National Academies Press: OpenBook
« Previous: Front Matter
Page 1
Suggested Citation:"Report Contents." National Academies of Sciences, Engineering, and Medicine. 2008. Ruggedness Testing of the Dynamic Modulus and Flow Number Tests with the Simple Performance Tester. Washington, DC: The National Academies Press. doi: 10.17226/14200.
×
Page 1
Page 2
Suggested Citation:"Report Contents." National Academies of Sciences, Engineering, and Medicine. 2008. Ruggedness Testing of the Dynamic Modulus and Flow Number Tests with the Simple Performance Tester. Washington, DC: The National Academies Press. doi: 10.17226/14200.
×
Page 2
Page 3
Suggested Citation:"Report Contents." National Academies of Sciences, Engineering, and Medicine. 2008. Ruggedness Testing of the Dynamic Modulus and Flow Number Tests with the Simple Performance Tester. Washington, DC: The National Academies Press. doi: 10.17226/14200.
×
Page 3
Page 4
Suggested Citation:"Report Contents." National Academies of Sciences, Engineering, and Medicine. 2008. Ruggedness Testing of the Dynamic Modulus and Flow Number Tests with the Simple Performance Tester. Washington, DC: The National Academies Press. doi: 10.17226/14200.
×
Page 4
Page 5
Suggested Citation:"Report Contents." National Academies of Sciences, Engineering, and Medicine. 2008. Ruggedness Testing of the Dynamic Modulus and Flow Number Tests with the Simple Performance Tester. Washington, DC: The National Academies Press. doi: 10.17226/14200.
×
Page 5
Page 6
Suggested Citation:"Report Contents." National Academies of Sciences, Engineering, and Medicine. 2008. Ruggedness Testing of the Dynamic Modulus and Flow Number Tests with the Simple Performance Tester. Washington, DC: The National Academies Press. doi: 10.17226/14200.
×
Page 6
Page 7
Suggested Citation:"Report Contents." National Academies of Sciences, Engineering, and Medicine. 2008. Ruggedness Testing of the Dynamic Modulus and Flow Number Tests with the Simple Performance Tester. Washington, DC: The National Academies Press. doi: 10.17226/14200.
×
Page 7
Page 8
Suggested Citation:"Report Contents." National Academies of Sciences, Engineering, and Medicine. 2008. Ruggedness Testing of the Dynamic Modulus and Flow Number Tests with the Simple Performance Tester. Washington, DC: The National Academies Press. doi: 10.17226/14200.
×
Page 8
Page 9
Suggested Citation:"Report Contents." National Academies of Sciences, Engineering, and Medicine. 2008. Ruggedness Testing of the Dynamic Modulus and Flow Number Tests with the Simple Performance Tester. Washington, DC: The National Academies Press. doi: 10.17226/14200.
×
Page 9
Page 10
Suggested Citation:"Report Contents." National Academies of Sciences, Engineering, and Medicine. 2008. Ruggedness Testing of the Dynamic Modulus and Flow Number Tests with the Simple Performance Tester. Washington, DC: The National Academies Press. doi: 10.17226/14200.
×
Page 10
Page 11
Suggested Citation:"Report Contents." National Academies of Sciences, Engineering, and Medicine. 2008. Ruggedness Testing of the Dynamic Modulus and Flow Number Tests with the Simple Performance Tester. Washington, DC: The National Academies Press. doi: 10.17226/14200.
×
Page 11
Page 12
Suggested Citation:"Report Contents." National Academies of Sciences, Engineering, and Medicine. 2008. Ruggedness Testing of the Dynamic Modulus and Flow Number Tests with the Simple Performance Tester. Washington, DC: The National Academies Press. doi: 10.17226/14200.
×
Page 12
Page 13
Suggested Citation:"Report Contents." National Academies of Sciences, Engineering, and Medicine. 2008. Ruggedness Testing of the Dynamic Modulus and Flow Number Tests with the Simple Performance Tester. Washington, DC: The National Academies Press. doi: 10.17226/14200.
×
Page 13
Page 14
Suggested Citation:"Report Contents." National Academies of Sciences, Engineering, and Medicine. 2008. Ruggedness Testing of the Dynamic Modulus and Flow Number Tests with the Simple Performance Tester. Washington, DC: The National Academies Press. doi: 10.17226/14200.
×
Page 14
Page 15
Suggested Citation:"Report Contents." National Academies of Sciences, Engineering, and Medicine. 2008. Ruggedness Testing of the Dynamic Modulus and Flow Number Tests with the Simple Performance Tester. Washington, DC: The National Academies Press. doi: 10.17226/14200.
×
Page 15
Page 16
Suggested Citation:"Report Contents." National Academies of Sciences, Engineering, and Medicine. 2008. Ruggedness Testing of the Dynamic Modulus and Flow Number Tests with the Simple Performance Tester. Washington, DC: The National Academies Press. doi: 10.17226/14200.
×
Page 16
Page 17
Suggested Citation:"Report Contents." National Academies of Sciences, Engineering, and Medicine. 2008. Ruggedness Testing of the Dynamic Modulus and Flow Number Tests with the Simple Performance Tester. Washington, DC: The National Academies Press. doi: 10.17226/14200.
×
Page 17
Page 18
Suggested Citation:"Report Contents." National Academies of Sciences, Engineering, and Medicine. 2008. Ruggedness Testing of the Dynamic Modulus and Flow Number Tests with the Simple Performance Tester. Washington, DC: The National Academies Press. doi: 10.17226/14200.
×
Page 18
Page 19
Suggested Citation:"Report Contents." National Academies of Sciences, Engineering, and Medicine. 2008. Ruggedness Testing of the Dynamic Modulus and Flow Number Tests with the Simple Performance Tester. Washington, DC: The National Academies Press. doi: 10.17226/14200.
×
Page 19
Page 20
Suggested Citation:"Report Contents." National Academies of Sciences, Engineering, and Medicine. 2008. Ruggedness Testing of the Dynamic Modulus and Flow Number Tests with the Simple Performance Tester. Washington, DC: The National Academies Press. doi: 10.17226/14200.
×
Page 20
Page 21
Suggested Citation:"Report Contents." National Academies of Sciences, Engineering, and Medicine. 2008. Ruggedness Testing of the Dynamic Modulus and Flow Number Tests with the Simple Performance Tester. Washington, DC: The National Academies Press. doi: 10.17226/14200.
×
Page 21
Page 22
Suggested Citation:"Report Contents." National Academies of Sciences, Engineering, and Medicine. 2008. Ruggedness Testing of the Dynamic Modulus and Flow Number Tests with the Simple Performance Tester. Washington, DC: The National Academies Press. doi: 10.17226/14200.
×
Page 22
Page 23
Suggested Citation:"Report Contents." National Academies of Sciences, Engineering, and Medicine. 2008. Ruggedness Testing of the Dynamic Modulus and Flow Number Tests with the Simple Performance Tester. Washington, DC: The National Academies Press. doi: 10.17226/14200.
×
Page 23
Page 24
Suggested Citation:"Report Contents." National Academies of Sciences, Engineering, and Medicine. 2008. Ruggedness Testing of the Dynamic Modulus and Flow Number Tests with the Simple Performance Tester. Washington, DC: The National Academies Press. doi: 10.17226/14200.
×
Page 24

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

1.3 Ruggedness Experiments 1.3.1 Background The purpose of ruggedness testing is to improve a test method by determining which controllable testing conditions most influence the results, and establishing limits for their control. A ruggedness evaluation should always precede an interlaboratory study for a test method. The purpose of an interlaboratory study is to establish the precision of a test method. It involves testing of multiple materials in multiple laboratories, and requires a significant commitment of time and resources. If critical testing conditions are not first iden- tified and controlled through a ruggedness evaluation, then an interlaboratory study will likely yield poor precision for the test method. Perhaps more important than a finding of poor precision, is the fact that data from the interlaboratory study is not generally useful for determining how to improve the precision of the test. This was the unfortunate finding of an interlaboratory study that was recently completed for the dynamic modulus test (3). This study identified high vari- ability in dynamic modulus data obtained from several labo- ratories, but was not able to establish reasons for the high variability or to identify procedural changes that would result in more acceptable testing error. By systematically varying testing conditions and quantifying their effect on the meas- ured data, a ruggedness evaluation is able to identify important sources of testing error and help establish limits to reduce testing error to a tolerable level. Since ruggedness testing is a critical part of the development of a test method, efficient statistical designs have been devel- oped and standardized for ruggedness tests. ASTM E1169, Standard Guide for Conducting Ruggedness Tests, describes the partial factorial Plackett-Burnam designs most often used in ruggedness testing. These designs are very efficient for simul- taneously evaluating the effect of changes in a number of operating conditions when there is no interaction between the operating conditions being evaluated. Inherent to this type of statistical design is the assumption that the effect of each of the operating conditions on the test result is independent. Therefore, the observed effect resulting from simultaneous variation of several operating conditions is simply the sum of the individual effects. Since ruggedness testing is concerned with the evaluation of the effect of changes in testing condi- tions and not necessarily the form of the effect, each testing condition is usually evaluated at only two levels. Replication should be included in the design when an estimate of the vari- ance of a single measurement is not known. ASTM D1067, Standard Practice of Conducting a Ruggedness or Screening Program for Test Methods for Construction Mate- rials, describes the two-level, seven-factor design with repli- cation recommended for ruggedness testing for construction materials tests. The factors to be evaluated and their two levels are determined from theoretical considerations or previous experience with the test. For this study, information obtained from testing completed in Project 9-19 and in Phase II of Project 9-29 was used to select the factors and their levels. The selection of the factors and their levels are discussed in detail later in this Chapter. Test data are collected for specific com- binations of the factors and their levels as outlined in Table 1. This table uses the nomenclature from ASTM D1067. The seven factors are designated by letters A through G. Capital letters indicate high levels for the factors while lower case letters indicate low levels. Thus, determination 1 will be made with factors A, B, and E at low levels and factors C, D, F, and G at high levels. With replication, the experiment requires 16 tests, two for each of the specific combinations indicated in Table 1. The order of the tests should be randomized within each replication of the experiment. Analysis of the resulting data is straightforward as described in ASTM D1067. It involves determining effects for each of the factors included in the partial factorial design, and an estimate of the variance of a single measurement. An F-test or linear regression can then be used to assess the statistical signif- icance of the factor effects relative to the variance of a single measurement. The major considerations in the design of a ruggedness test are (1) selection of the factors and their levels, (2) selection of a range of materials or test conditions for the evaluation, and (3) selection of an appropriate number of laboratories to par- ticipate in the ruggedness testing. The experimental design in Table 1 uses seven factors at two levels. This design is consid- ered appropriate for the proposed ruggedness testing for the simple performance tests. ASTM D1067 recommends using three to five materials covering the expected range of proper- ties to be measured in the test. The results from each material are analyzed separately; therefore, 16 measurements are needed for each material included in the evaluation. ASTM E1169 and ASTM D1067 differ on the number of laboratories to be included in the ruggedness testing. ASTM E1169 recommends 3 Determination Number Factor 1 2 3 4 5 6 7 8 A a A A a A A A A B b B B B b B B B C C C C c C C C c D D D D d d D D D E e E E E E E E e F F F F F F F f F G G G G G g G G g Table 1. Experimental design for a two level, seven factor ruggedness test.

