From the GAP 2019 transportation engineering conference: Danniel Rodriguez, Jose Garibay, and Soheil Nazarian from the Center for Transportation Infrastructure Systems (University of Texas – El Paso) examine asphalt-stabilized bases. ASBs provide structural underlayment to hot mix asphalt (HMA) surface course. An asphalt-treated base is generally composed of lower quality material than the HMA and also adheres to less stringent quality control standards in the form of wider gradation bands and lower to moderate asphalt content. The CTIS research shows interesting observations on how to optimize ASB performance, including with regard to indirect tensile strength. The research is valuable for a number of reasons, not the least of which is their consideration of economic constraints in infrastructure works and the need to find design-safe, beneficial uses of lower quality materials. Geosynthetics are also used in various situations with treated bases.
1. BACKGROUND ON ASPHALT CONTENT & TREATED BASES
Asphalt-stabilized base (ASB) is placed to provide a waterproof, structural underlayment to HMA surface course. ASB is generally composed of lower quality material than that of HMA and also adheres to less stringent quality control standards in the form of wider gradation bands and lower to moderate asphalt contents. The design of ASB is not a well-defined procedure (Nazzal, 2009). The recommendations for ASB design vary. The process can consist of electing a predefined gradation and asphalt content that satisfies performance testing, such as load wheel tracking (LWT) and indirect tensile strength (IDT) (Nazzal, 2009). The design may also be performed in a manner that is similar to that of designing for the optimum moisture content of high-quality granular base, or for the OAC of low-quality hot mix asphalt (HMA). The design of ASB has been evaluated by means of HMA convention through Marshall mix design method (Li, 2010), SUPERPAVE design and Bailey method (Hua, 2008). However, it is often not practical to adjust the gradation in order to satisfy all the volumetric and performance requirements considered in HMA designs. In actuality, the criteria for passing the LWT requirements and design strength are often waived.
Volumetrically, the OAC of an HMA is generally defined at a specified molded density. Figure 1 displays a typical design function for obtaining the OAC. As with typical HMA design, the procedures call for evaluation of the mix to meet minimum voids in mineral aggregate (VMA) requirements; depending on the type of HMA design mix. The minimum design VMA can be as low as 12% for dense-graded mix and up to 19% for stone matrix asphalt mix (Texas DOT, 2008). Specifying a minimum VMA is also not a common criterion in practice. Employing HMA design for ASB is almost always based on the developed relationship between the asphalt content and total density.
HMA design is generally performed through use of the Superpave Gyratory Compactor (SGC) or in some highway agencies (such as TxDOT) with the small Texas Gyratory Compactor (TGC). The laboratory study presented in this paper uses both compactors, following Texas specifications.
TGC accommodates HMA design and molds 4 in. x 2 in. specimens. Given the size of the mold, aggregates that are 0.75 in. diameter and larger are scalped out. Compaction occurs at 250°F after 2- hours of curing. The material is pressed with the ram to 50 psi, and then gyrated three times. This process is repeated until the pressure reaches 150 psi after one pump of the ram. The pressure is then pumped to 2,500 psi, and then relaxed to complete the molding process.
Designing with the SGC employs 6 in. x 4.5 in. specimens. Material is compacted at a temperature of 290°F after 2-hours of curing. The compactor tilts and gyrates at a rate of 30.0±0.5 gyrations per minute at an angle of 1.25±0.02°. During compaction, the ram application pressure maintains 87±2 psi perpendicular to the cylindrical axis of the specimen. Specimens are molded up to 75 gyrations or 100 gyrations. However, previous research has shown the locking points of most ASB mixes to be less than 75 gyrations (Hernandez, 2012).
RELATED: Geostructural Characterization of Foundation Materials Using Automated Plate Load Testing
Strength testing for design of ASB has involved unconfined compressive strength (UCS) and indirect tensile strength (IDT). UCS recommended testing requires 6 in. x 8 in. specimens. These size of specimens have been produced through use of the Texas Base Gyratory Compactor (TBGC). However, agency-adopted compaction method has shifted towards primary use of the SGC. It follows that the preference for strength testing consists of IDT on 6 in. x 4.5 in. specimens. The IDT performed in this study applies the compressive load at a controlled deformation rate of 2 in. per minute.
This study will propose a design method that is similar to that of designing high-quality base. The method is evaluated in comparison to HMA design through use of the SGC and TGC. It is also becoming popular for highway agencies to include high percentages of reclaimed asphalt pavement (RAP) and some recycled asphalt shingles (RAS) into the mix designs. Therefore, the materials used for the evaluation all contained at least 20% RAP content. Sample sets consisting of added 3% RAS content were also evaluated. The impact on the OAC by the design convention (high-quality base design vs. HMA design), number of gyrations and RAS content are assessed. Additionally, the impact of the RAS content on strength is also presented.
