Skip to contentUnited States Department of Transportation - Federal Highway AdministrationSearch FHWAFeedback

Pavements

A Review of Aggregate and Asphalt Mixture Specific Gravity Measurements and Their Impacts on Asphalt Mix Design Properties and Mix Acceptance

Maximum Specific Gravity of Asphalt Mixtures

Current Standard Test Methods

The test method most often used to determine Gmm is AASHTO T 209. Within the method, there are several options for determining the Gmm but all utilize the same basic principle of measuring the mass and volume of the loose mix sample to determine its maximum specific gravity. The options within AASHTO T 209 differ by the type of sample container and whether the container is filled with water or submerged in a water bath. There are three container choices: bowl, flask, or pycnometer.

An outline of the procedure is as follows:

  1. The dry mass of the loose mix samples are first determined and the mix is then placed in a tared container of one of the types previously mentioned.
  2. Water is added to the container to completely cover the sample and a vacuum is applied to remove entrapped air.
  3. The container is then filled with water and the mass determined or it is placed in a water bath and the mass determined.
  4. From these mass determinations, the volume of the loose mix and thereby its Gmm is determined

AASHTO T 209 also contains detailed procedures related to the calibration of flasks, bowls and pycnometers, as well as temperature corrections for the asphalt binder in the loose mix and the density of the water used in the test procedure if the test temperature differs from 25C (77F).

A survey conducted by the AMRL for the Aggregate Task Group (ATG) shows that out of 34 states that responded to the survey, 22 use AASHTO T209, and 12 states modify the test method to improve between laboratory precision. Most modifications reduce the options allowed in T 209.

The ASTM method for determining Gmm is D 2041. D 2041 is nearly the same as AASHTO T 209 with the exception that the calibration and volume correction issues are treated differently between the two methods. Whereas, T 209 provides calibration and volume correction procedures for tests that are conducted at temperatures substantially different than 25C (77F), D 2041 mandates that the test be conducted at temperatures of 25±1C (77±1.8F) to avoid the necessity of using correction factors.

The AASHTO and ASTM methods contain similar procedures for the determination of the Gmm for asphalt mixtures containing porous aggregate, commonly referred to as the "dryback" method. Essentially, the dryback procedure is aimed at determining how much water is absorbed into the coated particles during vacuum saturation. The tested sample is dried using a fan to a constant mass. The AASHTO method stipulates that this is only necessary for aggregate with water absorption greater than or equal to 1.5 percent. ASTM does not specify an absorption value, nor does it give any other criterion for determining whether a mixture should be tested using the alternate dryback procedure.

Precision Estimates of Current Standard Test Methods

The AASHTO and ASTM methods provide single operator and multilaboratory precision values for both procedures (non-porous and porous aggregate mixtures). The AASHTO precision values are shown in Table 8, and the ASTM precision values are shown in Table 9. No information is provided regarding the type of container used or whether the container was filled with water or weighed under water for non-porous aggregate mixtures. The ASTM acceptable range of two results for both single operator and multilaboratory conditions for non-porous aggregate mixtures are more than two times greater than the corresponding AASHTO values. The AASHTO and ASTM d2s precision values for both single operator and multilaboratory conditions for absorptive aggregate mixtures shown in Tables 8 and 9 are identical, implying that the same data set was used for the determination of the precision values.

Table 8 AASHTO T 209 Precision Estimates for Gmm

  Standard Deviation (1s) Acceptable Range of Two Results (d2s)
Single Operator Precision: Without supplemental dryback 0.0040 0.011
With supplemental dryback for absorptive aggregate mixtures 0.0064 0.018
Multilaboratory Precision: Without supplemental dryback 0.0064 0.019
*With supplemental dryback for absorptive aggregate mixtures 0.0193 0.055

* Values only apply to bowl determination of Gmm.

Table 9 ASTM D 2041 Precision Estimates for Gmm
  Standard Deviation (1s) Acceptable Range of Two Results (d2s)
Single Operator Precision: Without supplemental dryback 0.0080 0.023
*With supplemental dryback for absorptive aggregate mixtures 0.0064 0.018
Multilaboratory Precision: Without supplemental dryback 0.0160 0.044
*With supplemental dryback for absorptive aggregate mixtures 0.0193 0.055

* Values only apply to bowl determination of Gmm.

ASTM D 2041 precision estimates for mixtures containing aggregate with absorption of less than 1.5 percent or between 4 to 5 percent were evaluated in NCHRP 9-26 (2,11). The precision estimates for D 2041 from NCHRP 9-26 are presented in Table 10 and are much smaller than the corresponding values shown in Table 9.

The Proficiency Sample Programs also publish precision estimates for AASHTO T 209 and ASTM D2041 annually. These precision indices are shown in Table 11 below. Information about whether absorptive or non-absorptive aggregate mixtures used and how many laboratories used supplemental dryback for absorptive aggregate mixtures was not published on the AMRL website (1) at the time of this writing.

