NCHRP-RRD-369.pdf

Research Results Digest 369 February 2012 C O N T E N T S Background, 1 Objectives and Scope, 1 Survey of State DOT Laboratories, 2 Experiment Design, 2 Results and Analysis, 7 Findings and Conclusions, 23 BACKGROUND AASHTO T 209, Theoretical Maximum Specifi c Gravity and Density of Bituminous Paving Mixtures, describes a test method for determination of the theoretical maxi- mum specifi c gravity (Gmm) and density of uncompacted hot mix asphalt (HMA).1The Gmmand the density of HMA are fundamen- tal properties whose values are infl uenced by the composition of the HMA mixtures in terms of types and amounts of aggre- gates and asphalt materials. Gmmis used to calculate percent air voids in compacted HMA and to provide target values for the compaction of HMA. Gmmalso is essen- tial when calculating the amount of asphalt binder absorbed by the internal porosity of the individual aggregate particles in HMA. AASHTO T 209 requires application of a vacuum to a sample of HMA loose mix. The vacuum, combined with either manual or mechanical agitation, removes entrapped air in order to accurately determine the Gmm. The Gmmis then used to determine both the air void content and the in-place density of the HMA. In-place density is commonly used in the acceptance and pay-factor de- termination of HMA. Analysis of the AMRL Profi ciency Sample Program data has demonstrated that mechanical agitation provides less variation in test results when compared to manual ag- itation. However, several types of mechan- ical vibratory shakers are commonly used to apply agitation. It was not known if these different devices provide signifi cantly dif- ferent results when compared to one another. In addition, the effects on Gmmvalues of changes in vibration intensity from various settings of the vibrating devices had not been explored. OBJECTIVES AND SCOPE The goal of this research was to eval- uate the effect of using various devices and methods on measured values of Gmm. The specifi c objectives were to (1) compare the Gmmbetween manual and mechanical agitation; (2) investigate the relationship between the measured Gmmand the vibratory AASHTO T 209: EFFECT OF AGITATION EQUIPMENT TYPE ON THEORETICAL MAXIMUM SPECIFIC GRAVITY VALUES This digest summarizes key fi ndings of research conducted in NCHRP Project 10-87(01), “Precision Statements for AASHTO Standard Methods of Test,” by the AASHTO Asphalt Materials Reference Laboratory (AMRL) under the direction of the principal investigator, Dr. Haleh Azari. The digest is an abridgement of the full fi nal report, which is available for download at http:/apps.trb.org/cmsfeed/TRBNetProjectDisplay.asp?ProjectID=3049. Responsible Senior Program Officer: E. T. Harrigan NATIONAL COOPERATIVE HIGHWAY RESEARCH PROGRAM 1AASHTO T 209-10, Theoretical Maximum Spe- cifi c Gravity and Density of Hot Mix Asphalt (HMA). In Standard Specifi cations for Transpor- tation Materials and Methods of Sampling and Testing, 30th ed. American Association of State Highway and Transportation Officials, Washington, D.C., 2010. parameters of the mechanical vibratory tables and determine an optimum vibration intensity of the vi- brating devices; and (3) evaluate the effect on Gmm measurements of several variables, such as the order of placing water and mixture and the period of vac- uum and agitation. The research was conducted in ten major steps: 1. Survey the state highway agencies to deter- mine what specifi c mechanical equipment and methods are currently being used for determining the Gmmof asphalt mixtures. 2. Identify, based on the results of the survey, the most commonly used and the most unique equipment and methods used for measuring Gmm. 3. Select a variety of laboratory-prepared and plant-produced asphalt mixtures for the study, including (a) a fi ne-graded, low traffic volume ( 30 million ESALs) Superpave mix; and (c) a gap-graded or SMA high traffic volume Superpave mix. 4. Measure Gmmusing manual agitation and at several settings of various mechanical agitators. 5. Evaluate the frequency, acceleration, and kinetic energy at various settings of the vi- brating devices. 6. Evaluate the practical and statistical signif- icance of the differences between Gmmvalues obtained using (a) various settings of each vibratory device, (b) zero vibration, and (c) manual agitation, and use this informa- tion to determine the optimum setting of the various devices. 7. Evaluate the practical and statistical signifi - cance of the differences between the highest Gmmvalues from various mechanical devices and manual agitation. 8. Examine the relationship between the vi- bration properties of the vibrating devices and the highest Gmmvalue produced by the device. 