Investigating Effects Of Amine-Based Modifier On Recycled Asphalt Shingles Blending Index

This study was undertaken to investigate the effects of amine-based modifiers on the rheological characteristics of particle-filled viscous media such as recycled asphalt shingles (RAS). RAS are a recycled material that contains high concentrations of asphalt which has the potential for use in hot mix asphalt (HMA) when added to virgin asphalt. When using the RAS as a binder in HMA it is important to mix it with the virgin asphalt properly to achieve the best performance, which can also be enhanced by the incorporation of amine-based modifiers. Tear-off shingles were acquired from a roofing company and ground very fine so that 85% of the particles passed through sieve number 200. The virgin asphalt binder (PG 64-22) and three (20%, 30%, & 40%) percentages of grounded RAS were blended at a temperature of 180°C at a rotational speed of 400 rpm. These three mixtures were then blended with three different amine-based modifiers (1.5% of Rediset®, 0.5% of Evotherm®, and 5% of bio-binder by weight of mixture) at 135oC and a rotational speed of 400 rpm. The percentage of each modifier was selected based on recommendations of the manufacturers. The properties of the blended binder were studied using a rotational viscometer (RV) utilizing a Brookfield Viscometer DVIII-Ultra. Two different spindles were used to measure the viscosity of the binders at four different temperatures (105oC, 120oC, 135oC, and 150oC) and six (5, 10, 20, 25, 50, and 100) different rotational speeds. The analysis showed that viscosity increased with increasing percentages of RAS; however, the viscosities decreased after incorporation of the amine-based modifiers. Additionally, viscosity results were found to be different between the two spindles used. Viscosity measurement values were consistently higher when the vane spindle was used as compared to the smooth spindle. This can be attributed to incomplete blending of the RAS particles with asphalt matrix. However, the viscosity difference between the two spindles was reduced as the temperature was increased

This can be attributed to incomplete blending of the RAS particles with asphalt matrix. However, the viscosity difference between the two spindles was reduced as the temperature was increased and when modifiers were present. This, in turn, indicates an improvement of blending due to the addition of modifier and an increase in blending temperature. Furthermore, the coefficient of variation was significantly lower in cases where the vane spindle was used, indicating that the vane spindle could be more appropriate for measuring the viscosity of particle-filled viscous media.
An empirical relation was proposed to measure the blending behavior of the amine-based modified binders. The blending index was calculated using an empirical relation for all temperatures and rotational speeds. It was found that the blending index was affected by changes in temperature and shear speed. The blending index increased with increasing temperature. In addition, the bio-binder modified binder showed higher blending index compared to the other modified binders. Therefore, bio-binder is effective in reducing binder viscosity and enhancing blending between aged asphalt in RAS and un-aged asphalt  in the mixture.

