U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590
Federal Highway Administration Research and Technology
Coordinating, Developing, and Delivering Highway Transportation Innovations
Featuring developments in Federal highway policies, programs, and research and technology.
|This magazine is an archived publication and may contain dated technical, contact, and link information.|
|Federal Highway Administration > Publications > Public Roads > Vol. 62· No. 1 > Evaluating Accelerated Rut Testers|
Evaluating Accelerated Rut Testers
by Pedro Romero and Kevin Stuart
Pass or fail? This is what most departments of transportation (DOTs) want to know about an asphalt mixture before it is used to build or repair a road. If the mixture passes a rut-resistance test, it can be used on a road, but if it fails, the search continues until a good mixture is found.
Millions of dollars are spent each year on the maintenance and rehabilitation of our nation's vast highway network. To avoid costly pavement failure and to ensure that rehabilitation and maintenance efforts are maximized, state DOTs have adopted the use of accelerated rut testers and/or servo-hydraulic testing devices to help predict pavement performance. Recently, the Federal Highway Administration (FHWA) tested several devices at the Turner-Fairbank Highway Research Center (TFHRC) in McLean, Va., to determine which device offers the most accurate prediction of pavement rutting resistance.
Prediction of pavement rutting resistance starts with mechanical testing of asphalt mixtures. Accelerated rut testing devices - such as the Georgia Loaded-Wheel Tester, the French Pavement Rutting Tester, and the Hamburg Wheel-Tracking Device - try to simulate the distresses caused by many years of exposure to traffic within a matter of hours by rolling a wheel across a sample of asphalt concrete under specific temperature and humidity conditions. These devices apply a fixed load at a fixed temperature and do not account for seasonal changes in traffic, temperature, or environment.
Each of these devices was developed to respond to a particular set of conditions and to meet the needs of a specific region. The problem with these testers is that their performance criteria are based on limited databases obtained from local conditions and are lacking a solid mechanics-based analysis. Performance criteria and pass/fail values are established only after pavements have been built, materials tested, and performances measured over a period of several years. Once this is done, though, the resulting criteria are only valid for the local conditions under which those pavements were built.
Servo-hydraulic equipment offers the option of changing temperature and load configuration; however, tests are still performed under fixed conditions. That is, several tests have to be run to simulate different environments. With these drawbacks in mind, how certain can you be that the accelerated rut tester used by your DOT is really helping you differentiate a mixture that will fail from one with a promising service life?
The DOTs using these devices have either had to develop their own databases from which to establish specific relationships between observed damage in the device and actual performance in the field or had to adapt somebody else's database to their own environment. Neither of these options is a viable solution to this problem. Developing a database takes years (and thus defeats the purpose of accelerated testing), and adapting an existing database to a different environment will most likely yield inaccurate results.
How reliable is a "pass" or "fail" result then? This question can only be answered by a scientific evaluation of the results yielded by various accelerated testers. To find out which testers provide the most accurate results, they need to be evaluated under the same conditions, and one way to do so is by testing materials of known performance.
What made TFHRC the most suitable place for the evaluation of this technology? First, it has 12 test pavements that have been subjected to the same (controlled) testing conditions by an accelerated loading facility (ALF). ALF allows us to see more realistic performances of the mixtures and to compare them to what our laboratory testers predict. Second, TFHRC's Bituminous Mixtures Laboratory (BML) houses most of the commercially available rutting testers under the same roof. Having all equipment onsite is of great value.
As the first of its kind in North America, ALF has the capability of simulating 20 years of traffic loading in six months or less. ALF is a 29-meter-long structural frame containing a moving wheel assembly. The wheel assembly models one-half of a single axle and can apply loads ranging from 44.5 to 100.1 kilonewtons. It travels 18.5 kilometers per hour over a 9.8-meter test pavement section. To simulate highway traffic, ALF loads pavement in one direction, and the loads can be laterally distributed to simulate the side-to-side wander of trucks. ALF is computer-controlled, permitting operation 24 hours per day, seven days a week.
For a little more than three years, BML researchers prepared and tested specimens with the same material used in the 12 pavements tested by ALF. These specimens were tested in the French Pavement Rutting Tester; the Georgia Loaded-Wheel Tester; the Hamburg Wheel-Tracking Device; and the Superpave Shear Tester, a servo-hydraulic device. The lab results for each tester were then compared to the standard provided by the ALF testing results.