2. PROPOSED DESIGN PROCESS: HIGH-QUALITY BASE APPROACH
This approach follows a similar process to that of designing for the optimum moisture content of unbound granular bases. As shown in Figure 2, the design OAC is based on relationships between the asphalt content and total density, relative density, and indirect tensile strength (IDT) from a set of at least three laboratory specimens, molded with different asphalt contents. The objective of this design procedure was to obtain a practical OAC between 3% and 6%. This design calls for compacting the specimens with 75 gyrations of SGC.
A sample ASB design is presented in Figure 2 to demonstrate the proposed method. Design criteria of at least an IDT of 85 psi and a relative density of 97% are recommended (Texas DOT, 2013). Depending on the designer preference, the OAC values can be biased toward a “dry mix” and a “rich mix.” The steps associated with this procedure consists of obtaining the following parameters:
- AC max density: is defined as the AC at the maximum total density of the observed data set (Figure 2a).
- AC relative density: the AC values at a relative density of 97% for a “dry” design (left of vertex) and for a “rich” design (right of vertex), as shown in Figure 2b.
- AC strength: the AC values meeting strength of 85 psi for a “dry” design and for a “rich” design. See Figure 2c.
- AC critical: minimum of AC relative and AC strength, respectively for “dry” and “rich” design.
- OAC design: average of AC max density and AC critical, respectively for “dry” and “rich” design (See Figure 2d). In the case where the strength design curve exceeds the design criteria at every point, then the OAC design is defined at 97% relative density for a dry and rich design. This signifies that the strength has no impact on the design since the criteria is met at every asphalt content.
3. EVALUATION OF PROPOSED DESIGN METHOD
3.1 Experimental design
A study was performed to compare the OAC obtained from the proposed high-quality base design method with OAC obtained from low-quality HMA design method. Common base materials from Texas were used in this case study. Table 1 contains the experiment design. The experiment design was executed on virgin material mixed with RAP, and with virgin material mixed with RAP and RAS. Each sample set consisted of one trial specimen per asphalt content. Each sample set was prepared and tested for the high-quality base method and low-quality HMA method. That is, for each set, the volumetric properties were obtained first for OAC analysis with HMA design, and then followed with the IDT for OAC analysis with high-quality base method. In this paper, the OAC values obtained from the two methods are distinguished by: 1) OACbase-design and 2) OACHMA-design. The low-quality HMA method was further evaluated to compare the OAC obtained by compacting with SGC and TGC.
The test process is depicted in Figure 3. The specimens were mixed, molded, then measured for volumetric properties (bulk specific gravity and theoretical maximum specific gravity) and finally subjected to the IDT strength tests.
The tested materials represent typical ASB mix designs placed in Texas and were retrieved as raw, unmixed material. To incorporate RAS to the mix designs, 3% of the RAP percentage was replaced with RAS. The following are the material characteristics [material name, aggregate type, RAP mix percentage, RAP asphalt content]: [Mix A, Limestone Dolomite, 30%, 5.2%], [Mix B, Limestone Dolomite, 30%, 5.5%], [Mix C, Limestone Dolomite, 25%, 3.5%], [Mix D, Limestone Dolomite, 20% , 6.0%]. Figure 4 displays the combined gradations for the materials.
3.2. OAC results
As an example, Figure 5a displays the proposed high-quality base design procedure carried out on one of the test materials with 0% RAS content. The specimens were prepared with SGC at nominal AC of 3%, 4.5% and 6%. The resulting dry and rich OACbase-design are 3.7% and 6%, respectively. Since the IDT strengths of all specimens exceeded the design criteria, the OAC were controlled solely by density. This pattern was observed for most of the materials. The rich OACbase-design values for almost all cases were 6%. Figure 5b displays the results from the same specimens, but based on the low-quality HMA philosophy. The OAC in that case was 4.7% at a design density of 96% (typical design density used).
RELATED: Implementation of Geotechnical Asset Management
Following the low-quality HMA philosophy, but compacting with the TGC, yielded an OAC of 5.0%.