Background
The United States has the largest road network system in the world comprising more than 4 million miles of pavement. Of this, 2.3 million miles are surfaced with hot-mix asphalt (FHWA, 2011); therefore, hot-mix asphalt is the predominant material in pavement construction, rehabilitation, and maintenance projects (National Asphalt Association, 2005). Due to increases in population and living standards there is a significant annual increase in traffic. It is important that the entire pavement surface have sufficient capacity to bear the rapid growth of traffic volume, high axle loads, and severe climatic conditions. To fulfill this increasing demand in the US, thousands of miles of road are constructed and thousands of miles of road are maintained and rehabilitated each year. All types of road construction work require sufficient amounts of materials such as aggregate, asphalt binder, fuel, coal, and so forth. Most of these construction materials occur naturally and can be found at various sources (e.g., mines, wells). Extraction of these materials leads to their gradual depletion. Additionally, extracting these materials from quarries and transporting them to job sites is difficult and costly, thereby increasing overall construction costs.
Accordingly, transportation agencies are increasingly interested in investigating new technologies that will reduce the cost of asphalt pavement materials while maximizing long-term performance. The American Society of Civil Engineers (ASCE) 2009 Infrastructure Report Card revealed that 186 billion dollars is needed annually for rehabilitation and maintenance of the United States roadway system, but only 70.6 billion dollars is being invested annually. The cost of asphalt materials can be reduced by replacing the virgin asphalt (binder) with recycled products obtained from construction waste or other byproducts that contain large amounts of asphalt. Adopting the use of recycled products not only reduces the cost of asphalt materials but also diverts construction waste away from landfills. Using recycled products to manufacture asphalt pavement also lowers the energy required to produce the pavement and minimizes the process's impact on the environment. Al-Qadi, Imad, Elseifi, and Carpenter (2007) reported that the performance of properly designed asphalt mixes containing recycled products exhibits no differences compared to asphalt mixes containing only virgin materials. When compared to conventional virgin mixes some recycled products even improve performance in certain applications.
Recycled asphalt pavement (RAP), recycled asphalt shingles (RAS), and Crumb rubber are the most common sources of secondary recycled materials used in road construction. RAP is old pavement that has been milled from the roadway, crushed into smaller aggregate sizes, and stockpiled. At the end of an asphalt pavement's "service life," the pavement is still valuable because it contains mineral aggregates and asphalt cement that can be reheated and reincorporated into new hot mix asphalt (HMA  (Krivit, 2007). The cost of disposing waste shingles in landfills can be as high as $90 to $100 per ton (Malik, Teto and Mogawer, 2000 (Janisch & Turgeon, 1996). In this study it was found that the use of up to 5% RAS was beneficial in that it caused a slight increase in asphalt cement hardness. Janisch and Turgeon's (1996) findings, increases in energy prices, and the gradual depletion of natural resources all served to stress the compulsory need to adopt new methodologies that would benefit the environment, users, and the industry. Although recycling by-product materials is beneficial in most cases because of the reduced consumption of virgin materials, it is imperative that the performance of the highway is not compromised.
Benefits that may derive from the recycling of by-product materials in HMA include: 1. reduced consumption of virgin materials, 2. reduced emissions and energy consumption during processing and manufacturing as a result of using fewer virgin materials, 3. reduced amount of by-product material disposed in landfills, 4. diminished consternation of the public regarding emissions, and 5. improved economic competitiveness in the asphalt paving construction industry.
Clearly, recycling asphalt shingles in HMA could be a valuable approach in the road construction industry.

Problem Statement
Disposing of roofing shingles at the end of their service lives results in the accumulation of large quantities of old asphalt materials in landfills. If these materials are not properly treated, serious environmental problems can result. The fabrication of additional HMA from virgin aggregate and bituminous materials only compounds the problem. RAS have been considered a valuable construction material because they can be included in new hot mix asphalt for use in both the construction and maintenance of paved surfaces. According to the FHWA, the use of RAS in the production of new asphalt materials results in economic, environmental, and energy savings (Roof to rood).
Recently, RAS have been increasingly used as a coarse material (dry process) and as a part of the fabrication of binder material (wet process). The latter method is more commonly used in the pavement industry. In this method RAS is blended with virgin binder using a shearing mechanism. Blending of RAS with virgin asphalt decreases the quantity of virgin material required and thereby lowers construction costs. The new challenge is to use higher percentages of RAS in pavement construction without reducing performance. Elseifi, Salari, Mohammad, Hassan, Daly and Dessouky, 2012 demonstrated a method called the "wet process" in which they increased the percentage of RAS without depleting the required performance criteria. In this wet process, RAS were ground into a very fine form called "ground RAS." The ground RAS was then blended with virgin binder material at high temperatures and high shear rates prior to mixing with the aggregate so that the RAS mixture could act as a binder. Using the wet process facilitated better control of the chemical and physical reactions taking place in the binder blend. . Elseifi, et al., 2012 reported that using this method up to 20% RAS can be used in road construction materials without compromising the performance of the road. Although Elseifi, et al.'s (2012) wet process was based on manufactured recycled shingles. A search of the extant literature revealed that no research has been conducted focusing only on tear-off roofing shingles and their blending behavior with virgin asphalt; therefore, this research study focused on the blending behavior, performance, and characteristics of RAS with virgin asphalt in the presence of three different modifiers as reflected in the resultant blending index.

Research Objectives
The main disadvantage of using high percentages of RAS in construction is the increased stiffness of the mixture. This, in turn, can make the mixture hard to mix and compact. To address this problem, certain amine-based modifiers are added to the mixture to reduce the stiffness and enhance workability. To evaluate the effects of various modifiers, viscosity can be measured to determine the rheological behavior of the modified mixtures. The objective of this research was to evaluate the physical and rheological properties and performance of mixtures modified with RAS with and without the presence of amine-based modifiers. In addition, this work evaluated the effectiveness of the designed modifiers in RAS-modified mixtures in terms of improving the blending between aged and un-aged asphalt. In this study the viscosity of samples was measured using two different types of spindles. The measured viscosity data obtained from this research has the potential to introduce new concepts that might be used to determine the blending index of mixtures, thereby providing a new approach for further research and development of new modifiers for flexible pavements.