Table 10 Precision Estimates for ASTM D2041 Evaluated in NCHRP 9-26
  Standard Deviation (1s) Acceptable Range of Two Results (d2s)
Single Operator Precision: Without supplemental dryback for aggregate with less than 1.5% absorption 0.002 0.006
With supplemental dryback for aggregate with 4 to 5% absorption 0.005 0.0013
Multilaboratory Precision: Without supplemental dryback for aggregate with less than 1.5% absorption 0.004 0.011
With supplemental dryback for aggregate with 4 to 5% absorption 0.010 0.027
Table 11 AASHTO T 209/ASTM D2041 Precision Indices Published by AMRL
Year Sample No. No. of Labs Single Operator Multilaboratory
Participated* Data Used** 1s d2s 1s d2s
2007 21/22 475 430 0.0037 0.0105 0.0057 0.0160
2006 19/20 435 415 0.0060 0.0170 0.0083 0.0234
2005 17/18 405 398 0.0072 0.0203 0.0115 0.0325
2004 15/16 358 352 0.0059 0.0166 0.0086 0.0244
2003 13/14 305 300 0.0041 0.0117 0.0080 0.0227
2002 11/12 281 271 0.0043 0.0123 0.0073 0.0207
2001 9/10 235 230 0.0053 0.0151 0.0070 0.0198
2000 7/8 221 214 0.0048 0.0135 0.0080 0.0225
1999 5/6 152 148 0.0052 0.0148 0.0078 0.0221
1998 3/4 53 51 0.0059 0.0166 0.0077 0.0216

*Total number of laboratories participated in the program each year
**Number of laboratories whose data were used to determine precision estimates

Almost all of the annual precision estimates are smaller than the D2041 precision statements shown in Table 9. This suggests that the D2041 precision statements should be re-established.

Alternatives to Current Standard Test Methods

There are two additional procedures for the determination of Gmm worthy of discussion: 1) CoreLok, and 2) pressure meter method. The CoreLok is a vacuum sealing device that has been discussed previously and has been adapted for the determination of Gmm. The pressure meter concept for asphalt mixtures is based on the pressure meter used for determining the air content of concrete mixtures. The advantages and disadvantages of each of these alternate methods are shown in Table 12.

Table 12 Advantages and Disadvantages of Alternate Methods for Gmm
Method AASHTO and/or ASTM Designation Advantages Disadvantages
Vacuum Sealing or CoreLok (Instrotek) (6) D 6857
  • Simple to perform
  • Less time consuming than current AASHTO or ASTM procedures
  • Potential for reduced variability with more experience
  • Equipment and bag cost
  • No dryback procedure
  • Not accurate for mixtures containing porous aggregate
Pressure Meter (Franko and Lee) (12) None
  • Similar results to AASHTO T 209 for mean and standard deviation
  • Fast test
  • Cumbersome piece of equipment (large and heavy)
  • Equipment needs design changes to be more user friendly
  • Relatively unknown method in asphalt testing
  • Limited research has been conducted

Recent research related to Gmm testing has focused on the evaluation of alternative methods for determining the Gmm and not on improving the accuracy or precision of the current AASHTO or ASTM methods. As shown in Table 12, Franko and Lee (12) adapted the pressure meter test for asphalt mixtures. This test, similar to that used for the measurement of air content in concrete mixtures, was successful at matching AASHTO T 209 with respect to accuracy and precision. The main drawback is the excessive weight and size of the equipment. The test procedure, with additional refinement, appears to be a viable alternative to the current AASHTO and ASTM procedures.

Sholar et al. (6) evaluated a vacuum sealing device, commercially known as the CoreLok, for the determination of Gmm for HMA containing porous limestone aggregate and mixtures containing non-porous granite aggregate. The CoreLok produced results similar to AASHTO T 209 for non-porous aggregate mixtures. However, the CoreLok consistently determined higher Gmm values for asphalt mixtures containing porous aggregate. The researchers determined that this was the result of the CoreLok test method not having a dryback procedure.

As mentioned previously in the Background section, if Gmb is held constant and the Gmm changes by +0.010, the calculated air voids can change about +0.4 percent. The exact change is dependent on the initial Gmm. For example, if the Gmm of a mixture is 2.550 and is increased to 2.560, with a constant Gmb of 2.450, the air voids will increase from 3.92 percent to 4.30 percent, an increase of 0.38 percent. The AASHTO multilaboratory precision is 0.019 for mixtures containing non-porous aggregate. If the Gmm changes by 0.019 (an extreme case not likely to be exceeded more than 5 percent of the time by definition), then the air voids would change by 0.71 percent. The ASTM multilaboratory precision is 0.044 for mixtures containing non-porous aggregate. If the Gmm changes by 0.044, then the calculated air voids would change by 1.63 percent. For mixtures containing porous aggregate, the AASHTO and ASTM multilaboratory precision is 0.055. If the Gmm changes by 0.055, then the air voids would change by 2.03 percent.

As can be seen, the ASTM multilaboratory precision for non-porous aggregate and the AASHTO/ASTM multilaboratory precision for porous aggregate can result in between-laboratory air void values that are very different, yet are still considered valid according to the precision statement. One of the possible reasons for the reduction in precision are the variations allowed when performing the test. One way of addressing this issue is for each agency to conduct an interlaboratory precision study encompassing a representation of contractors, consultants, and agency labs that perform Gmm testing. In addition, each agency could further specify the exact types of testing equipment and procedure to be used, such as specifying a particular type of container and method for determining the mass of the container (container filled with water and weighed or weighed under water).

The Florida Department of Transportation, in an effort to improve the precision of the AASHTO T 209 method, has specified the following: 1) flasks will be the only container allowed, 2) the flasks will be filled with water and weighed and 3) the dryback procedure is required to account for the use of porous aggregate. A precision study was conducted and the following d2s precision values were determined: single operator (0.013) and multilaboratory (0.016). In essence, reducing the options in the test method improved the precision.

 
Updated: 11/17/2011
 

FHWA
United States Department of Transportation - Federal Highway Administration