research is documented in the following reports: 0-4177-1, Recycled Asphalt Pavement and Crushed Concrete Backfill: State-of-the-Art Review and Material Characteristics 0-4177-2, Recycled Asphalt Pavement and Crushed Concrete Backfill: Results from Initial Durability and Geotechnical Tests 0-4177-3, Evaluation of Crushed Concrete and Recycled Asphalt Pavement as Backfill for Mechanically Stabilized Earth Walls

To obtain copies of a report: CTR Library, Center for Transportation Research, (512) 232-3126, email: [email protected]

Your Involvement Is Welcome! Disclaimer This research was performed in cooperation with the Texas Department of Transportation and the Federal Highway Administration. The contents of this report reflect the views of the authors, who are responsible for the facts and accuracy of the data presented herein. The contents do not necessarily reflect the official view or policies of the FHWA or TxDOT. This report does not constitute a standard, specification, or regulation, nor is it intended for construction, bidding, or permit purposes. Trade names were used solely for information and not for product endorsement. The engineer in charge was Ellen Rathje (Texas No. 94549).

The University of Texas at Austin Center for Transportation Research Library 3208 Red River #115 Austin, TX 78705-2650