Experimental Plan
This study was designed to investigate the effect of amine based modifiers on RAS binder with or without incorporation of modifiers. Each of these 12 specimens was processed using both spindle type SC4-27 and spindle type V73.
In this research the proportion of Redised® was 1.5%, Evotherm® was 0.5%, and biobinder was 5% by weight total mixture. To make a homogeneous mixture, each combination of RAS and virgin asphalt were first blended for 60 minutes at 400 rpm and 180 o C; similarly, each modifier was blended separately with RAS modified mixture (already prepared) at 400 rpm and 130 o C for 20 minutes. From each blended mixture 10.5 grams of binder was poured into an aluminum chamber. The chamber was then placed into a preheated Thermosel® for 20 minutes.
To measure viscosity a Brookfield Engineering viscometer was chosen and tested. A dynamic shear rheometer (DSR) test was conducted to measure the viscoelastic properties of the binders.
An experiment set up used in this study is shown in Figure 1.1; furthermore, in detail each mixture design and modification will described in chapter three in section 3.3 and 3.4.

CHAPTER 2 Literature Review
This chapter is designed to introduce RAS as a construction material for use in asphalt pavement. Some previous applications of RAS in HMA used in the pavement industry are discussed. A thorough literature review regarding the laboratory tests which are performed in this research are also included.