using a single laboratory that has experience with the test being evaluated, while ASTM D1067 recommends using three laboratories. Since the data from each laboratory must be evaluated separately, the use of multiple laboratories in the ruggedness testing does not improve the quality of the statis- tical analysis. As stated in ASTM D1067, the primary benefit obtained from the inclusion of multiple laboratories in ruggedness testing is an additional review of the validity of the test method and the need for added clarity in the operating instructions. Two laboratories were included in the rugged- ness experiment. Tests were conducted in AAT’s laboratory using the Interlaken SPT and in the FHWA Mobile Asphalt Laboratory using the IPC Global SPT. 1.3.2 Ruggedness Testing Plan for Dynamic Modulus This section discusses the ruggedness testing plan that was developed for the dynamic modulus test. It discusses the selection of the materials, testing conditions and factors that were included evaluation 1.3.2.1 Materials and Testing Conditions Temperature and loading rate are the two factors that most influence the dynamic modulus of asphalt concrete mixtures. Figure 1 presents a dynamic modulus master curve generated using the reduced testing protocol developed in Phase IV of this project (Temperatures of 4, 20, and 40°C and loading frequencies of 10, 1, 0.1, and 0.01 Hz). Aggregate type and gradation, volumetric properties, and binder grade will result in a shifting of the master curve, but the overall range will not change significantly. As shown, the range of dynamic modu- lus values can be covered using the following temperature and frequency combinations: • High modulus, 4°C at 10 Hz • Intermediate modulus, 20°C at 0.1 Hz • Low modulus, 40°C at 0.01 Hz Project 9-19 has suggested that confined tests may be nec- essary for gap- and open-graded mixtures. It is likely that the sensitivity of dynamic modulus measurements to confining pressure effects will be different for dense compared to gap- and open-graded mixtures. Therefore, two mixtures were used in the ruggedness testing: a 9.5-mm dense-graded mixture with a PG 64-22 binder, and a 12.5-mm Stone Matrix Asphalt (SMA) mixture with a PG 76-22 binder. Since, as discussed below, one of the factors to be considered in the ruggedness evaluation is air versus water for temperature conditioning, a moisture sensitive dense-graded mixture was used. Smaller nominal maximum aggregate size mixtures were selected to minimize testing error associated with specimen preparation and thereby accentuate the planned effects. Table 2 presents mixture proportions for the mixtures used in the ruggedness testing. The dense-graded mixture uses a somewhat moisture sensitive diabase from Northern Virginia having a typical ten- sile strength ratio of 75 percent in AASHTO T283, Standard Method of Test for Resistance of Compacted Bituminous Mix- ture to Moisture Induced Damage. The SMA mixture uses a combination of diabase from Northern Virginia and Lime- stone from West Virginia. Tests were conducted using the 4 1.0E+03 1.0E+04 1.0E+05 1.0E+06 1.0E+07 1.0E-06 1.0E-04 1.0E-02 1.0E+00 1.0E+02 1.0E+04 1.0E+06 Reduced Frequency E* , p si 4 C 20 C 40 C Fit -3.0000 -2.0000 -1.0000 0.0000 1.0000 2.0000 3.0000 0 10 20 30 40 50 Temperature, C Lo g Sh ift F ac to r Selected Testing Conditions Figure 1. Typical dynamic modulus master curve.

three combinations of temperatures and frequencies listed above. 1.3.2.2 Factors and Levels The dynamic modulus test includes a number conditions that require some level of control in order to minimize test- ing error. The sections below discuss the selection of factors and their levels for the ruggedness testing. Temperature. Temperature is the most important factor affecting the mechanical properties of asphalt concrete mix- tures and must be carefully controlled to obtain precise test data. For the SPT, the temperature is controlled by first equil- ibrating the specimen in a separate conditioning chamber to the test temperature. A dummy specimen of the same size as the test specimen with a thermocouple installed at the middle and exposed to the same thermal history as the test specimen is used to determine when temperature equilibrium is achieved. Once the specimen is equilibrated at the test temperature, a maximum time limit has been specified to in- strument the specimen, install it in the test chamber, and have the test chamber return to the test temperature. The current tolerance on the temperature in the equilibration chamber is ±0.5°C from the target temperature. The time limit for transfer is 3 min. Both of these were successfully met in the Phase II evaluation testing that resulted in an acceptable coefficient of variation of 13 percent. In the ruggedness test- ing, the effect of increasing the equilibration tolerance and the specimen transfer time were evaluated, since less stringent control on these factors may reduce the overall testing time. A test temperature tolerance of ±1.0°C and specimen transfer times of 3 and 5 min were investigated. A related factor that will be investigated in the ruggedness testing is the fluid for conditioning the test specimens. Currently air is specified as the fluid in the conditioning chamber, and the specimen equilibration time at each tem- perature may be as long as 4 hours for the temperature sequence of 4, 20, and 40°C recommended in the reduced dy- namic modulus testing procedure developed in Phase IV of this project. However, it is well known that water has better thermal conductivity than air, and the overall time to com- plete the testing could be substantially reduced if the speci- mens could be equilibrated in water baths set to the testing temperatures. For example, the Marshall stability test, AASHTO T245, Standard Method of Test for Resistance to Plastic Flow of Bituminous Mixtures Using Marshall Appara- tus, requires temperature equilibration times of 30 min in water baths and 2 hours in ovens, both set to the specified test temperature. If in the ruggedness testing the dynamic modu- lus is not significantly affected by the use of water as a condi- tioning fluid, it may be possible to complete the testing at all three temperatures required for master curves in a single day. Air versus water as conditioning fluids was, therefore, in- cluded in the ruggedness testing program. Loading rate. Loading rate has a similar effect as temper- ature on the mechanical properties of asphalt concrete. In fact, this is the basis of the time-temperature superposition concept used in the development of dynamic modulus master curves. Although loading rate has a major effect on the mechanical properties of asphalt concrete, it will not be included in the ruggedness testing because the load standard error computed by SPT software is very sensitive to variations in the frequency of the applied load. Limiting the load standard error to 10 per- cent or less ensures that the frequency of the applied load will be the same as the specified loading frequency. Axial strain. Research has shown the dynamic modulus to be sensitive to the applied axial strain, particularly at high temperatures or low frequencies of loading (4). AASHTO TP 62 has a very wide tolerance of 50 to 150 μstrain for the axial strain, which may be partially responsible for the poor test precision reported in the recently completed interlaboratory study for the dynamic modulus test (3). In the SPT, a control loop has been specified with a tolerance of 75 to 125 μstrain, and in the Phase II evaluation the axial strains were controlled within 80 to 110 μstrain. Axial strain level was a factor in the ruggedness testing with the factor levels set at 75 and 125 μstrain as specified in the equipment specifications for the SPT. Confining pressure. Research has also shown the dynamic modulus at high temperatures and low frequencies of loading to be sensitive to confining pressure (4). Neither the Proj- ect 9-19 test methods (1) nor AASHTO TP 62 address confined dynamic modulus testing. Currently the SPT requires control of confining pressure to ±2 percent of the specified value. The maximum confining pressure available in the SPT is 210 kPa; therefore, the maximum deviation from the target is ±4.2 kPa. In the Phase II evaluation, this level of control was easily main- tained by the two devices. The ruggedness testing included 5 Property 9.5 mm Dense 12.5 mm SMA Binder Content, % 5.7 6.5 Sieve Size, mm 19 100 100 12.5 100 97 9.5 91 81 4.75 68 30 2.36 40 19 1.18 31 15 0.6 22 13 0.3 12 12 0.15 7 10 Gradation, % passing 0.075 4.8 8.3 Table 2. Composition of the mixtures.

confined tests with confining pressures of 135 and 140 kPa to verify that the current level of confining pressure control is adequate. Unconfined tests were performed with and without the membrane to determine if the level of confinement pro- vided by the membrane is significant. If tests at multiple con- fining pressures are desired, the procedure will be simplified if the membrane can remain in place during the unconfined testing. End friction reducer. A major assumption in the dy- namic modulus test is that the stresses are distributed uni- formly over the specimen. Friction between the loading platen and the specimen produces shear stresses which result in a deviation from this assumption. The effects of friction can be minimized by using long specimens and making measurements near the middle. The test specimen size for the simple perfor- mance tests was determined in an extensive specimen size and geometry study conducted in Project 9-19 (5). The specimen diameter of 100 mm was selected to provide flow data that are independent of specimen size. The height to diameter ratio of 1.5 was selected to provide dynamic modulus and flow data that are independent of specimen height. In the Project 9-19 specimen size and geometry study, an end friction reducing element consisting of two latex sheets separated by silicon grease was used. The reduction of end friction in these tests was probably a significant factor in the conclusions concern- ing specimen size. The greased latex sheets are not conducive to production testing; therefore, in Project 9-29 Teflon™ sheets were used in the evaluation testing. The type of end friction reducer, greased latex versus Teflon™ was included in the ruggedness evaluation to verify that either approach is acceptable. Specimen properties. Air void content and end paral- lelism are two specimen properties that must be controlled. With available specimen fabrication techniques, an air void tolerance of ±0.5 percent of the target is obtainable with care- ful control. It is desirable to increase the air void tolerance to minimize the number of specimens rejected. The Hirsch model, which was developed to estimate the effect of volu- metric properties on the dynamic modulus can be used to assess the effect of air voids on the dynamic modulus (6). Fig- ure 2 shows the potential error caused by a 1.0 percent change in air voids. As shown the error is dependent on the modulus of the mixture and varies from about 3 percent for low and high modulus values to 9 percent for intermediate modulus values. This analysis shows that variability in specimen air voids is a sig- nificant contributor to the overall test variability and that a high degree of control over air void content is needed. However, the current tolerance of ±0.5 percent is probably the tightest con- trol obtainable using current specimen fabrication techniques. Therefore, air void content was not considered in the rugged- ness testing. The current tolerance of ±0.5 percent should be used until specimen fabrication equipment is improved. Like end friction, the degree of parallelism of the specimen ends affects the distribution of stresses in the specimen. The uniform stress distribution assumed in the analysis of the dynamic modulus data requires smooth, parallel ends. Sawed 6 0 1 2 3 4 5 6 7 8 9 10 1.0E+04 1.0E+05 1.0E+06 1.0E+07 Dynamic Modulus, psi % E rr or D ue to 1 % C ha ng e in A ir Vo id s 4 % Air Voids 7 % Air Voids Figure 2. Estimated testing error for current air void tolerance.

