Rutting of flexible pavement is a widespread problem nationally, including Oklahoma. Rutting is defined as the longitudinal depression along the wheel path due to progressive movement of materials under repeated traffic load. Recent studies have shown that rutting potential of hot mix asphalt (HMA) samples can be evaluated in the laboratory during the design phase of a project using an Asphalt Pavement Analyzer (APA). The rutting susceptibility is evaluated by subjecting HMA samples to moving wheel loads and measuring permanent deformation at selected points along the wheel path as a function of the number of loading cycle. A pressurized rubber hose is placed between the moving wheel and the HMA sample to approximately simulate traffic loading on a pavement in the field. Both rectangular beam and cylindrical samples can be used. A typical test usually involves 8,000 cycles of loading on three beam samples or six cylindrical samples or a combination. The Asphalt Vibratory Compactor (A VC) is used to prepare beam samples, while cylindrical samples are either prepared using a Superpave Gyratory Compactor or an Asphalt Vibratory Compactor. Temperature, magnitude and frequency of moving load, hose pressure and number of cycle can be varied between tests and within the same test, if so desired. Effect of moisture can also be considered by conducting a test under submerged condition. The University of Oklahoma (OU) received funding for a project (Item 2153) to procure an Asphalt Pavement Analyzer and an Asphalt Vibratory Compactor for the Ray Broce Materials Laboratory at OU. This project, funded jointly by the Oklahoma Department of Transportation (ODOT), the Federal Highway Administration (FHWA) and the Oklahoma Asphalt Pavement Association (OAPA), has two major goals: (a) exploratory testing of selected mixes to gain confidence and experience in using APA for evaluation of rut potential; and (b) establishing "baseline data" for selected mixes having low and high rut susceptibility. The following tasks were identified to accomplish the project goals: (i) Procurement and installation of AP A and A VC; (ii) Demonstration and training; (iii) Selection of mixes and collection of materials (ingredients); (iv) Preparation of sample; (v) Exploratory rut testing; (vi) Analysis of exploratory test data; (vii) Conducting tests for baseline data; (viii) Analysis of baseline data; (ix) Preparation of final report. The APA and the AVC were purchased in August 1999. A new electrical panel was installed in the Broce lab to meet the power requirements. Also, compressed air supply in the lab was upgraded to provide compressed air to both pieces of equipment. The installation was completed in September 1999. The manufacturer, Pavement Technologies, Inc. of Georgia, conducted a weeklong demonstration and training in October 1999 that involved calibration of data acquisition system (DAS) for wheel load, horizontal and vertical displacements, DAS setting for beam and cylindrical samples, operation of temperature and preset counter controllers, rubber hose replacement, rut depth measurement (both manual and automated), sample preparation using A VC, safety training, and complete rut and fatigue testing. Three mixes, one for exploratory testing and two for baseline data, were selected in cooperation with ODOT. In addition, ten plant-produced mixes were selected for testing by both the ODOT Materials Division and the OU Team for comparison of results and to address the issue of reliability. Later, another limestone superpave mix was added for extensive testing in developing baseline data. The mix design for exploratory testing of one of the mixes (3012-0APA-99037) was selected from ODOT standards and specifications for type B-insoluble mix. About sixty-four samples were tested for rutting. About half of these samples were prepared using A VC, while Superpave Gyratory Compactor ( SGC) was used for the remaining samples. Two different temperatures (60° and 64°C) and four different asphalt contents (4.5%, 5%, 5.5% and 6%) were used for this series of tests. In the initial stage, over 50 percent samples did not meet the target air void (7 ± 1 % ), particularly for samples prepared using A VC. Sample quality and air void compliance improved with time and experience of the research team. The rut values (8,000 cycles) varied between 2.0 mm and 6.4 mm and the average rut depth were found to be more sensitive to temperature than asphalt content. Although, one of the goals of exploratory testing was to address "reproducibility" of data, this goal could not be achieved partly because of the difficulties in achieving the target air void at the initial stage. Also, it became evident that rut potential evaluation using AP A is not a trivial exercise because of the complexities and difficulties involved in preparing "identical" samples and testing, particularly rut measurement (location, averaging, level of accuracy, sensitivity, etc.). This task w.as completed in June 2000. Based on discussions at the Project Panel Meeting, the project was extended in August 2000 for a year to address the following items that were not addressed in the work plan of the original proposal (Item 2153). (1) Comparison of data for the ten plant- produced mixes with the ODOT data for the same mixes and packaging ofthe data; (2) a better control on achieving the air void requirement; (3) reproducibility of test data; (4) correlation between rutting and resilient modulus, (5) density gradient analysis. An extension for one year is sought to address these issues. Addressing these issues is considered important in enriching our knowledge and confidence in AP A as a tool for performance-based testing of HMA. However, efforts during the past year have focused on the first three items, and equipment has been procured to pursue the remaining two items. Evaluation of rut potentials for ten plant-produced mixes was completed in September 2000. These mixes were selected in cooperation with ODOT Materials Division. Seven of these mixes were type B-insoluble, and three recycled asphalt materials (RAP). For each mix six cylindrical (SGC) and two beam samples were prepared and tested, giving a total of 76 samples. A majority of these samples met the target air void (7 ± 1%). The measured rut depth values varied between 1 mm and 8 mm. The rut depths form beam samples were consistently higher than the corresponding cylindrical samples. Such variations are attributed to sample geometry and rut measurement details. ODOT Materials Division has conducted rut tests using APA on the same ten plant produced mixes. These data was collected from ODOT, and compared with the corresponding data obtained by the OU Team. There was not a significant difference in measured rut depths for the same mix, therefore, additional rut tests were not conducted An effort was made sort out bad data, if there is any. Ranking of these mixes according to their rut potential was completed in December 2000. ODOT participated in the NCAT Test Track project and provided materials and mix designs for two test sections. Tn a meeting, the Oklahoma Asphalt Task Force suggested that the OU Broce Lab participate in rut testing of both mixes. We tested 12 samples (6 SGC cylindrical) x 2 mixes) for rutting. The rut depth from the track will be compared with the AP A data when the field data becomes available. Two gravel mixes (301 l-OK99-63070 and 30I l-OK99-63071) were selected, in cooperation with ODOT, for the development of "baseline data." For each of the two mixes, we tested 24 samples for rutting (1 gradation x I-PG binder x I-aging x I-temperature x 4 asphalt contents x 6 samples (4 SGC cylindrical samples and 2 AVC beam samples). At that stage, it was possible to prepare HMA samples to target air voids fairly accurately. Several samples were tested under wet condition and with different loading conditions as well as hose pressure. The baseline data can be used for calibration of AP A As such, the baseline data are reproducible. Since it is very difficult to produce AP A samples that are identical, addressing the issue of reproducibility is a difficult task. With significant experience over the past years in using AP A and A VC for evaluation of rutting, a duplicate series of tests (24 samples) were conducted to address reproducibility. Later, it was realized that the baseline data was lacking Superpave mixes, so a limestone mix designed in accordance with the superpave method was added with the test matrix. The limestone mix was designed using 13 different asphalt binders (unmodified and modified) that are currently used in Oklahoma. A total of I 04 cylindrical SGC samples were prepared and tested for rutting, and the results statistically analyzed to enrich the baseline database. Twelve Superpave samples were prepared in OU laboratory. Half of these samples were tested for rutting at OU, while the remaining half will be tested at ODOT. Similarly, another 12 samples were prepared at ODOT using the same aggregate and binders used at OU. The rut test values thus obtained was compared to address the issue of repeatability and reproducibility.