Past Studies of Recycled Asphalt Materials
Environmental measures are becoming more dominant factors in the decision-making process in infrastructure and construction projects. Additionally, global crude oil prices have increased rapidly in recent decades. The price of liquid asphalt has grown dramatically; the price of asphalt increased from $235/ton in 2004 to more than $635/ton in 2013 (New York Department of Transportation, 2013). As a product derived from petroleum distillation, asphalt is becoming less available because of improvements in cooking technologies that allow refineries to produce synthetic fuel from asphalt. This, in turn, reduces the supply of asphalt available for road construction (Cleveland, 1993). An increasing concern for sustainable development, in addition to the emphasis on material conservation, reuse, and recycling, has encouraged a number of government and highway agencies to commission research investigations to characterize, and optimize the production of pavement materials. The use of recycled materials can provide additional value. They have been used in applications that show performance similar to conventional materials and cost effectiveness has been demonstrated (Iswandaru & Wilson, 2006). These successes have driven researchers and pavement industry companies to address the issue of using more and more recycling materials derived from waste products. For example, biobinder, RAS, and RAP can be used as alternative asphalt resources while looking for substitutes for virgin asphalt (Fini, Kalberer et al., 2011). Similarly, with regard to RAS, in the US more than 11 million tons of asphalt roofing materials are produced each year. Ten million tons are post-consumer (tear-off), and one million tons are pre-consumer manufacturing scrap. Asphalt roofing shingles have been used in paving practices since the early 1990s as a portion of aggregate, and more recently have also been used as a binder in hot mix asphalt. Due to the presence of large quantities of asphalt in RAS, most state agencies that regulate road construction have allowed RAS to be used with certain maximum percentages in hot mix asphalt.
The maximum allowable percentage of RAS in most states is approximately 5% by weight of the total aggregate. Some states limit RAS type to manufacturing scrap only, while others allow for the application of tear-offs as well. For example, following the supplemental specification issued by the Ohio Department of Transportation in 2011, the state of Ohio allowed the use of either manufacturer's RAS or tear-off RAS depending on the particular pavement course (Ohio Department of Transportation, 2011).  Several authors of previous studies mentioned that the introduction of shingles into an asphalt mixture can increase the stiffness of the mixtures which can, in turn, promote pavement resistance to rutting. Ali et al., 1995 andBurak et al., 2004 studied the effects on engineering properties that resulted from introducing roofing shingles into HMA. They found that the Marshall Stability values increased when adding 1% shingle into the mixture but that further increasing the percentage of shingle caused a decrease in stability values. However, they noted that at concentrations of up to 5% shingle the stability values of the mixtures were still higher than the minimum values called out in the superpave specification criteria. In addition, this and other studies showed that by incorporating 5% shingle in pavement construction contractors can reduce the construction cost by $2.79/ton (Brock et al., 1989). In another study, Foo et al. (1999) reported that the introduction of shingles into virgin asphalt can cause a significant increase in the stiffness of the asphalt binder. The use of shingles in a HMA mixture will generally improve the rutting resistance; however, the mixture may show lower fatigue life and lower thermal cracking resistance. In such cases it was recommended that the use of an appropriate softener (bio-binder) in virgin binder could improve the fatigue and low temperature performance of the mixture (Fini, Al-Qadi et al., 2011). There are several studies and innovations related to facilitating the application of RAS without compromising the "workability" and mechanical properties of the mixture (e.g., application of softer binder, mechanical grinding, and wet processing). Recently, a bio-based additive was produced that is able to make the binder softer, thereby enhancing workability and mixing (Mogawer et al. 2012, Fini, Al-Qadi et al., 2011, Beale, 2011Williams, 2013).
Another innovation was grinding the RAS to ultra-fine particle size and blending it with asphalt binder through a wet process. This facilitated the incorporation of higher percentages of RAS in hot mix asphalt (Elseifi et al., 2012). In this wet process, the ground RAS is blended with the binder at a high temperature prior to mixing it with the aggregates. This method permits better control of the chemical and physical reactions which occur in the binder blend. Results of the rheological and stability testing for this wet process indicated that 20% RAS can be used successfully in HMA. Fini et al. (2011) studied the effects of bio-binders on mixtures containing RAS. A biobinder derived from swine manure was added to the base PG 52-28 asphalt binder at a concentration of 5% by weight of asphalt binder and a bio-modified binder was created. Due to the chemical and physical nature of the bio-binder, its introduction along with RAS allowed mixing at a lower temperature of 124ºC and compaction occurred at 113ºC. This study showed that the presence of bio-binder led to improved blending between the aged asphalt and the virgin asphalt. In addition, it was found that bio-binder improved the workability and compaction of the RAS content mixture. In another study by Fini, Al-Qadi et al. (2011) on the analysis of dynamic modulus of mixtures it was shown that incorporation of 40% RAP to the control mixture increased the mixture stiffness. The introduction of the bio-modified binder decreased the 40% RAP mixture's stiffness; therefore, it indicated that the bio-modified binder can effectively reduce the mixing and compaction temperatures and help to reduce the stiffness effect caused by the introduction of high percentages of RAP and RAS in the mixtures.
Based on previous research studies it became clear that aged RAS are one of the constituents that help increase the viscosity of the mixture but may cause stress on the pavement during preparation, mixing, compaction, and during its life of operation. To address this issue different types of modifiers and additives are being used in construction according to their properties and design guidelines/specifications. In an attempt to establish a suitable design method, a new method called warm mix asphalt (WMA) was introduced recently. Marisa et al.
(2012) conducted a Marshall Stability Test, immersion compression test, and water sensitivity test on a warm recycled mix. They concluded that the temperatures for the production and compaction of the mixtures influenced the final results. The best result was obtained from mixtures compacted at 90 o C.

Recycled Asphalt Shingles (RAS)
Shingles are manufactured for 15-20 years of service. After their life service time they are replaced by new roofing shingles which produces a large quantity of waste/scrap shingles.
Reuse of recycled asphalt shingles was identified by the U.S. Environmental Protection Agency (EPA) as a top priority. Constituents of typical asphalt shingle include 20-35% asphalt cement, 2-15% cellulose felt, 20-38% mineral granule/aggregates, and 8-40% mineral filler/stabilizer. Due to the high content of asphalt in shingles, the primary application of RAS is production of hot mix asphalt. Most states' departments of transportation (DOT) approved 5% (depends upon the type) RAS in HMA. Research by Button et al. (1995) and Grodinsky (2002) revealed that the use of more than 5% by weight RAS in HMA affected adversely the creep stiffness and tensile strength of HMA. Consequently, this 5 % RAS application uses only 10-20% of the total asphalt shingle waste generated (Turley, 2010). To make use of the additional waste another potential application of RAS could be incorporation into structural fill including highway embankment fills or backfill behind retaining walls.
Asphalt shingles contain approximately 30% AC by mass (Foo, 1999); therefore, using RAS in HMA decreases the amount of virgin AC required, and decreases the costs to produce HMA. It can also enhance the properties of the HMA when small amounts of RAS are incorporated; however, this improvement may be dependent upon the source and quality of the RAS. The roofing application of shingles and the demolition of the roofing shingles are shown in   The granular material in asphalt shingles is composed of coal slag and crushed rock coated with ceramic metal oxides. It is generally uniform in size, ranging from 0.3mm to 2.36 mm, and is hard and angular when powdered limestone (70% passing the No. 200 sieve) is also added as a stabilizer (Newcomb, 1993;Ross, 1997) which makes the mixture stiffer.

