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 Bottom-up fatigue cracking is one of the main distress types in flexible pavement. Current design methods of flex- ible pavement assume that cumulative damage occurs where each load cycle uses up a portion of the finite fatigue life of the asphalt layer regardless of load magnitude or traffic vol- ume. The concept of endurance limit assumes that there is a strain value below which fatigue damage may not occur or can be healed during unloading. The fact that traffic loads are separated by ârest periodsâ may allow for partial or full heal- ing of the accumulated damage, which in turn increases the number of load repetitions before failure. Therefore, if the pavement is thick enough to keep strains below the endur- ance limit, the fatigue life of the pavement can be considerably extended. This concept has significant design and economic implications. In 1972, Monismith and McLean (1) first proposed an endurance limit of 70 microstrain for asphalt pavements. More recently, Nishizawa et al. (2) analyzed in-service pavements in Japan and reported an endurance limit of 200 microstrain. Wu et al. back-calculated Falling Weight Deflectometer (FWD) data and reported strains at the bottom of the asphalt layer between 96 and 158 microstrain for a long-life pavement in Kansas (3). Bhattacharjee et al. (4) obtained endurance limit values through uniaxial testing that ranged from 115 to 250 microstrain. Studies performed at the University of Illinois (5, 6) showed that fatigue life becomes significantly lon- ger if the strain is kept below approximately 100 microstrain. In NCHRP Project 9-38, beam fatigue and uniaxial tension testing were conducted to determine fatigue life (7). By con- ducting a small strain-controlled beam fatigue test, a fatigue life in excess of 50 million cycles was achieved. Data from the Long Term Pavement Performance (LTPP) studies were also analyzed to determine if they support the endurance limit concept. The results obtained from the study support the exis- tence of an endurance limit in HMA mixes (7). Another major concept recently investigated by research- ers is the HMA healing phenomenon. Healing of micro- damage was proposed as the primary reason for the increased fatigue life at low strain levels (8, 9, 10). Healing is generally considered as the capability of a material to self-recover its mechanical properties (stiffness or strength) to some extent upon resting due to the closure of micro-cracks. Phillips (11) proposed that healing of asphalt binders is a three-step pro- cess consisting of (1) the closure of micro-cracks due to wet- ting (adhesion of two crack surfaces together driven by surface energy); (2) the closure of macro-cracks due to consolidating stresses and binder flow; and (3) the complete recovery of mechanical properties due to diffusion of asphaltene struc- tures. Kim and Little (12) developed a mechanical approach to identify the healing potential of asphalt concrete. They performed cyclic loading tests with varying rest periods on notched beam specimens of sand asphalt mixtures. They con- cluded that rest periods enhance the fatigue life through heal- ing and relaxation mechanisms. There is mounting evidence that healing and the endurance limit of HMA are related to each other. It has been observed both in laboratory studies of fatigue at low strain levels with rest periods and in thick, properly constructed pavements that bottom-initiated fatigue cracking is almost non-existent. The HMA endurance limit, however, does not reflect an absence of load induced damage in the HMA. Rather, it results from a balance of damage caused by loading and healing or dam- age recovery occurring during rest periods (6). The endur- ance limit of HMA is, therefore, not a single value, but varies depending on loading and environmental conditions applied to the HMA. Considering an endurance limit in flexible pave- ment design requires the consideration of the effects of load- ing, environment, and material properties on both damage accumulation and healing. These findings on the endurance limit of HMA served as the research hypothesis upon which NCHRP Project 9-44 (13) was formulated. In summary, the literature provides endurance limit val- ues for certain conditions, but there is no general predic- tive model currently available to estimate these values under C H A P T E R 1 Introduction