The comparisons showed that no device was clearly better than the others, not even the Superpave Shear Tester (SST). For the most part, all devices were able to separate good from bad mixtures when these had been made with the same aggregate and different binders (ranging from a very soft PG 58 to a stiff PG 78). But, when mixtures with two different aggregate gradations (nominal maximum aggregate of 19 millimeters and 37.5 millimeters) were tested, no device was able to distinguish the mixtures that performed well from those that did not, even though ALF testing showed significant differences in pavement performance.
There are several possible reasons for this. One of them is that most devices do not have proper boundary conditions. When tested in the wheel-tracking devices, lab specimens are always surrounded by steel molds and are resting on a steel base, which is never the case with the testing of real pavements. On the other hand, stress development in laboratory rut testers is never representative of real-life conditions because the size and pressure at the test wheel may not be representative of real wheels used in real pavements.
Specimen size also might have played a role in the lack of perfect agreement between the laboratory results and the ALF results. In some situations, the specimen dimensions were too small in comparison to the materials used, and thus, may not have contained a representative sample of what was actually found in the field. For any test to be valid, the load applied to a specimen should always be in proportion to the specimen size. This was not the case in the most of the evaluated devices, which were designed for fixed conditions.
Our goal in this study was to validate selected mechanical tests used to predict rutting in asphalt pavements by comparing laboratory results with those observed at the ALF. Even though we concluded that none of these devices is perfect, our findings must be interpreted carefully. Using a mechanical test to predict pavement performance is better than not using any at all. However, when conducting a test, be aware that the performance criteria were established for a specific set of conditions and that if those conditions differ from your circumstances, new criteria should be established. Failing to do so could lead to inaccurate pass or fail values.
The same holds true for the Superpave Shear Tester. Even though this device was developed using a mechanics-based approach, there are no performance-prediction models to date that can be used to analyze its data. FHWA has a contract with the University of Maryland to develop such a model(s). Several issues should be carefully looked at prior to conducting a mechanical test. This should be done from an engineering perspective, considering specimen size, load configuration, and temperature because not all tests available in the market will meet the specific requirements of your paving situation. Next time you ask yourself whether the mixture passed or failed, remember that the answer doesn't mean much unless the performance criteria were established for your set of conditions.
For more information, contact Kevin Stuart, (703) 285-2627, email@example.com, or contact Pedro Romero, (703) 285-2911, firstname.lastname@example.org.
Dr. Pedro Romero is the technical manager of the Bituminous Mixtures Laboratory at TFHRC. He has worked as a contractor for FHWA for almost four years. He has a bachelor's degree in civil engineering from the U.S. Coast Guard Academy and a master's and doctorate from Penn State University.
Kevin D. Stuart is a research highway engineer for FHWA and manager of the Bituminous Mixtures Laboratory at TFHRC. He has worked in FHWA's Office of Research and Development for more than 17 years. He has a bachelor's and master's degree in civil engineering from Penn State University.
TFHRC's Bituminous Mixtures Laboratory
The Bituminous Mixtures Laboratory (BML) at the Turner-Fairbank Highway Research Center specializes in research of asphalt pavement mixtures. This lab supports FHWA's efforts to develop, evaluate, and improve materials, mixture-design technology, and performance-based tests for asphalt paving mixtures. The lab's activities are aimed at extending the life and improving the performance of asphalt pavement, reducing vehicle tear-and-wear, and shortening construction delays. Researchers at the BML are constantly evaluating new equipment and test procedures used for the prediction of pavement performance. Currently, further research is being conducted to help in the development of a device that can be used to predict pavement performance as accurately as possible.
The Hamburg Wheel-Tracking Device
The Hamburg Wheel-Tracking Device measures the combined effects of rutting and moisture damage by rolling a steel wheel across the surface of an asphalt concrete slab that is immersed in hot water. This device was developed in the 1970s by Esso A.G. of Hamburg, Germany. The concept is based on a similar British device that had a rubber tire. The city of Hamburg finalized the test method and developed a pass/fail criterion that would guarantee mixtures with very low susceptibility to rutting. The Hamburg Wheel-Tracking Device costs approximately $60,000.
This device was originally used to measure rutting susceptibility. The test required 9,540 wheel passes at temperatures of 40 C or 50 C. After increasing the number of wheel passes to 19,200, Hamburg discovered that some mixtures could deteriorate due to moisture damage shortly after 10,000 passes. The test uses water instead of an environmental air chamber as a means of obtaining the required test temperature.
Test specimens are secured with plaster of paris in reusable steel containers. Each specimen is placed into a container so that its surface is level with the top edge of the container. The container and the specimen are then placed into the wheel-tracking device. The container rests on steel, thus providing a rigid, load-bearing base for the specimen.