Types of Recycled Asphalt Shingles
Understanding the composition and properties of asphalt shingles is necessary to fully characterize asphalt mixtures in which they are incorporated. The American Society for Testing and Materials (ASTM) clearly specifies shingles according to their production in documents ASTM D225 and ASTM D3462. The specifications in ASTM D225 apply to asphalt shingles made with organic (cellulose or wood fiber) backing, and ASTM D3462 contains specifications for asphalt shingles made with fiberglass backing.

Organic shingles.
Organic shingles are made of paper (felt)-saturated asphalt cement (AC). These types of shingles are heavier and contain more AC. In cold regions, such as the northern USA and Canada, these shingles are used due to the higher flexibility conferred by the large AC content. The increased flexibility makes them less likely to crack in cold weather.

Fiberglass shingles. Fiberglass shingles contain a base layer (mat) of fiberglass
coating. These types of shingles are easier to work with and install because the fiberglass base makes the shingles lighter in weight. Fiberglass shingles provide greater resistance to moisture and fire than organic shingles.

Typical Asphalt Shingle Composition
The percentages of the individual component materials in asphalt are different in shingles manufactured with organic felt compared to shingles manufactured with fiberglass felt. Brock  Shingles are manufactured by saturating and coating both sides of organic or fiberglass backing felt with liquid asphalt. The asphalt used to coat the felt material is different from asphalt used in paving materials. The asphalt used in roofing shingles is much harder and stiffer because the manufacturers use an "air-blown" process to increase the viscosity of the asphalt.

Asphalt Cement Content in Tear-off Shingles
Weathering a portion of the surface granules on roofing shingles results in a greater overall percentage of AC compared to new shingles. Oxidation and volatilization of the lighter organic compounds in roofing shingles makes the AC in tear-off shingles stiffer. As a result, using higher percentages of RAS in HMA can lead to the mix being stiffer than a virgin mix.
Tear-off shingles tend to include nails, paper, wood, and other debris that makes recycling a longer process (Mallick, 2000). Care and consideration should be taken when RAS is added to HMA to avoid this potential contamination.

Benefits of Recycled Asphalt Shingles in Hot Mix Asphalt
The benefits of using shingles in HMA include cost savings, environmental preservation, and the potential for improved performance. Recycling RAS in HMA avoids the expense associated with the disposal of shingle waste and reduces the amount of material entering landfill sites, thereby benefitting the environment. The amount of virgin AC required in HMA mixes can be reduced by incorporating RAS; this reduces costs. A relatively small number of shingles can displace a large percentage of AC (Foo 1999) in hot HMA. Additionally, studies revealed increased resistance to high-temperature rutting in HMA that contained factory waste shingles (Foo, 1999). The benefits of using RAS include:

CHAPTER 3 Materials Used and Experiment Methodology
In this chapter each of the materials that were used in this study are characterized. Asphalt

Evotherm®.
Evotherm® is a warm mix additive/modifier used successfully in warm mix technology in asphalt pavement construction. Evotherm® WMA is a comprehensive chemical additive system designed to allow the production and compaction of high quality asphalt pavements at temperatures much lower than those needed in conventional HMA. The benefit is the reduced consumption of energy when manufacturing the asphalt mixes. Various job sites studied by Michel, Frederic and Faucon (2003) achieved energy savings of approximately 40% percent, with measured gains ranging from 35% to 55% depending on the moisture content of the aggregate materials and the ambient weather conditions. Additionally, the reduction in processing temperatures caused a significant drop in the emission rates of stack gases and particulates at the mix plant. One study showed a 48% reduction in greenhouse gases, 58% reduction in nitrogen oxides, and 41% reduction in sulfur dioxide, which is responsible for acid rain (Michel et al., 2003).