9. Investigate the effect on Gmmof the order in which mixture and water are placed in the vacuum fl ask or bowl. 10. Investigate the effect on Gmmof changing the duration of the vacuum and agitation process. SURVEY OF STATE DOT LABORATORIES The survey of the state DOTs included nine questions to identify the candidate devices for the study, how the devices are operated by each state, and whether any of the states test methods deviate from those prescribed by AASHTO T 209. The 35 responses to the survey are organized and presented in Appendix A of the project final report, which can be accessed at http:/apps.trb.org/cmsfeed/ TRBNetProjectDisplay.asp?ProjectID=3049. Based on the results of the survey, the most com- monly used mechanical agitators were selected for the laboratory experiment so that the research fi nd- ings would apply to the widest number of laborato- ries. Several unique setups also were selected to compare the application of non-typical methods to typical methods. EXPERIMENT DESIGN A laboratory experiment was designed to mea- sure Gmmof various mixture types using different de- vices and a variety of agitation levels. The experiment also investigated the effects on Gmmof factors such as the order in which water and mixture are placed in the pycnometer and the vacuum and agitation duration. Test Apparatus and Setup Seven devices were selected for investigation, as follows: 1. Humboldt Vibrating Table (H-1756); 2. Gilson Vibro-Deaerator (SGA-5R); 3. Syntron Vibrating Table (VP-51 D1); 4. Orbital Shaker Table (SHKE 2000); 5. HMA Vibrating Table (VA-2000); 6. Aggregate Drum Washer (with vacuum lid); and 7. Corelok Vacuum Sealing Device. Table 1 provides a brief description of each unit. The Humboldt, Gilson, and HMA tables were selected because together they make up more than 80% of the devices used by the state laboratories. Despite being less common, the Orbital shaker (similar to the Barnstead shaker), Aggregate washer, and Corelok offered unique features and thus the op- portunity to investigate differences between these devices and the more common setups. The Syntron shaker was selected because it is used with a unique setup by the Minnesota DOT. 2 The setup used for measuring Gmmincludes an agitator, vacuum container, a vacuum bowl or vacuum flask (pycnometer), a balance, a vacuum pump, a moisture trap, a vacuum measurement device, a ma- nometer, a bleeder valve, a thermometric device, a water bath, and a drying oven that conforms to the re- quirements of Sections 6.2 to 6.11 of AASHTO T 209. Vibratory frequency and amplitude measure- ments were made using a triaxial accelerometer, a signal conditioner, and SignalView computer soft- ware. An accelerometer produces an electrical sig- nal that is a function of mechanical vibration. A sig- nal conditioner obtains the signal voltage and acts as an interface between the accelerometer and the com- puter, which processes and displays the signals. The accelerometer was attached to the top of the vacuum container lid with wax adhesive to capture the frequency and acceleration of vibration. The fre- quency measurements were recorded to the nearest 0.1 Hz and acceleration measurements were recorded to the nearest 0.01 m/s2in vertical, horizontal, and perpendicular axes. Looking down at the container from the top, the x-axis extended from the left to the right of the device, the y-axis perpendicular to the x- axis forming a plane parallel to the table, and the z-axis perpendicular to the x-y plane. Specimen Preparation Test specimens were either prepared in the lab- oratory or acquired from the fi eld. Dense-graded 4.75-mm, 12.5-mm, 25.0-mm, and 37.5-mm nomi- nal maximum aggregate size (NMAS) mixtures were prepared in the laboratory. Dense-graded 9.5-mm and 3 Table 1 Description of the devices selected for the refi nement of AASHTO T 209 study. DeviceManufacturerAgitation TypeDescription Humboldt Vibrating Table (H-1756) Gilson Vibro-Deaerator (SGA-5R) Syntron Vibrating Table (VP-51 D1) Orbital Shaker Table (SHKE 2000) HMA Vibrating Table (VA-2000) Aggregate Drum Washer (with vacuum lid) Corelok (vacuum sealing device) Humboldt Mfg. Co. Gilson Co., Inc. FMC Technologies, Inc. Thermofi sher Scientifi c HMA Lab Supply Karol-Warner Co. InstroTek, Inc. Vibratory Vibratory Vibratory Orbital Vibratory Rotary No agitation (vacuum seal- ing method) The unit has a dial for adjusting the fre- quency and amplitude of vibration. Different intensities are indicated by numbers on the dial from 1 to 10. The unit has a dial for adjusting the fre- quency and amplitude of vibration. Different intensities are indicated by bars with different thicknesses. No number is associated with the bars. The unit comes with a dial-rheostat for adjusting the amplitude of the vibration. The dial-rheostat is part of a separate control box, which allows for remote control if desired. The unit has an adjustable knob that controls the speed of the shaker plat- form in an orbital pattern in the range of 0 to 350 rpm. The unit has a fi xed intensity. This table was the most frequently used by the state DOTs. The unit rotates slowly at the rate of 25 rpm and tumbles the loose mixture while the vacuum is applied. The unit vacuum-seals the loose asphalt