specimen ends are not perfectly smooth, nor parallel. Since the friction reducer helps minimize the effects caused by end roughness, end parallelism is the critical specimen geometry property that must be considered. Based on measurement of a number of specimens, a tolerance of 1.0 degree was estab- lished in Phase I of this project. To meet this tolerance requires careful control of the sawing operation. To verify that this level of control is acceptable, specimens with sawed ends and milled ends were included in the dynamic modulus ruggedness testing program. 1.3.2.3 Summary Table 3 summarizes the factors and factor levels that were in- cluded in the ruggedness testing for the dynamic modulus test. Dynamic modulus tests were performed for each of the combi- nations of material, confinement, temperature, and loading rate listed in Table 4. Confined tests were only performed at high temperatures where past research has shown confining effects to be significant. Since the dynamic modulus is a non-destruc- tive test, the testing program required the fabrication of 32 spec- imens, 16 for each mixture. Tests on these 32 specimens were performed for the four combinations of temperature and con- finement listed in Table 4 in the following order, unconfined at 4°C, unconfined at 20°C, confined at 40°C then unconfined at 40°C. For each temperature/confinement combination, the order of the determinations from Table 1 was randomized. The entire ruggedness testing program was performed in two labo- ratories: AAT’s laboratory using the ITC SPT and FHWA’s Mo- bile Asphalt Laboratory using the IPC SPT. 1.3.3 Ruggedness Testing Plan for the Flow Number Tests This section discusses the ruggedness testing plan that was developed for the flow number test. It discusses the selection of the materials, testing conditions, and factors that were in- cluded in the evaluation 1.3.3.1 Materials and Testing Conditions The ruggedness testing for the flow number test included materials and testing conditions that result in a wide range of permanent deformation properties. It also included tests on dense- and gap-graded mixtures because it is likely that the sensitivity of the flow number test to confining pressure effects will be different for dense- compared to gap-graded mixtures. To evaluate rutting resistance, the flow number test will be performed at a high pavement temperature represen- tative of the project location and pavement layer depth to evaluate the rutting resistance of the mixture. In NCHRP Project 9-33, criteria have been developed for the flow number test based on the 50 percent reliability 7-day average maximum high pavement temperatures computed using the LTPPBind software (7). Table 5 summarizes these temperatures for selected cities (8). Based on these temperatures, mixtures in- 7 Unconfined Tests Confined Tests Factor Low High Low High Equilibrium Temperature X – 1 °C X + 1 °C X – 1 °C X + 1 °C Specimen Transfer Time 3 min 5 min 3 min 5 min Specimen Conditioning Fluid Air Water Air Water Strain Level 75 μstrain 125 μstrain 75 μstrain 125 μstrain Confining Pressure No membrane Membrane 135 kPa 140 kPa Specimen End Parallelism Milled Sawed Milled Sawed Friction Reducer Greased latex Teflon™ Greased latex Teflon™ Table 3. Summary of factors and levels for the dynamic modulus ruggedness test. Temperature, °C/Frequency, Hz Mixture Confinement 4/1.0 20/0.1 40/0.01 Unconfined X X X Dense-graded Confined X Unconfined X X X SMA Confined X Table 4. Materials and conditions for the dynamic modulus ruggedness test. City 50 Percent Reliability Design High Pavement Temperature, °C (8) 98 Percent Reliability High Temperature Grade, Fast Traffic, 3 to 10 million ESAL (8) Atlanta, GA 51 64 Chicago, IL 47 64 Fairbanks, AK 38 52 Fargo, ND 46 64 Houston, TX 52 70 Indianapolis, IN 48 64 Miami, FL 51 64 Oklahoma City, OK 52 70 Phoenix, AZ 58 76 Reno, NV 51 64 Washington, DC 49 64 Table 5. LTPPBind design high pavement temperatures for 50 percent reliability.

corporating PG 64-22 binders should be tested at approxi- mately 50 °C. A temperature of 50 °C was selected for use in the flow number ruggedness testing. The same two mixtures selected for the dynamic modulus ruggedness were used in the ruggedness testing for the flow tests. Table 6 summarizes the testing conditions for the flow tests. Tests were performed on the dense-graded mixture with and without confinement, but only confined tests were per- formed on the SMA mixture. All tests were performed at 50 oC. 1.3.3.2 Factors and Levels Many of the same factors discussed for the dynamic modu- lus ruggedness test were included in the ruggedness testing for the flow number test. The sections below discuss each of these factors. Temperature. The same temperature factors: equilibrium temperature tolerance, transfer time, and conditioning fluid were included in the ruggedness testing for the flow number test. The factor levels were ±1.0 degree for equilibrium tem- perature, 3 min and 5 min for specimen transfer time, and air and water as conditioning fluids. Loading rate. The duration of the load pulse and dwell time between load pulses are important factors affecting the accumulation of permanent deformation in the flow number test. The duration of the load pulse was not included in the ruggedness testing because the load standard error computed by the SPT software is very sensitive to variations in the duration of the load pulse. Limiting the load standard error to 10 percent or less ensures that the load pulse will be sinusoidal with a duration of 0.1 sec. The equipment specifi- cations currently do not include a tolerance on the dwell time between load pulses. It is specified as 0.9 sec, and current computer control systems are very accurate allowing it to be controlled within a millisecond or less. A tolerance should be included in the specification; therefore, the dwell time was in- cluded in the ruggedness testing. The levels for this factor were set at 0.85 and 0.95 sec. Only the IPC equipment had the capability to adjust the dwell time in the flow number test. Deviatoric stress. Research has shown that the flow num- ber test is sensitive to the applied deviatoric stress (9). The equipment specifications currently apply a ±2.0 percent tol- erance on the deviatoric stress. This level of control was taken from other similar tests for asphalt concrete. Deviatoric stress was included in the ruggedness tests with the factor levels set at 135 and 145 kPa for unconfined tests, and 945 and 985 kPa for confined tests. Confining pressure. Research has also shown that the flow number test is sensitive to confining pressure (9). Cur- rently the SPT specification requires control of confining pressure to ±2.0 percent of the specified value. The maximum confining pressure available in the SPT is 210 kPa; therefore, the maximum deviation from the target is ±4.2 kPa. In the Phase II evaluation, this level of control was easily maintained by the two devices. The ruggedness testing included confined tests with confining pressures of 135 and 140 kPa to verify that the current level of confining pressure control is adequate. Contact stress. The contact stress used in the flow num- ber test applies a small creep load to the specimen during the test. The effect of this loading has not been evaluated in past research. A contact stress of 5 percent of the deviatoric stress was recommended in the Project 9-19 test procedures (2). In the ruggedness testing, contact stresses of 3.7 and 7.5 per- cent were evaluated. End friction reducer. A major assumption in the flow number test is that the stresses are distributed uniformly over the specimen. Friction between the loading platen and the specimen produces shear stresses which result in a deviation from this assumption. The effects of friction can be minimized by using long specimens. The test specimen size for the simple performance tests was determined in an extensive specimen size and geometry study conducted in Project 9-19 (5). The specimen diameter of 100 mm was selected to provide flow data that are independent of specimen size. The height to diameter ratio of 1.5 was selected to provide dynamic modulus and flow data that are independent of specimen height. In the Project 9-19 specimen size and geometry study, an end fric- tion reducing element consisting of two latex sheets separated by silicon grease was used. The reduction of end friction in these tests was probably a significant factor in the conclusions concerning specimen size. The greased latex sheets are not conducive to production testing; therefore, in Project 9-29 Teflon™ sheets were used in the evaluation testing. The type of end friction reducer, greased latex versus Teflon™ was included in the ruggedness evaluation to verify that either approach is acceptable. Specimen properties. Air void content and end paral- lelism are two specimen properties that must be controlled. With available specimen fabrication techniques, an air void tolerance of ±0.5 percent of the target is obtainable with care- 8 Mixture Confinement Confining Stress, kPa Deviator Stress, kPa Anticipated Flow Unconfined 0 140 Low Dense-graded Confined 140 965 Moderate SMA Confined 140 965 High Table 6. Mixture and test conditions for the ruggedness testing for the flow number tests.

ful control. It is desirable to increase the air void tolerance to minimize the number of specimens rejected. Project 9-19 included a subset of flow number tests on mixtures with varying asphalt and air void contents (10). Flow number data from this study are plotted in Figure 3 and Figure 4 to show the effects of air voids. Although there is large scatter in the data, the air void content has a large effect over the 4 to 7 per- cent air void range likely to be used in laboratory testing. Using the trend lines shown, a 1 percent change in air voids produces a 56 percent change in the flow number for uncon- fined tests and a 20 percent change in flow number for con- fined tests. Like the dynamic modulus, this analysis shows that 9 10 100 1000 10000 100000 0 2 4 6 8 10 12 14 Air Void Content, % Fl ow N um be r 3.90 % AC 4.55% AC 5.20 % AC 5.90 %AC 1 10 100 1000 10000 0 2 4 6 8 10 12 Air Void Content Fl ow N um be r 3.90% AC 4.55% AC 5.20% AC 5.90% AC Figure 3. Effect of air voids on unconfined flow number [data from Project 9-19 (10)]. Figure 4. Effect of air voids on confined flow number [data from Project 9-19 (10)].