This device tests two slabs simultaneously with two reciprocating solid steel wheels. Slabs are 320 millimeters (mm) long and 260 mm wide, and they can be 40, 80, or 120 mm thick. The wheels have a diameter of 203.5 mm and a width of 47 mm. The load is fixed at 685 newtons, and the average contact stress given by the manufacturer is 203.5 megapascals (MPa). Given that the contact area increases with rut depth, contact stress is variable. According to the manufacturer, a contact stress of 0.73 MPa approximates the stress produced by one rear tire of a double-axle truck. The average speed of each wheel is approximately 1.1 kilometers per hour. Each wheel travels approximately 230 mm before reversing direction, and the device operates at 53 ± 2 wheel passes per minute.
The French Pavement Rutting Tester
The French Pavement Rutting Tester tests slabs for permanent deformation using a reciprocating pneumatic tire at a constant temperature of 60 C. This device was developed by Laboratories des Ponts et Chaussees in France to evaluate mixtures that either incorporate materials that may lead to rutting, are subjected to heavy traffic, or have no performance history. This tester costs approximately $85,000.
This machine uses two reciprocating tires to test two slabs simultaneously. Slabs are 500 mm long, 180 mm wide, and they can be 50 mm or 100 mm thick. The wheel load must be equal on both slabs to avoid asymmetric pressures on the tire assembly. However, the two slabs do not have to be replicates. In fact, the manufacturer recommends testing mixtures in random order to account for any variations that might occur over time.
The tires used by the French Pavement Rutting Tester have a diameter of 415 mm and a width of 110 mm. (Note that the width of the slab is only 180 mm.) The standard tire pressure is 0.60 ± 0.03 MPa, and the maximum is 0.71 MPa. It takes approximately 0.1 seconds for the tires to travel from one end of the slab to the center, where the speed is highest. The average speed of each wheel is 7 kilometers per hour. Each wheel travels 380 mm before reversing direction, and the device operates at approximately 67 cycles per minute.
The Georgia Loaded-Wheel Tester
The Georgia Loaded-Wheel Tester (GLWT) measures rutting susceptibility by rolling a steel wheel across a pressurized hose positioned on top of an asphalt concrete beam at a temperature of 41 C. GWLT was developed by the Georgia Department of Transportation and has been refined several times. The model used at TFHRC costs approximately $10,000.
Beams are first aged for 24 hours at room temperature and then for an additional 24 hours at a temperature of 40.6 C. After aging, one beam is positioned in GWLT, and a stiff, 29-mm-diameter rubber hose is positioned on top of it. The hose is pressurized with air at 0.69 MPa. A steel wheel that is loaded with weights rolls back and forth on top of the hose for 8,000 cycles to create a rut. During the test, the sides of the beam are confined by steel plates, except for the top 12.7 mm. The wheel travels approximately 330 mm at an average speed of two kilometers per hour before reversing in direction. The device operates at 33 cycles or 67 passes per minute. Newer models are available with slightly different configurations.
The Superpave Shear Tester
The Superpave Shear Tester (SST) is a closed-loop feedback, servo-hydraulic system that imparts axial and shear loads and confinement pressure to asphalt concrete specimens at controlled temperatures. The response of asphalt concrete to these loads is used in performance-prediction models such as Superpave. This system was developed under the Strategic Highway Research Project. The model used at TFHRC costs approximately $230,000.
SST consists of six components: (1) The test-control system consists of hardware and software (pre-programmed algorithms) that control load applications and acquire data during a test. (2) The testing chamber contains a rigid reaction frame that ensures precise specimen displacement measurements, and a shear table that holds specimens during testing and can impart shear loads. The chamber can host specimens up to 200 mm in diameter and height. (3) Its temperature is computer-controlled by an environmental system that keeps temperature constant during testing (within a range of 0 C to 70 C). (4) The hydraulic system provides the required force to load the specimen for different test conditions. A Series 3410 hydraulic motor powers two actuators, each with a capacity of approximately 32 kN. The vertical actuator applies the axial force to the test specimen, and the horizontal actuator drives the shear table, which, in turn, imparts shear loads to the specimen. (5) The air-pressurization system allows pressurization of the specimen at high rates (70 kilopascals per second for Superpave testing). (6) Several linear variable transducers are used to sense changes in the specimen diameter and to measure the deformation of the specimen and the relative motion of the specimen platens. SST has been preprogrammed to conduct six Superpave tests. However, many more test configurations are possible.
Page Owner: Office of Corporate Research, Technology, and Innovation Management
Scheduled Update: Archive - No Update
Technical Issues: TFHRC.WebMaster@dot.gov