Shown in Table 2 are the OAC values obtained with 75 and 100 gyrations following the proposed highquality base design and low-quality HMA design procedures. The results with 0% and 3% RAS are also included. Figure 6 summarizes the average OAC values from the four materials following each design process. The error bars correspond to the respective minimum and maximum OAC values of the four mixes for each design method. Also shown in the figure is the typical OAC range derived in practice (4+0.3% OAC). The proposed high-quality base design method yields an average dry OAC of 3.7%. The addition of 3% RAS does not seem to affect the average OAC. This OAC is slightly lower than that currently derived in practice from following the HMA design methods. HMA design with a) SGC and with b) TGC produce average OACs of approximately 4.8%. This average OAC is slightly greater than that obtained during practice. This greater OAC is due to the wider AC band evaluated for each data set in this experimental design plan in comparison to that performed during practice. This experimental design plan caps each data set to 6% versus 5%-5.5%, as performed in practice. A maximum AC of 5%-5.5% will force a lower OAC value due to the linear fit at 96% target molded density. For common practice, it is recommended for the OAC design sets to be evaluated at a wide AC band (up to 6%) in order to attain truly representative design OAC values. The addition of RAS does not seem to influence the average OAC when following the HMA design method. Likewise, the use of SGC vs. TGC also does not significantly affect the OAC.
Figure 7 displays the OAC values for a tested material following each design method to identify any sensitivity to the number of gyrations. For this and the other materials evaluated, the OACs are not too sensitive to the number of applied gyrations, considering that the gyrations are within the general vicinity of the locking points.
Overall, the OAC is more sensitive to the design method followed (high-quality base design vs. low-quality HMA design). The proposed high-quality base design method will yield a slightly dryer OAC than HMA design procedure.
4. IMPACT OF RAS ON ASB STRENGTH
The impact of RAS on IDT strength was assessed by comparing the results from specimens with 0% RAS content to those with 3% RAS. As an example, Figure 8a and Figure 8b display the absolute and relative increase in strength by adding 3% RAS to one of the tested materials. The IDT strength increases with the addition of 3% RAS.
A global analysis of 3% RAS influence on strength at varying asphalt contents is displayed in Figure 9. For each asphalt content, the average, minimum, and maximum strength ratios of 3% RAS to 0% RAS are presented. The strength ratios are almost always equal to or greater than unity; illustrating an increase in IDT strength by adding 3% RAS to the mix designs. The increase in strength is shown to mostly range between 10% and 25%.
An evaluation of ASB design practice was performed to recommend a procedure that would yield moderate to dry OAC design. A proposed high-quality base design procedure was compared to that of a low-quality HMA to assess the impacts on the obtained OAC. The influence of the type of compactor (SGC vs. TGC), number of gyrations, and the presence of RAS was also assessed. The evaluation was performed on 4 commonly used mixtures in Texas. The results show that high-quality base design process to yield OAC values that are slightly dryer when compared to OACs obtained from HMA design method. In essence, the contractor can choose to follow either design procedure (proposed high-quality base design or HMA design) to obtain an acceptable ASB OAC. For common practice, it is recommended for the OAC design sets to be evaluated at a wide AC band (up to 6%) in order to attain representative design OAC values. The presence of RAS in the mix, the number gyrations (near the locking point), or the type of compactor had no significant influence on the derived design OAC. However, the presence of 3% RAS in the mix designs was shown to increase the IDT strength of the mix by 10% to 25%.
The authors would like to thank the Texas Department of Transportation for their guidance and support to perform this study. The authors would like to acknowledge engineering personnel from the TxDOT Construction and Laboratory divisions for their assistance in acquiring raw material for this study. The authors would also like to thank the undergraduate research assistants that performed the laboratory testing.
Hernandez, H., Garibay, J., and Nazarian, S. (2012). Development of a New Mix Design Method and Specification Requirements for Asphalt Treated Base. Research Report 0-6361-1, Center for Transportation Infrastructure Systems, The University of Texas at El Paso.
Hua, Q. (2008). Dense-Graded Asphalt Treated Base Design Method Investigation. Presented at Fifth International Symposium on Transportation and Development, Beijing, China.
Li, Z., Chen,Y., and Xing, Z. (2010). Experimental Study of Asphalt Treated Base Binder Course for Pavement Design. Presented at International Conference of Chinese Transportation Professionals, Beijing, China.
Nazzal, M., and Mohammad, L. (2009). Evaluation of Low Cost Asphalt Treated Base Mixtures. Presented at Louisiana Transportation Engineering Conference, Baton Rouge, LA.
Texas Department of Transportation. (2008). Design of Bituminous Mixtures: Tex-204-F. TxDOT,. Texas Department of Transportation. (2013). Molding, Testing, and Evaluating Asphalt Black Base Materials: Tex-126-E. TxDOT.