variability in specimen air voids is a significant contributor to the overall test variability and that a high degree of control over air void content is needed. However, the current tolerance of ±0.5 percent is probably the tightest control obtainable using current specimen fabrication techniques. Therefore, air void content was not a factor considered in the ruggedness testing. The current tolerance of ±0.5 percent should be used until specimen fabrication equipment is improved. As discussed for the dynamic modulus test, end parallelism was included as a factor in the ruggedness testing for the flow number test. Different conclusions concerning the effects of end parallelism may be drawn from the small strain dynamic modulus test and the large strain flow number test. Addi- tionally, the platen configurations are different in the two tests. For the dynamic modulus test, a ball joint that allows the top platen to conform to the plane of the specimen is used. For the flow number test, the platens are fixed in a par- allel arrangement. Specimens with sawed ends and milled ends were included in the dynamic modulus ruggedness test- ing program. 1.3.3.3 Flow Number Test Summary Table 7 summarizes the factors and factor levels that were included in the ruggedness testing for the flow number test. As shown, dwell time and contact stress were included only in the unconfined tests. The effect of dwell time was evaluated only with the IPC equipment. The effect of contact stress was evaluated only with the ITC equipment. Flow number tests were performed at 50°C for three combinations of material and confinement: dense-graded, unconfined; dense-graded, confined; and SMA, confined. Since the flow number test is a destructive test, the testing program required the fabrication and testing of 48 specimens, 16 for each mixture/confinement combination. For each mixture/confinement combination, the order of the determinations from Table 7 was randomized. The data for each mixture/confinement combination will be analyzed separately to draw conclusions on appropriate levels of control for the individual factors. The entire ruggedness testing program was performed in two laboratories: AAT’s laboratory using the ITC SPT and FHWA’s Mobile Asphalt Laboratory using the IPC SPT. 1.4 Equipment Effects Experiment Since equipment from multiple vendors will be used in future interlaboratory studies for the SPT, a study was per- formed after the ruggedness testing to quantify differences in data obtained with equipment from various manufacturers. The objective of this experiment was to verify that the same material properties are obtained in the dynamic modulus and flow number tests using devices from different manufacturers. Table 8 presents the design of this experiment. For each test condition, four replicate tests were performed with each device on the 9.5-mm dense-graded mixture. Analysis of variance techniques was used to analyze the data from each column of the experiment in Table 8. This approach assumes homogeneity of variances for data obtained from the various devices. Based on the data collected in Phase II of NCHRP Project 9-29, this is a reasonable assumption. Using 4 repli- cates per cell provides 9 degrees of freedom for the error term, and 2 degrees of freedom for the equipment effect. As shown in Figure 5 this results in an efficient design as a larger number of replicates have only a minor effect on the critical F-statistic used in the analysis of variance to detect the significance of differences caused by equipment effects. Since the dynamic modulus test is nondestructive, only 12 specimens, four for each device were needed to complete the dynamic modulus portion of the study. The flow number, which is a destructive test, required 24 specimens. 10 Unconfined Tests Confined Tests Factor Low High Low High Equilibrium Temperature 49 °C 51 °C 49 °C 51 °C Specimen Transfer Time 3 min 5 min 3 min 5 min Conditioning Fluid Air Water Air Water Dwell Time (IPC only) 0.85 sec 0.95 sec Not included Not included Contact Stress (ITC only) 5 kPa 10 kPa Not included Not included Deviatoric Stress 135 kPa 145 kPa 945 kPa 985 kPa Confining Stress Not included Not included 135 kPa 140 kPa Specimen End Condition Milled Sawed Milled Sawed Friction Reducer Greased latex Teflon™ Greased latex Teflon™ Table 7. Summary of factors and levels for the flow number ruggedness test.

11 Condition Dynamic Modulus Test Flow Number Test Temperature, °C 10 20 35 35 50 50 Confining Stress, kPa 0 0 0 135 0 140 Deviatoric Stress, kPa To obtain 100 µ strain 140 965 Manufacturer Replicates Interlaken 4 4 4 4 4 4 IPC Global 4 4 4 4 4 4 MDTS 4 4 4 4 4 4 Table 8. Equipment effects experiment. 0 1 2 3 4 5 6 7 0 2 4 6 81 3 5 7 9 Replicates Per Cell Cr iti ca l F V al ue , α = 0. 05 Figure 5. Effect of replicates per cell on critical F-statistic for design in Table 8.

12 2.1 Analysis Approach Linear regression is an efficient method for analyzing the ruggedness data. For each combination of mixture, laboratory, temperature, frequency, and confinement, the ruggedness test data can be fit to a linear model of the form: Y = B0 + B1X1 + B2X2 + B3X3 + B4X4 + B5X5 + B6X6 + B7X7 + Error (1) where: Y = measured value X1, X2, X3, X4, X5, X6, X7 = seven factors included in the ruggedness testing B0, B1, B2, B3, B4, B5, B6, B7 = model coefficients Error = model error From this analysis, the statistical significance of the model coefficients can be determined. For statistically significant factors, the model coefficients can then be used to estimate values for each of the factors that will keep their effect below a specified level. 2.2 Dynamic Modulus Test The results of the dynamic modulus ruggedness testing are presented in Appendix A. The dynamic modulus ruggedness experiment included the factors listed in Table 3. The re- sponses measured in the dynamic modulus ruggedness ex- periment are listed in Table 9. These include the measured dynamic modulus and phase angle, and the computed data quality indicators. Regression equations of the form of Equation 1 were de- veloped for each parameter listed in Table 9. The results are summarized in Table 10 through Table 13 for the dynamic modulus and phase angle. Table 10 and Table 11 present re- sults for tests in AAT’s laboratory with the ITC equipment while Table 12 and Table 13 present results for tests in FHWA’s laboratory using the IPC Global equipment. These tables present p-values indicating the significance of the re- gression coefficients for each of the factors included in the ruggedness experiment. The p-value is the probability of re- jecting the null hypothesis when it is in fact true. For this analysis, it is the probability that the regression coefficient for a particular ruggedness factor is zero when the analysis indicates that it is either greater or less than zero. Thus, low p-values indicate the regression coefficient is statistically significant and the ruggedness factor affects the results of the test. The key to analyzing the ruggedness test in this manner is selecting critical p-values above which the regression coeffi- cient is not significant, and it can be concluded that the ruggedness factor does not affect the test result over the range tested. It is important to keep the objective of ruggedness test- ing in mind when selecting critical p-values. The objective of ruggedness testing is to identify those controllable factors that likely affect a test, and to establish levels for their control. This is different from the usual objective of regression analysis, which is to develop a model to predict an outcome. A predic- tive model should only include variables that are highly re- lated to the predicted outcome, so a very low p-value of 0.05 or less is normally used to detect significant variables for pre- diction models. However, for analysis of ruggedness test data, selecting a very low p-value may result in the erroneous con- clusion that one or more of the ruggedness factors does not affect the test result over the range tested and controlling that factor is not important. For this analysis higher critical p-values than used in regression modeling should be selected. A critical p-value of 0.10 was used. In Table 10 through Table 13, factors with p-values less than or equal to 0.10 are shown in bold. The analysis was not performed for the un- confined 40°C data for the dense graded mixture tested in the IPC device because the quality of the data was poor. The modulus measured in the equipment for this condition was below the calibrated limit of the machine. C H A P T E R 2 Results and Analysis of Ruggedness Experiments

2.2.1 Factors Affecting Dynamic Modulus and Phase Angle Table 14 was constructed to combine the results from both mixtures tested in both laboratories. It presents the percentage of times a specific factor was found to be significant. The notes indicate when a factor was significant for only one laboratory or only one material. Table 14 shows that the factors included in the ruggedness experiment were not found to be significant very often indicat- ing that the degree of control provided for the dynamic mod- ulus test by the SPT is reasonable. From this analysis, it is clear that the transfer time, end condition, and friction reducer have little effect on the dynamic modulus and phase angle. The effects of the statistically significant factors are shown in Figure 6 and Figure 7 for the dynamic modulus and phase angle, respectively. In this analysis, a factor was considered significant if it was found to be statistically significant in 25 per- cent or more of the tests. For the factors controlled by the SPT: temperature, strain level, and confinement, these figures show the change in modulus and phase angle over the tolerance range of the SPT. For the user-selected factors: air or water as the conditioning fluid and with or without a membrane for unconfined tests, these figures show the higher modulus or phase angle condition. For example for 40°C confined tests, the dynamic modulus when water is used as the conditioning fluid is 6 percent higher than when air is used. Data on the repeatability of the dynamic modulus test were collected in Phase II of this project (11). The Phase II experi- ment included eight replicates of two mixtures tested by single operators in two laboratories. Laboratory and mixture effects were found to not be significant, therefore, the 32 observa- tions were pooled to obtain estimates of the repeatability of the dynamic modulus and phase angle. The coefficient of vari- ation for the dynamic modulus obtained from this experi- ment was 13 percent and the standard deviation of the phase angle was 1.7 degrees. It is likely that the repeatability of the dynamic modulus test will improve in the future as specimen fabrication techniques are improved and operators become more familiar with the equipment. However due to 13 Parameter Type Dynamic Modulus Material Property Phase Angle Material Property Load Standard Error Data Quality Indicator Deformation Standard Error Data Quality Indicator Deformation Uniformity Data Quality Indicator Phase Uniformity Data Quality Indicator Table 9. Dynamic modulus test data. Dynamic Modulus Phase Angle Factors 4 C 20 C 40 C 40 C Confined 4 C 20 C 40 C 40C Confined Equilibrium Temperature (-1 vs +1 C) 0.12 0.10 0.87 0.71 0.66 0.95 0.37 0.46 Transfer time (3 vs 5 min) 0.44 0.40 0.50 0.73 0.94 0.66 0.43 0.62 Conditioning Fluid (Water vs Air) 0.45 0.74 0.04 0.66 0.97 0.74 0.60 0.39 Strain Level 0.18 0.43 0.03 0.02 0.95 0.15 0.14 0.37 Membrane (No vs Yes) 0.09 0.21 0.52 NA 0.83 0.72 0.49 NA Confinement (135 vs 145 kPa) NA NA NA 0.63 NA NA NA 0.57 End Condition (Mill vs Saw) 0.39 0.49 0.70 0.32 0.84 0.73 0.70 0.39 Friction Reducer (Teflon vs Latex) 0.27 0.09 0.00 0.15 0.98 0.92 0.52 0.73 Dynamic Modulus Phase Angle Factors 4 C 20 C 40 C 40 C Confined 4 C 20 C 40 C 40C Confined Equilibrium Temperature (-1 vs +1 C) 0.59 0.17 0.99 0.66 0.00 0.53 0.28 0.06 Transfer time (3 vs 5 min) 0.50 0.43 0.68 0.79 0.33 0.89 0.91 0.69 Conditioning Fluid (Water vs Air) 0.64 0.65 0.49 0.88 0.83 0.53 0.85 0.97 Strain Level 0.90 0.99 0.35 0.01 0.63 0.94 0.15 0.04 Membrane (No vs Yes) 0.91 0.42 0.33 NA 0.06 0.22 0.06 NA Confinement (135 vs 145 kPa) NA NA NA 0.22 NA NA NA 0.06 End Condition (Mill vs Saw) 0.24 0.92 0.76 0.26 0.66 0.91 0.15 0.17 Friction Reducer (Teflon vs Latex) 0.15 0.64 0.62 0.85 0.16 0.25 0.49 0.96 Table 10. Significance of dynamic modulus ruggedness test factors on dynamic modulus and phase angle for the dense mixture tested in AAT’s Laboratory with the ITC SPT. Table 11. Significance of dynamic modulus ruggedness test factors on dynamic modulus and phase angle for the SMA mixture tested in AAT’s Laboratory with the ITC SPT.