2different conditions and accounting for healing. Also, the lit- erature does not provide a clear relationship between endur- ance limit and healing, which is one of the main contributions of this research. Previous studies showed that fatigue life is primarily influ- enced by mix stiffness (E) and also affected by binder content and air voids. Lower asphalt contents and lower air voids led to higher stiffness, while higher asphalt contents and lower air voids led to higher fatigue lives (14, 15). Tayebali et al. (16) also found that as air voids increased, fatigue life decreased for both controlled strain and controlled stress tests. It was concluded that stiffer mixes would perform better for thick pavements, while lower stiffness mixes would perform better for thin pavements. Fatigue life and endurance limit are also affected by tempera- ture and binder grade. Because the endurance limit of HMA is tied closely to the healing potential of the binder, healing occurs more rapidly at higher temperatures and softer binder grades, and the strain level that can be tolerated with no damage accu- mulation is increased (9). Verstraeten et al. (17) concluded that the longer the rest periods and the higher the temperatures, the greater the beneficial effect. Over the last 50 years, several researchers have studied the significance of rest periods between load applications dur- ing fatigue testing of HMA. Different findings have been pre- sented in the literature showing different opinions on the effect of rest period. Some researchers thought that the rest period only leads to a temporary modulus recovery without actually extending the fatigue life, while others found that the modulus recovery did extend fatigue life by a certain amount. Van Dijk and Visser (18) found that increased rest periods can increase fatigue life by a factor of 1 to 10 times. Other test results indicated that increasing the rest period had no sig- nificant effect on fatigue performance under certain assump- tions and test conditions (19). Raithby and Sterling (20) found that fatigue life does not increase significantly for rest periods greater than ten times the loading time (or 1s rest period) and the waveform influence was less important than the duration of rest periods. Van Dijk and Visser (18) showed increased fatigue lives with increasing rest periods. Bonnaure et al. (21) concluded that increasing the rest period between the loading cycles increases fatigue life. They also showed that the maxi- mum beneficial effect of rest periods on the fatigue life was for a rest period equal to 25 times the loading cycle (or 0.625s). Objectives The objectives of this research were to (1) determine the fatigue endurance limit for HMA and relate it to healing that occurs during the rest period between load applications and (2) develop models that relate the endurance limit of HMA to material properties, loading conditions, and temperature. These objectives were achieved by conducting both beam and uniaxial fatigue laboratory experiments to identify the mixture and pavement design features related to the endurance limit for bottom-initiated fatigue cracking. Issues studied included incorporating rest periods between loading cycles and the effect of rest period on the healing and endurance limit of HMA. This report summarizes the design, features, results, and products of the beam fatigue and uniaxial fatigue experiments conducted in NCHRP Project 9-44A to meet these objectives. Appendixes 1 and 2 are comprehensive treatises describing all aspects of the beam fatigue and uniaxial fatigue studies, respectively, including tabulations of the experimental data. These appendixes are adapted from dissertations presented by Dr. Mena Souliman and Dr. Waleed Zeiada to Arizona State University in partial fulfillment of the requirements for their Doctor of Philosophy degrees. Appendix 3 is a description of the Microsoft Access® Mechanistic-Empirical Distress Predic- tion Models (M-E_DPM) database in which all relevant data and results from the project are stored. Appendixes 1, 2, and 3, and the M-E_DPM database are not published herein, but are available online on the TRB website (http://trb.org) and can be found by searching for NCHRP Report 762. HMA Endurance Limit and Healing A rational procedure was developed to relate the HMA healing phenomenon to the endurance limit. If the fatigue test is conducted with and without rest period between load applications, the typical stiffness ratio (SR) (ratio between the current stiffness and the initial stiffness) versus the number of load cycles will be as shown in Figure 1. Both curves start at an SR of one (no damage). The curve for the test without rest period is steeper than the other curve because of the continu- ous deterioration during the test. The test with rest period shows higher SR values during the test because of healing that 0.5 Nf w/o RP 0.0 1.0 At Endurance Limit, SR = 1.0 at all values of N Test with rest period Test without rest period St iff ne ss R a tio Loading Cycles, N Healing Index, HI Figure 1. Typical SR versus number of load cycles for tests with and without rest period.