mixture in a plastic bag. 19.0-mm NMAS mixtures were obtained from con- struction sites at the National Institute of Standards and Technology, Gaithersburg, Maryland, and gap- graded stone matrix asphalt (SMA) 9.5-mm, 19.0-mm, and 25.0-mm NMAS mixtures were obtained at construction sites in Richmond, Virginia. The mix- ture designs of the dense-graded laboratory-prepared and plant-produced mixtures are provided in Table 2; however, the mixture designs for the SMA mixtures were not available from the contractor. Plant-produced samples were obtained in con- formance to the requirements of AASHTO T 168 and stored in sealed boxes until the time of testing.2 To prepare the plant mixtures for testing, they were fi rst heated in their boxes at 135 ± 5°C (275 ± 9°F) for about 2 hours. The materials were then worked until a loose mixture condition was obtained. Me- chanical splitter and quartering methods were used to split the mixtures to the appropriate size for test- ing in accordance with AASHTO R 47.3Mixtures were then dried in the oven at 105 ±5°C (221 ±9°F) to constant mass. HMA particles were further sepa- rated by hand so that the particles of the fi ne aggre- gate portion were no larger than 6.3 mm (14in.). The mixtures were then cooled to room temperature before weighing and testing. The laboratory mixtures were designed accord- ing to the Superpave mix design procedure. Non- 4 Table 2 Mix designs of the dense-graded laboratory-prepared and plant-produced mixtures. Laboratory-Prepared MixturesPlant-Produced Mixtures 4.75-mm12.5-mm25.0-mm37.5-mm9.5-mm19.0-mm PercentPercent PercentPercentPercentPercent Sieve (mm)Passing (%)Passing (%)Passing (%)Passing (%)Passing (%)Passing (%) 50.00100100100100100100 37.5010010010097100100 25.001001009791100100 19.00100100867810098 12.5010092715910087 9.501007863459474 4.75935245295337 2.36673429193327 1.18442218122220 0.6023151181415 0.301611751010 0.151175477 0.0758.04.44.43.665.1 AC %5.85.24.03.65.24.4 D. B. Ratio1.50.91.01.11.181.23 2AASHTO T 168-03, Sampling Bituminous Paving Mixtures. In Standard Specifi cations for Transportation Materials and Methods of Sampling and Testing,30th Ed. (CD-ROM), Amer- ican Association of State Highway and Transportation Offi- cials, Washington, D.C., 2010. 3AASHTO R 47-08, Reducing Samples of Hot Mix Asphalt (HMA) to Testing Size. In Standard Specifi cations for Trans- portation Materials and Methods of Sampling and Testing, 30th ed. American Association of State Highway and Trans- portation Officials, Washington, D.C., 2010. absorptive limestone-dolomite aggregate and PG 64-22 asphalt were mixed at 157°C (315°F) and short-term conditioned for 2 hours at 145°C (293°F) according to AASHTO R 30.4The mixtures then were separated by hand so that the particles of the fi ne aggregate portion were not larger than 6.3 mm. Samples then were cooled to room temperature before weighing and testing. The 4.75-mm and 9.5-mm mixtures were pre- pared in 1,500-g batches. The 12.5-mm mixtures were prepared in 2,000-g batches, and the 19-mm and 25-mm mixtures were prepared in 2,500-g batches. The 37.5-mm mixture also was prepared in 2,000-g batches given that the 4,000-g batch weight required by AASHTO T 209 for 37.5-mm and larger mixes could not fit in the flask or pycnometer. In this respect, four 2,000-g specimens of 37.5-mm mixture were tested, combined into two pairs, and the weight measurements from each of the two specimens per pair were added and served as values for one replicate. Measurement of Test Data Gmmmeasurements using vibratory, orbital, and rotary devices were conducted following AASHTO T 209. The cooled, separated particles of asphalt mixture were placed in a tared vacuum container, and the dry mass of the sample was recorded. A suf- fi cient amount of 25°C (77°F) distilled water then was added to cover the sample completely. A deviation from the AASHTO T 209 test method was conducted on several mixtures in which the specifi ed weight of the dry sample was added to the fl ask or pycnometer after water was placed in the con- tainer. The purpose of this deviation was to examine the effect on the release of airand thus on the Gmmof the order of placement of mixture and water. After adding 0.001% of wetting agent, the container or fl ask was sealed and subjected to vibration at 27.5 ±2.5 mm Hg of vacuum for 15 minutes. For three of the mixtures, agitation-vacuum times of 10 minutes, 20 minutes, and 25 minutes also were used with the Gilson vibratory device to examine the effect on Gmm of the duration of agitation. Subsequent to the release of the vacuum, the con- tainer was (a) immersed in a distilled-water bath for 10 minutes for mass measurement in water or (b) fi lled with distilled water, kept in the water bath for 10 min- utes, then dried and placed on the scale for mass de- termination in the air. The weight measurements were obtained to the nearest 0.1 g. In addition to the weight measurements, the acceleration and frequency in the x-