the non-homogeneous nature of asphalt mixtures, it is unlikely that the repeatability will improve to that obtained with the dy- namic shear rheometer (DSR) on homogeneous asphalt binder samples. The coefficient of variation for DSR measurements on original binder samples is 3.4 percent (12). Considering these levels of repeatability, it may be reasonable to expect the coefficient of variation for the dynamic modulus to improve to approximately 8 percent and the standard deviation of the phase angle to improve to 1.5 degrees. Using these limits, the following observations were made concerning the dynamic modulus test: 1. Temperature control of ±0.5°C is adequate. This range re- sults in a change in modulus that is less than 6 percent and a change in phase angle that is less than 1 degree. 2. Confining pressure control of ±2 percent is adequate. This range results in a change in modulus that is less than 1 percent and a change in phase angle that is less than 0.5 degrees. 3. Either air or water can be used as a conditioning fluid. 4. Strain control of ±25 μstrain is adequate for unconfined tests, but not for confined tests. For unconfined tests this range results in a change in modulus that is less than 4 per- cent and a change in phase angle that is less than 1.7 degrees. However for confined tests, the strain control must be im- proved to ±15 μstrain to keep the effect on the modulus below 8 percent. 5. Unconfined tests can not be performed with the membrane in place. Either the membrane adds a level of confinement that significantly affects the modulus and phase angle at high temperatures or since the instrumentation is mounted 14 Dynamic Modulus Phase Angle Factors 4 C 20 C 40 C 40 C confined 4 C 20 C 40 C 40C Confined Equilibrium Temperature (-1 vs +1 C) 0.02 0.00 0.03 0.00 0.06 0.27 Transfer time (3 vs 5 min) 0.90 0.30 0.11 0.73 0.54 0.56 Conditioning Fluid (Water vs Air) 0.05 0.03 0.02 0.07 0.32 0.35 Strain Level 1.00 0.09 0.51 0.53 0.01 0.01 Membrane (No vs Yes) 0.36 0.22 NA 0.00 0.01 NA Confinement (135 vs 145 kPa) NA NA 0.38 NA NA 0.29 End Condition (Mill vs Saw) 0.96 0.43 0.13 0.42 0.28 0.20 Friction Reducer (Teflon vs Latex) 0.91 0.88 0.32 0.34 0.13 0.78 Table 12. Significance of dynamic modulus ruggedness test factors on dynamic modulus and phase angle for the dense mixture tested in FHWA’s Laboratory with the IPC SPT. Dynamic Modulus Phase Angle Factors 4 C 20 C 40 C 40 C confined 4 C 20 C 40 C 40C Confined Equilibrium Temperature (-1 vs +1 C) 0.03 0.00 0.00 0.82 0.15 0.03 0.36 0.42 Transfer time (3 vs 5 min) 0.50 0.81 0.72 0.28 0.59 0.15 0.46 0.37 Conditioning Fluid (Water vs Air) 0.87 0.80 0.45 0.00 0.89 0.43 0.83 0.04 Strain Level 0.49 0.49 0.29 0.30 0.78 0.87 0.12 0.88 Membrane (No vs Yes) 0.85 0.74 0.05 NA 0.02 0.06 0.02 NA Confinement (135 vs 145 kPa) NA NA NA 0.10 NA NA NA 0.04 End Condition (Mill vs Saw) 0.39 0.88 0.34 0.72 0.98 0.90 0.78 0.42 Friction Reducer (Teflon vs Latex) 0.62 0.97 0.89 0.62 0.64 0.49 0.29 0.19 Table 13. Significance of dynamic modulus ruggedness test factors on dynamic modulus and phase angle for the SMA mixture tested in FHWA’s Laboratory with the IPC SPT. Factors Dynamic Modulus Phase Angle Equilibrium Temperature (-1 vs +1 °C) 47 33 Transfer time (3 vs 5 min) 0 0 Conditioning Fluid (Water vs Air) 33 131 Strain Level 27 20 Membrane (No vs Yes) 18 54 Confinement (135 vs 145 kPa) 252 503 End Condition (Mill vs Saw) 0 0 Friction Reducer (Teflon vs Latex) 134 0 Notes: 1 FHWA Laboratory with IPC 2 SMA in FHWA Laboratory with IPC 3 SMA only 4 Dense in AAT Laboratory with ITC Table 14. Percentage of times each ruggedness factor was found to be significant.

outside the membrane, the membrane affects the defor- mation measurements. 2.2.2 Factors Affecting Data Quality Indicators Similar analyses were performed for the data quality indi- cators. The results are summarized in Table 15 and Table 16 for tests in AAT’s laboratory using the ITC equipment and in Table 17 and Table 18 for tests in the FHWA’s laboratory using the IPC equipment. Like the tables for modulus and phase angle, these tables present p-values indicating the sig- nificance of the regression coefficients for each of the factors included in the ruggedness experiment. Again to highlight the important effects, p-values of 0.10 or less are shown in bold. Table 19 presents a summary table with the percentage of times a specific factor was found to be significant. The notes indicate when a factor was significant for only one laboratory or only one material. Like the measured material properties, the data quality indicators were not affected very often by the ruggedness factors. The sections that follow discuss each of the data quality indicators. 15 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 Temperature, +0.5 C Time, min -Water / +Air Strain, +25 microstrain -Without / +With Membrane - Milled / +Sawed -Teflon / +Latex Confinement, +2% Change in Dynamic Modulus, % Confined 40 C 40 C 20 C 4 C Figure 6. Effect of statistically significant ruggedness factors on the dynamic modulus. -4 -3 -2 -1 0 1 2 3 4 5 Temperature, +0.5 C Time, min -Water / +Air Strain, +25 microstrain -Without / +With Membrane - Milled / +Sawed -Teflon / +Latex Confinement, +2% Change in Phase Angle, Degree Confined 40 C 40 C 20 C 4 C Figure 7. Effect of statistically significant ruggedness factors on the phase angle.