3 occurs during the rest period after each load application. A larger separation between the two curves indicates more heal- ing, and vice versa. If the curve for the test with rest period remains horizontal, it indicates that full healing occurs after each load cycle. Healing Index (HI) (Equation 1) was defined as the dif- ference between the SRs for the tests with and without rest period at Nf w/o RP (number of cycles to failure for the test with- out rest period) as shown in Figure 1, where failure is defined when the SR reaches 0.5. HI SR SR Nw RP w oRP at f w oRP= â[ ] ( )1 where, SR w/ RP = stiffness ratio with rest period SR w/o RP = stiffness ratio without rest period The approach used in this research is to develop a regres- sion relation between SR and various factors in the form of Equation 2: SR = f BG, AC, V , T, , N, RP (2)a( )ε In Equation 2, SR is the stiffness ratio, BG is the binder grade, AC is the binder content, Va is the air voids, T is the temperature, e is the initial strain, N is the number of load applications, and RP is the rest period between load applications. If the SR in Equation 2 is set to one, the strain, e, will become the endurance limit, which implies that full healing occurs after each loading cycle. Note that an SR of 1 is equivalent to an HI of 0.5, which means full healing. Note also that setting the SR to 1 to determine the endurance limit is a better approach than setting the HI to 0.5. The SR approach can be used at any number of load repetitions (N), but the HI approach can be used at Nf w/o RP only. Also, since it is assumed that full heal- ing occurs after each loading cycle at the endurance limit, the number of loading cycles is redundant and may be removed from Equation 2 without large effect. The rest period in the laboratory is inversely related to the average annual daily truck traffic (AADTT) in the field. If the AADTT is low, the time between trucks or axle loads is large, which corresponds to large rest periods between stress applications. Materials, Mix Design, and Fatigue Testing A 19-mm Superpave mix design was selected for the proj- ect that met the requirements of typical mixtures used for paving arterial roads in Arizona (22). Three asphalt concrete mixes using three binder grades (PG 58-28, PG 64-22, and PG 76-16) were prepared to ensure that a wide range of stiffness would be encountered. The same aggregate gra- dation was used for all three mixtures. Figure 2 shows the designed aggregate gradation distribution curve and the specification limits and Table 1 shows the composite aggre- gate properties. Mixes were designed according to the requirements of Maricopa Association of Governments (MAG) 710 specifi- cations (22) for High Traffic Gyratory mixes with Ndesign = 100. Table 2 shows the volumetric mix design for the three binders. The optimum binder contents were 4.8, 4.5, and 4.7 percents for the PG 58-28, PG 64-22, and PG 76-16 mixtures, respectively. Two types of fatigue tests were conducted, beam (flexure) fatigue and uniaxial fatigue. Beam fatigue tests were per- formed on HMA mixtures prepared with all three binder grades (PG 58-28, PG 64-22, and PG 76-16), whereas the uniaxial fatigue tests were performed on asphalt mixtures prepared with the PG 64-22 binder grade only. 0 10 20 30 40 50 60 70 80 90 100 Pe rc e n t P a ss in g Sieve Size (mm) Spec. Limits Agg. Gradation 0.45 Max. Density 19 . 0 12 . 5 9. 5 6. 3 4. 8 2. 4 2. 0 1. 2 0. 6 0. 1 0. 2 0. 3 0. 4 25 . 0 Figure 2. Designed aggregate gradation distribution. Property Value Specifications Bulk (Dry) Sp. Gravity 2.614 2.35-2.85 âSSD Sp. Gravity 2.638 Apparent Sp. Gravity 2.677 Water Absorption (%) 0.90 0-2.5 â Sand Equivalent Value 71 Min 50 Fractured Face One (%) 99 Min 85 Fractured Face Two (%) 96 Min 80 Flat & Elongation (%) 1.0 Max 10 Uncompacted Voids (%) 46.8 Min 45 L.A. Abrasion @ 500 Rev. 16 Max 40 Table 1. Composite aggregate properties.
4Volumetric Property Binder Type Specifications PG 58-28 PG 64-22 PG 76-16 Target Asphalt Content (%) 4.8 4.5 4.7 4.5-5.5 Bulk Specific Gravity (Gmb) 2.365 2.367 2.351 N/A Theoretical Max. Sp. Gr. (Gmm) 2.461 2.467 2.454 N/A Design Air Voids (%) 3.9 4.1 4.2 3.8-4.2 VMA (%) 13.9 13.5 14.3 Min. 13 VFA (%) 71.9 69.9 70.8 N/A Asphalt Sp. Gr. (Gb) 1.024 1.024 1.042 N/A Table 2. Volumetric mix design for different binder types.