16 Load Standard Error Deformation Standard Error Deformation Uniformity Phase Uniformity Factors 4 C 20 C 40 C 40 C Confined 4 C 20 C 40 C 40 C Confined 4 C 20 C 40 C 40 C Confined 4 20 40 40 C Confined Equilibrium Temperature 0.67 0.62 0.66 0.85 0.68 0.05 0.09 0.16 0.01 0.50 0.78 0.71 0.98 0.94 0.69 0.97 Transfer time 0.75 0.54 0.79 0.75 0.82 0.61 0.61 0.73 0.15 0.52 0.69 0.70 0.22 0.81 0.06 0.76 Conditioning Fluid (Water vs Air) 1.00 0.69 0.20 0.73 0.95 0.08 0.54 0.09 0.78 0.94 0.17 0.34 0.28 0.51 0.02 0.38 Strain Level 0.06 0.30 0.05 0.45 0.07 0.13 0.09 0.05 0.01 0.89 0.06 0.74 0.46 0.35 0.00 0.04 Membrane (No vs Yes) 0.34 0.10 0.88 NA 0.25 0.00 0.72 NA 0.21 0.99 0.81 NA 0.28 0.35 0.76 NA Confinement (135 vs 145 kPa) NA NA NA 0.38 NA NA NA 0.41 NA NA NA 0.32 NA NA NA 0.92 End Condition (Mill vs Saw) 0.68 0.32 0.81 0.40 0.78 0.32 0.99 0.46 0.26 0.53 0.21 0.48 0.19 0.31 0.20 0.66 Friction Reducer (Teflon vs Latex) 0.84 0.37 0.56 0.97 0.76 0.01 0.94 0.29 0.31 0.37 0.25 0.59 0.95 0.50 0.12 0.08 Load Standard Error Deformation Standard Error Deformation Uniformity Phase Uniformity Factors 4 C 20 C 40 C 40 C Confined 4 C 20 C 40 C 40 C Confined 4 C 20 C 40 C 40 C Confined 4 20 40 40 C Confined Equilibrium Temperature 0.70 0.59 0.33 0.01 0.64 0.61 0.70 0.74 0.50 0.07 0.59 0.24 0.57 0.62 0.32 0.89 Transfer time 0.65 0.74 0.19 0.13 0.70 0.86 0.07 0.12 0.87 0.22 0.39 0.00 0.07 0.23 0.34 0.49 Conditioning Fluid (Water vs Air) 0.58 0.09 0.84 0.62 0.51 0.37 0.69 0.28 0.62 0.57 0.78 0.53 0.68 0.91 0.41 0.90 Strain Level 0.06 0.22 0.01 0.58 0.08 0.71 0.85 0.30 0.91 0.86 0.80 0.30 0.69 0.13 0.47 0.60 Membrane (No vs Yes) 0.69 0.54 0.62 NA 0.55 0.17 0.41 NA 0.68 0.84 0.94 NA 0.85 0.47 0.41 NA Confinement (135 vs 145 kPa) NA NA NA 0.46 NA NA NA 0.27 NA NA NA 0.00 NA NA NA 0.04 End Condition (Mill vs Saw) 0.62 0.36 0.71 0.16 0.74 0.33 0.89 0.03 0.08 0.19 0.86 0.02 0.92 0.26 0.36 0.52 Friction Reducer (Teflon vs Latex) 0.55 0.29 0.36 0.23 0.71 0.93 0.75 0.30 0.76 0.91 0.24 0.66 0.07 0.33 0.30 0.03 Load Standard Error Deformation Standard Error Deformation Uniformity Phase Uniformity Factors 4 C 20 C 40 C 40 C Confined 4 C 20 C 40 C 40 C Confined 4 C 20 C 40 C 40 C Confined 4 20 40 40 C Confined Equilibrium Temperature 0.33 0.01 0.44 0.98 0.10 0.79 0.90 0.76 0.35 0.33 0.53 0.15 Transfer time 0.93 0.42 0.29 0.17 0.06 0.62 0.05 0.62 0.33 0.57 0.27 0.10 Conditioning Fluid (Water vs Air) 0.38 0.47 0.29 0.52 0.19 0.04 0.74 0.33 0.38 0.38 0.54 0.86 Strain Level 0.18 0.00 0.24 0.33 0.14 0.43 0.26 0.87 0.73 0.46 0.51 0.03 Membrane (No vs Yes) 0.93 0.11 NA 0.65 0.39 NA 0.92 0.81 NA 0.06 0.21 NA Confinement NA NA 0.28 NA NA 0.65 NA NA 0.16 NA NA 0.37 End Condition (Mill vs Saw) 0.54 0.16 0.40 0.08 0.64 0.17 0.07 0.16 0.72 0.84 0.12 0.02 Friction Reducer (Teflon vs Latex) 0.53 0.11 0.30 0.42 0.79 0.34 0.27 0.62 0.34 0.31 0.18 0.17 Load Standard Error Deformation Standard Error Deformation Uniformity Phase Uniformity Factors 4 C 20 C 40 C 40 C Confined 4 C 20 C 40 C 40 C Confined 4 C 20 C 40 C 40 C Confined 4 20 40 40 C Confined Equilibrium Temperature 0.06 0.08 0.00 0.09 0.91 0.31 0.00 0.88 0.67 0.55 0.10 0.83 0.30 0.24 0.67 0.68 Transfer time 0.14 0.26 0.03 0.90 0.22 0.78 0.85 0.13 0.60 0.90 0.68 0.49 0.48 0.43 0.80 0.97 Conditioning Fluid (Water vs Air) 0.16 0.42 0.01 0.00 0.52 0.16 0.01 0.03 0.98 0.10 0.08 0.80 0.67 0.75 0.56 0.77 Strain Level 0.18 0.08 0.00 0.00 0.29 0.04 0.63 0.95 1.00 0.35 0.14 0.45 0.29 0.77 0.79 0.96 Membrane (No vs Yes) 0.67 0.49 0.00 NA 0.26 0.71 0.00 NA 0.24 0.19 0.21 NA 0.09 0.22 0.38 NA Confinement NA NA NA 0.01 NA NA NA 0.76 NA NA NA 0.69 NA NA NA 0.72 End Condition (Mill vs Saw) 0.57 0.97 0.03 0.44 0.83 0.25 0.75 0.44 0.26 0.83 0.90 0.65 0.26 0.54 0.75 0.96 Friction Reducer (Teflon vs Latex) 0.60 0.53 0.19 0.11 0.78 0.67 0.83 0.55 0.49 0.41 0.51 0.62 0.33 0.65 0.73 0.33 Table 15. Significance of dynamic modulus ruggedness test factors on data quality indicators for the dense mixture tested in AAT’s Laboratory with the ITC SPT. Table 16. Significance of dynamic modulus ruggedness test factors on data quality indicators for the SMA mixture tested in AAT’s Laboratory with the ITC SPT. Table 17. Significance of dynamic modulus ruggedness test factors on data quality indicators for the dense mixture tested in FHWA’s Laboratory with the IPC SPT. Table 18. Significance of dynamic modulus ruggedness test factors on data quality indicators for the SMA mixture tested in FHWA’s Laboratory with the IPC SPT.

2.2.2.1 Load Standard Error The load standard error is a measure of how well the SPT applies a sinusoidal loading to the specimen. During Phase II of this project, a maximum load standard error of 10 percent was associated with good quality data (11). For the load stan- dard error, transfer time, end condition, and friction reducers did not appear to affect the results. The effects of the remain- ing factors are shown in Figure 8. Although some ruggedness factors were statistically significant, it is clear from Figure 8 that these have only a minor effect on the load standard error when the allowable range of 10 percent is considered. 2.2.2.2 Deformation Standard Error The deformation standard error is a measure of how close the deformations measured in the SPT are to a sinusoid. During Phase II of this project a maximum deformation standard error of 10 percent was associated with good qual- ity data (11). For the deformation standard error, transfer time, confinement, end condition, and friction reducers did not appear to affect the results. The effects of the remaining factors are shown in Figure 9. From Figure 9, it is clear that the deformation standard error for high-temperature tests is higher when water is used as the conditioning fluid and when the unconfined dynamic modulus is measured with the membrane in place. These two conditions should, therefore, be avoided. 2.2.2.3 Deformation Uniformity The deformation uniformity is a measure of how close the individual deformation measurements made on a sample 17 Load Standard Error Deformation Standard Error Deformation Uniformity Phase Uniformity Equilibrium Temperature 40 27 20 0 Transfer time 71 13 13 202 Conditioning Fluid (Water vs Air) 20 33 131 72 Strain Level 53 40 132 20 Membrane (No vs Yes) 182 18 0 181 Confinement 251 0 0 252 End Condition (Mill vs Saw) 71 13 20 71 Friction Reducer (Teflon vs Latex) 0 72 0 202 Notes: 1 SMA Mixture in FHWA Laboratory with IPC 2 AAT Laboratory with ITC Table 19. Percentage of times each ruggedness factor was found to be significant. -10 -8 -6 -4 -2 0 2 4 6 8 10 Temperature, +0.5 C Time, min -Water / +Air Strain, +25 microstrain -Without / +With Membrane - Milled / +Sawed -Teflon / +Latex Confinement, +2% Change in Load Standard Error, Percent Confined 40 C 40 C 20 C 4 C Figure 8. Effect of statistically significant ruggedness factors on the load standard error.

agree with one another. During Phase II of this project a maximum deformation uniformity of 20 percent was associ- ated with good quality data (11). For the deformation uni- formity, only the temperature and end condition were found to be statistically significant. Figure 10 shows the effect of these two factors on the deformation uniformity. The tem- perature effect is small considering the allowable value of 20 percent for good quality data. The end condition effect is larger, but not consistent over the temperature ranges. For unconfined tests at 4°C and confined tests at 40°C, the data from milled ends are more variable. On the other hand, the data from the sawed ends are more variable in the un- confined tests at 20°C and 40°C. Thus, the effects of the ruggedness factors on the deformation uniformity are small and not consistent. 2.2.2.4 Phase Uniformity The phase uniformity is a measure of how close the indi- vidual phase angle measurements made on a sample agree with one another. During Phase II of this project a maximum phase uniformity of 3 degrees was associated with good quality 18 -10 -8 -6 -4 -2 0 2 4 6 8 10 Temperature, +0.5 C Time, min -Water / +Air Strain, +25 microstrain -Without / +With Membrane - Milled / +Sawed -Teflon / +Latex Confinement, +2% Change in Deformation Standard Error, Percent Confined 40 C 40 C 20 C 4 C Figure 9. Effect of statistically significant ruggedness factors on the deformation standard error. -20 -15 -10 -5 0 5 10 15 20 Temperature, +0.5 C Time, min -Water / +Air Strain, +25 microstrain -Without / +With Membrane - Milled / +Sawed -Teflon / +Latex Confinement, +2% Change in Deformation Uniformity, Percent Confined 40 C 40 C 20 C 4 C Figure 10. Effect of statistically significant ruggedness factors on the deformation uniformity.

data (11). For phase uniformity, temperature, conditioning fluid, and end condition did not appear to affect the results. Figure 11 shows the effect of the remaining factors on the phase uniformity. The effects are generally small and not consistent over the testing conditions, except for the membrane effect. Phase angles are more variable in unconfined tests when the membrane is used. 2.2.3 Summary Table 20 summarizes the results of the analysis of the rugged- ness test data for the dynamic modulus test. For statistically significant ruggedness factors, Table 20 presents the effect of the factor on the measured modulus and phase angle and the data quality indicators. Table 20 also presents acceptable values based on anticipated test variability. The following conclu- sions were drawn for each of the ruggedness factors: 1. Equilibrium temperature. The current temperature control of ±0.5°C in SPT is acceptable. Temperature changes over this level are expected to result in less than a 6 percent change in the dynamic modulus and less than a 0.6 degree change in the phase angle. 2. Transfer time. The transfer time over the range of 3 to 5 min was not found to be a significant factor in the measured material properties, and had only a minor effect on the data quality. The transfer time can be increased to 5 min. 3. Conditioning fluid. The use of water as a condition- ing fluid results in significantly poorer quality test data for confined test conditions. Air should, therefore, be used as the 19 -3 -2 -1 0 1 2 3 Temperature, +0.5 C Time, min -Water / +Air Strain, +25 microstrain -Without / +With Membrane - Milled / +Sawed -Teflon / +Latex Confinement, +2% Change in Phase Uniformity, Degree Confined 40 C 40 C 20 C 4 C Figure 11. Effect of statistically significant ruggedness factors on the phase uniformity. Factors Control Dynamic Modulus Phase Angle Load Standard Error Deformation Standard Error Deformation Uniformity Phase Uniformity Equilibriu m Tem perature 0.5 °C < 6 % < 0.6 ° < 0.5 % < 0.5 % < 2 % NS Transfer Ti me 3 versus 5 min NS NS NS NS NS < 0.5 ° Conditioning Fluid Air versus Water < 6 % < 1 ° < 0.6 % < 0.5 % unconfine d <7 % confined NS NS Strain Level 25 μstrain < 4 % unconfined < 12 % confined < 1.6 ° < 0.6 % < 0.5 % unconfine d <1.7 % confined NS < 0.7 ° Membrane Without versus With < 11 percent 3.7 ° < 0.6 % < 4.6 % NS < 1.1 ° Confinem ent 2 % < 0.8 % < 0.2 ° < 0.2 % NS NS < 0.1 ° End Condition Milled versus Sawed NS NS NS NS < 4.7 % NS Friction Reducer Greased Latex versus Teflon NS NS NS NS NS < 0.9 ° Acceptable 8 % 1.7 ° 5 % 5 % 10 % 1.5 ° NS = not statistically significant Table 20. Summary of the effect of ruggedness test factors on material properties and data quality indicators in the dynamic modulus test.

conditioning fluid. If the specimens are to be conditioned in a water bath, they should be sealed in plastic to keep the water from penetrating the specimen. 4. Strain level. The current strain control of ±25 μstrain is acceptable for unconfined tests. However, the strain con- trol should be improved to ±15 μstrain to accommodate con- fined testing which may be necessary for some mixture types. 5. Membrane. Unconfined tests should not be performed with the membrane on the specimen. The membrane in- creases the dynamic modulus and phase angle for moderate to high temperature tests. It also significantly reduces the quality of the deformation and phase angle data. 6. Confinement. The current confining pressure control of ±2 percent is acceptable in confined tests. Over this range of control, the dynamic modulus and phase angle are ex- pected to vary by 0.8 percent and 0.2 degrees, respectively. 7. End condition. There was no significant difference in the measured material properties between milled specimen ends and sawed specimen ends. The effect of end condition on the data quality was small and not consistent. The use of sawed spec- imen ends is acceptable for dynamic modulus tests in the SPT. 8. Friction reducer. There was no significant difference in the measured material properties between greased latex and Teflon™ as the end friction reducer. The effect of the fric- tion reducer on the data quality was small and not consistent. The use of either greased latex of Teflon™ as the end friction reducer is acceptable for dynamic modulus tests in the SPT. 2.3 Flow Number Test The results of the flow number ruggedness testing are pre- sented in Appendix B. The flow number ruggedness experi- ment included the factors listed in Table 7. The responses measured in the flow number ruggedness experiment included the flow number and the permanent strain after selected num- ber of load cycles. Flow did not occur in all of the confined tests. Table 21 summarizes the data that was analyzed for the flow number tests. Regression equations of the form of Equation 1 were de- veloped for each of the marked cells in Table 21. The results are presented in Table 22 and Table 23 for the unconfined tests on the dense graded mixture; Table 24 and Table 25 for the confined tests on the dense graded mixture; and Table 26 and Table 27 for the confined tests on the SMA mixture. Each table presents p-values indicating the significance of the re- gression coefficients for each of the factors included in the ruggedness experiment. As discussed previously for the dy- namic modulus, low p-values indicate the regression coeffi- cient is statistically significant and the ruggedness factor affects the results of the test. A critical p-value of 0.10 was used in this analysis. Factors with p-values equal to or less than 0.1 are shown in bold in Table 22 through Table 27. Table 28 and Table 29 were constructed to combine the re- sults for the tests in both laboratories. These tables present the percentage of times a specific factor was found to be significant. Table 28 presents the results for the unconfined tests, while Table 29 presents the results for the confined tests. The sec- tions that follow discuss the results for the flow number and the measured permanent strains. 2.3.1 Factors Affecting Flow Number In order to analyze the flow number, all specimens tested in both laboratories must exhibit flow. Flow occurred in all of the unconfined tests on the dense-graded mixture and about 25 percent of the confined tests on the dense-graded mixture. 20 Parameter Dense Unconfined Dense Confined SMA Confined Flow Number X εp at 500 cycles X X X εp at 1000 cycles X X X εp at 2000 cycles X X X εp at 5000 cycles X εp at 8000 cycles X Table 21. Flow number test data. Permanent Strain at Factors Flow Number 500 cycles 1000 cycles 2000 cycles Equilibrium Temperature (-1 vs +1 C) 0.06 0.51 0.22 0.05 Transfer time (3 vs 5 min) 0.26 0.87 0.93 0.66 Conditioning Fluid (Water vs Air) 0.64 0.06 0.02 0.02 End Condition (Mill vs Saw) 0.29 0.02 0.02 0.04 Friction Reducer (Teflon vs Latex) 0.02 0.94 0.42 0.03 Axial Stress (135 vs 145 kPa) 0.14 0.80 0.43 0.12 Contact Stress (5 vs 10 kPa) 0.63 0.30 0.27 0.35 Permanent Strain at Factors Flow Number 500 cycles 1000 cycles 2000 cycles Equilibrium Temperature (-1 vs +1 C) 0.03 0.31 0.23 0.33 Transfer time (3 vs 5 min) 0.76 0.14 0.26 0.44 Conditioning Fluid (Water vs Air) 0.74 0.22 0.15 0.18 End Condition (Mill vs Saw) 0.99 0.07 0.26 0.79 Friction Reducer (Teflon vs Latex) 0.39 0.29 0.25 0.22 Axial Stress (135 vs 145 kPa) 0.98 0.77 0.74 0.64 Dwell Time (0.85 vs 0.95 sec) 0.30 0.69 0.67 0.55 Table 22. Significance of flow number ruggedness test factors for unconfined tests with the ITC SPT on the dense mixture. Table 23. Significance of flow number ruggedness test factors for unconfined tests with the IPC SPT on the dense mixture.

None of the SMA specimens exhibited flow in the confined tests. Only temperature and end friction reducer were found to have a statistically significant effect on the flow number in un- confined tests. Figure 12 shows the effect of these two factors. For temperature the flow number decreases by 7.5 percent for an increase in temperature of 0.5°C while the flow number is 20 percent higher when Teflon™ is used as the end friction reducer. As expected, increasing temperature decreases the flow number. Apparently the Teflon™ end friction reducer is less effective than the greased latex membranes resulting in greater end friction and a higher flow number. 21 Permanent Strain at Factors Flow Number 500 cycles 1000 cycles 2000 cycles Equilibrium Temperature (-1 vs +1 C) 0.02 0.01 0.00 Transfer time (3 vs 5 min) 0.14 0.07 0.02 Conditioning Fluid (Water vs Air) 0.37 0.43 0.53 End Condition (Mill vs Saw) 0.12 0.12 0.05 Friction Reducer (Teflon vs Latex) 0.01 0.00 0.00 Axial Stress (945 vs 985 kPa) 0.07 0.05 0.02 Confining Stress (135 vs 145 kPa) 0.21 0.09 0.02 Table 24. Significance of flow number ruggedness test factors for confined tests with the ITC SPT on the dense mixture. Permanent Strain at Factors Flow Number 500 cycles 1000 cycles 2000 cycles 5000 cycles 8000 cycles Equilibrium Temperature (-1 vs +1 C) 0.29 0.33 0.48 0.88 0.87 Transfer time (3 vs 5 min) 0.34 0.35 0.38 0.34 0.40 Conditioning Fluid (Water vs Air) 0.79 0.94 0.88 0.75 0.59 End Condition (Mill vs Saw) 0.77 0.94 0.92 0.80 0.71 Friction Reducer (Teflon vs Latex) 0.02 0.07 0.21 0.42 0.42 Axial Stress (945 vs 985 kPa) 0.88 0.85 0.68 0.74 0.98 Confining Stress (135 vs 145 kPa) 0.25 0.25 0.37 0.65 0.37 Permanent Strain at Factors Flow Number 500 cycles 1000 cycles 2000 cycles 5000 cycles 8000 cycles Equilibrium Temperature (-1 vs +1 C) 0.70 0.71 0.52 0.63 0.77 Transfer time (3 vs 5 min) 0.84 0.86 0.50 0.31 0.41 Conditioning Fluid (Water vs Air) 0.98 0.40 0.26 0.49 0.80 End Condition (Mill vs Saw) 0.71 0.82 0.93 0.66 0.89 Friction Reducer (Teflon vs Latex) 0.00 0.00 0.01 0.08 0.07 Axial Stress (945 vs 985 kPa) 0.19 0.24 0.28 0.48 0.74 Confining Stress (135 vs 145 kPa) 0.05 0.05 0.14 0.40 0.53 Table 26. Significance of flow number ruggedness test factors for confined tests with the ITC SPT on the SMA mixture. Table 27. Significance of flow number ruggedness test factors for confined tests with the IPC SPT on the SMA mixture. Factors Flow Number 500 cycles 1000 cycles 2000 cycles Equilibrium Temperature (-1 vs +1 C) 100 0 0 50 Transfer time (3 vs 5 min) 0 0 0 0 Conditioning Fluid (Water vs Air) 0 50 50 50 End Condition (Mill vs Saw) 0 100 50 50 Friction Reducer (Teflon vs Latex) 50 0 0 50 Axial Stress (135 vs 145 kPa) 0 0 0 0 Contact Stress (5 vs 10 kPa)1 0 0 0 0 Dwell2 0 0 0 0 Notes: 1 ITC only 2 IPC only Table 28. Significance of flow number ruggedness test factors on unconfined tests. Permanent Strain at Factors Flow Number 500 cycles 1000 cycles 2000 cycles Equilibrium Temperature (-1 vs +1 C) 0.45 0.32 0.26 Transfer time (3 vs 5 min) 0.39 .030 0.28 Conditioning Fluid (Water vs Air) 0.58 0.82 0.95 End Condition (Mill vs Saw) 0.01 0.06 0.23 Friction Reducer (Teflon vs Latex) 0.08 0.10 0.11 Axial Stress (945 vs 985 kPa) 0.79 0.56 0.46 Confining Stress (135 vs 145 kPa) 0.35 0.39 0.41 Table 25. Significance of flow number ruggedness test factors for confined tests with the IPC SPT on the dense mixture.

Data on the repeatability of the flow number test were collected in Phase II of this project (11). The Phase II ex- periment included eight replicates of two mixtures tested by single operators in two laboratories. Laboratory and mixture effects were found to not be significant, therefore, the 32 observations were pooled to obtain estimates of the repeatability of the flow number. The coefficient of varia- tion for the flow number from the Phase II analysis was found to be 35 percent while the coefficient of variation for the measured permanent strain was found to be only 14 per- cent. The high variability of the flow number was attributed to difficulties detecting the exact point where the perma- nent strain rate begins to increase. Future improvements may be made to the flow point detection algorithm, but it is unlikely that the repeatability of the flow number will be less than that for the measured permanent strain. Based on this analysis, the temperature control of ±0.5°C is accept- able. However, flexibility can not be permitted in the selec- tion of the end friction reducer. Since the greased latex membranes provide less friction and were recommended in Project 9-19, these friction reducers should be used in the flow number testing. 2.3.2 Factors Affecting Permanent Strain Figure 13 through Figure 17 show the effects of the signif- icant ruggedness factors on the permanent strain measured after 500, 1,000, 2,000, 5,000, and 8,000 cycles, respectively. The analysis for the permanent strain can only be performed when data are available for all specimens tested in both labs. The dense graded mixture specimens began to fail after 2,000 cycles. The SMA mixture specimens began to fail after 8,000 cycles. Considering the permanent strain measured in the flow number test has a coefficient of variation of 14 percent, several observations can be made based on the data shown in Fig- ure 13 through Figure 17. First, the machine control factors of temperature, axial stress, contact stress, dwell time, and confining pressure have little effect on the measured perma- nent strains over the control range provided by the SPT. Also 22 Permanent Strain at Factors 500 cycles 1000 cycles 2000 cycles 5000* cycles 8000* cycles Equilibrium Temperature (-1 vs +1 C) 25 25 25 0 0 Transfer time (3 vs 5 min) 0 25 25 0 0 Conditioning Fluid (Water vs Air) 0 0 0 0 0 End Condition (Mill vs Saw) 25 25 25 0 0 Friction Reducer (Teflon vs Latex) 100 100 50 50 50 Axial Stress (± 2 %) 25 25 25 0 0 Confining Stress (135 vs 145 kPa) 25 50 25 0 0 ∗ SMA only Table 29. Summary of significance of ruggedness test factors on confined tests. -30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30 Temperature, +0.5 C Time, min -Water / +Air - Milled / +Sawed -Teflon / +Latex Axial Stress, + 2% Contact Stress, +2% Dwell, +0.01 sec Confining Stress, +2% Change in Flow Number, % SMA Confined Dense Confined Dense Unconfined Figure 12. Effect of statistically significant ruggedness factors on the flow number.

the transfer time has little effect on the measured permanent strains over the range of 3 to 5 min. However, the three user- selectable factors, conditioning fluid, end condition, and end friction reducer, have a major effect on the measured perma- nent strains. In unconfined tests, the permanent strain was much higher when water was used as the conditioning fluid. Recall, the dense-graded mixture that was used had marginal resistance to moisture damage when tested in accordance with AASHTO T283. Apparently water that penetrates the voids in this mix- ture results in some level of moisture damage during the re- peated load test. The conditioning fluid was not significant in the confined tests, probably because less water entered the specimens because these specimens were conditioned with the confining membrane in-place. Although the ends were uncovered, the path for water infiltration from the ends is much longer resulting in less water absorption by the speci- men during conditioning. Clearly, water can not be used as a conditioning fluid in the flow number test. The measured permanent strains are higher when greased latex membranes are used as end friction reducers. Appar- ently this type of end friction reducer is more effective than Teflon™ resulting in less end friction and greater permanent deformation in the test. Flexibility can not be permitted in the selection of the end friction reducer. Since the greased latex membranes provide less friction and were specified in Project 9-19, these friction reducers should be used in the flow number testing. 23 -30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30 Temperature, +0.5 C Time, min -Water / +Air - Milled / +Sawed -Teflon / +Latex Axial Stress, + 2% Contact Stress, +2% Dwell, +0.01 sec Confining Stress, +2% Change in Permanent Strain, % SMA Confined Dense Confined Dense Unconfined -30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30 Temperature, +0.5 C Time, min -Water / +Air - Milled / +Sawed -Teflon / +Latex Axial Stress, + 2% Contact Stress, +2% Dwell, +0.01 sec Confining Stress, +2% Change in Permanent Strain, % SMA Confined Dense Confined Dense Unconfined Figure 13. Effect of statistically significant ruggedness factors on the permanent strain after 500 load cycles. Figure 14. Effect of statistically significant ruggedness factors on the permanent strain after 1,000 load cycles.

The specimen end condition also has a major effect on the measured permanent strains in the dense-graded mixture, but not the SMA mixture. Dense-graded specimens with milled ends had consistently higher permanent strains. Apparently, the smooth, milled ends of dense-graded mixture further reduced end friction resulting in an increase in permanent deformation. Because end milling is time consuming sawed ends meeting the specimen end condition requirements in the Equipment Specification for the Simple Performance Test System should be used. 2.3.3 Summary Table 30 summarizes the results of the analysis of the ruggedness test data for the flow number test. For statistically significant ruggedness factors, Table 30 presents the effect of the factor on the flow number and the measured perma- nent strains after 2,000 load cycles. Table 30 also presents acceptable values based on anticipated test variability. The following conclusions were drawn for each of the rugged- ness factors: 1. Equilibrium temperature. The current temperature control of ±0.5°C in the SPT is acceptable. Temperature changes over this level are expected to result in less than a 7 percent change in the flow number and less than a 5 percent change in the permanent strain. 2. Transfer time. The transfer time over the range of 3 to 5 min was found to be a significant factor only for the 24 -30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30 Temperature, +0.5 C Time, min -Water / +Air - Milled / +Sawed -Teflon / +Latex Axial Stress, + 2% Contact Stress, +2% Dwell, +0.01 sec Confining Stress, +2% Change in Permanent Strain, % SMA Confined Dense Confined Dense Unconfined -3 0 - 25 -2 0 - 15 -1 0 - 5 0 5 1 0 1 5 2 0 2 5 3 0 Temperature, +0.5 C Time, min -W ater / +Air - Mi lled / +Saw ed -Tef lon / +Latex Ax ial St ress, + 2% Cont act St ress, +2 % Dw ell, +0.01 sec Confining Stress, +2% Change in Permanent Strain, % SMA Confined Dens e Confine d De ns e Un conf ined Figure 15. Effect of statistically significant ruggedness factors on the permanent strain after 2,000 load cycles. Figure 16. Effect of statistically significant ruggedness factors on the permanent strain after 5,000 load cycles.

permanent strains in the confined tests. Increasing transfer time to 5 minutes is expected to result in no change to the flow number and less than a 4 percent change in the meas- ured permanent strain. Based on an acceptable range of 7 percent which is one-half of the coefficient of variation of the flow number test, the transfer time can be increased to 5 min. 3. Conditioning fluid. The use of water as a condition- ing fluid can result in moisture damage in the specimen during repeated loading if sufficient water penetrates the specimen. Air should, therefore, be used as the conditioning fluid. If the specimens are to be conditioned in a water bath, they should be sealed in plastic to keep the water from penetrating the specimen. 4. End condition. The method of preparing the specimen ends had a major effect on the permanent strain measured in both unconfined and confined tests. Milled ends resulted in larger permanent deformations for the dense-graded mixture probably because end friction was less with the smoother milled end. Because end milling is time consuming sawed ends meeting the specimen end condition requirements in the Equipment Specification for the Simple Performance Test System should be used. 5. Friction reducer. Of all the factors included in the ruggedness testing, the end friction reducer had the greatest effect on the flow number and the measured permanent de- formation. Flow numbers were much lower and permanent de- formation much higher when the greased latex friction reducer 25 -30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30 Temperature, +0.5 C Time, min -Water / +Air - Milled / +Sawed -Teflon / +Latex Axial Stress, + 2% Contact Stress, +2% Dwell, +0.01 sec Confining Stress, +2% Change in Permanent Strain, % SMA Confined Dense Confined Dense Unconfined Figure 17. Effect of statistically significant ruggedness factors on the permanent strain after 8,000 load cycles. Unconfined Confined Factors Control Flow Number εp, 2000 cycles Flow Number εp, 2000 cycles Equilibrium Temperature 0.5 °C < 7 % < 5 % NF < 2% Transfer time 3 versus 5 min NS NS NF < 4% Conditioning Fluid Air versus Water NS < 24 % NF NS End Condition Milled versus Sawed NS < 14 % NF < 15 % Friction Reducer Greased Latex versus Teflon < 20 % < 24 % NF < 25 % Axial Stress 2 % NS NS NF < 1% Contact Stress 2 % NS NS NA NA Dwell 0.01 sec NS NS NA NA Confinement 2 % N A NA NF < 3 % Acceptable 10 % 7 % 10 % 7 % NA = not included NF = no flow detected NS = not statistically significant Table 30. Summary of the effect of ruggedness test factors on the flow number and permanent strain.

was used. Flexibility can not be permitted in the selection of the end friction reducer. Since the greased latex membranes pro- vide less friction and were specified in Project 9-19, these fric- tion reducers should be used in the flow number testing. 6. Axial stress. The axial stress control of ±2 percent in the SPT is acceptable. Stress variations over this level are ex- pected to result in no change in the flow number and less than a 1 percent change in the permanent strain. 7. Contact stress. The contact stress control of ±2 per- cent in the SPT is acceptable. Stress variations over this level had no significant effect on the flow number or the measured permanent strains. 8. Dwell time. Data from the flow number test was not affected by a range in dwell time of 0.1 sec. The computer control used in the SPT is capable of controlling the dwell time much more precisely at this level. 9. Confinement. The current confining pressure control of ±2 percent is acceptable in confined tests. Over this range of control, the permanent strain is expected to vary by less than 3 percent. 26

Next: Chapter 3 - Results and Analysis of Equipment Effects Experiment »
Ruggedness Testing of the Dynamic Modulus and Flow Number Tests with the Simple Performance Tester Get This Book
×
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

TRB’s National Cooperative Highway Research Program (NCHRP) Report 629: Ruggedness Testing of the Dynamic Modulus and Flow Number Tests with the Simple Performance Tester examines ruggedness testing that was conducted with a simple performance tester for the dynamic modulus and flow number tests.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!