Virgin asphalt binder.
Un-modified binder which was classified as PG 64-22 according to Superpave specifications was selected for this study. This bitumen is a petroleumbased refined product. Typical heating temperature of the bitumen is 177˚C with a flash point of 325˚C. Preferred storage temperatures range between 140˚C and 168˚C. The use of unnecessarily high temperatures results in increased hardening, oxidation, and heating costs. PG 64-22 is primarily used in paving for both new construction and pavement rehabilitation. (U.S. Oil & Refining Corporation, 2005). It was used in an attempt to offset the potential mixture stiffening resulting from the use of a high percentage of RAS in the mixture. Based on the viscosity of the binder, the mixing temperature was 180ºC. Table-3.1 shows properties of the virgin binder.

Bio-binder.
Bio-binder is derived from non-petroleum-based renewable resources such as wood or corn. Recent research efforts have suggested that using a bio-binder along with a petroleum-based asphalt can produce a bio-modified binder (Fini, Al-Qadi, Zada B. and Beale, 2011 and; Williams, 2013); therefore, the bio-binder could be an alternative to petroleum-based asphalts. In this study bio-binder used was produced by thermochemical liquefaction processing of swine manure under relatively high temperature (T = 340 o C) and pressure (P = 10.3 MPa) for specific residence times (RT = 80 min.) is used to produce bio oil and utilizes the heavy residue remaining in this process as an asphalt modifier. Table 3.2 shows the chemical composition of bio-binder and asphalt. optimum binder content to conform to the mix design criteria; its stability is lower than mixture fabrication at high temperatures. The use of a lower temperature leads to less energy consumption and lower emissions production at asphalt mixing plants. Hamzaha, Golchina and Ching (2013) study showed that the optimum binder content (OBC) of warm mix asphalt (WMA) was slightly lower than the OBC for HMA. Furthermore, the higher Rediset® content slightly decreased the stability of the asphalt mixture. This implied that higher Rediset® content has a softening role in the asphalt mixtures (Hamzaha et al., 2013). Table 3.3 shows the recommended concentration of Rediset® for various applications.

Preparation of RAS
The tear-off shingles used in this study were obtained from a local roofing company in Greensboro, North Carolina. Dirty particles like iron nails, wood, paper, pieces of glass, and other debris were separated from the shingles. The separated shingles were then ground utilizing an industrial Hamilton Beach grinder. Grounded RAS was then gradated to isolate the required particle size samples using sieve analysis. The grounded RAS were put on the top of the sieve and shaken for 20 minutes using an automatic shaking mechanism. A typical Hamilton Beach grinder is shown in Figure 3.1. The various sieve sizes are shown in Table 3.4. 2. 5% of bio-binder, 1.5% of Rediset®, and 0.5% of Evotherm® by weight of total mixture were blended separately.
The details of the mixing proportions and titles given to each modified binder are shown in Table 3.5. In the title of each modified binder, first two letters (MB) stands for modified binder; followed by two digits which stand for percentages of recycled asphalt shingles (RAS) and a single letter (E, R, B) which stands for the type of amine-based modifiers: "E" for Evotherm®, "R"for Rediset®, "B" for Bio-binder from swine manure and "N" for no modifiers.

Mixing (Blending) Process
To perform the mixing, 12 aluminum cans were taken and filled by preheated asphalt binder of 150 grams. Among those 12 cans, four cans were blended with 20% grinded recycled asphalt shingles, four cans were blended with 30% of RAS, and the last four cans were blended with 40% of RAS at high temperature and shear speed. Single cans representing each percentage (a total of three cans) were kept separate and used as control samples. Three sets of cans were made from the rest of the cans (a total of nine cans) with each of the three groups containing 20%, 30%, and 40% RAS. The first set of cans was blended with 0.5 % Evotherm®. Similarly, the second set of cans was blended with 1.5% Rediset®, and the third set of cans was blended with 5% bio-binder at high temperature and rotational speed. Each blended mixture was then poured into a small aluminum chamber. Altogether there were 12 specimens for spindle SC27 and an equal number of specimens for the V73 spindle. Modification details are presented in the next section.

Experiment Method
3.5.1 Viscosity measurements. The viscosity of the prepared specimens was measured at different temperatures and shear rates using a Brookfield viscometer (RV-DVIII Ultra) followed by the ASTM D4402 test procedure. To prepare the test specimens; after blending, two specimens were prepared from each modification by pouring 10.5 grams of blended binder into the tiny aluminum chambers shown in Figure 3.2. Altogether twenty-four specimens were prepared for viscosity measurements: one set (twelve) of specimens was for smooth spindle (SC 27) and the second set (twelve specimens) was for vane spindle (V73) and these tubes (aluminum chambers) were then placed in the preheated Thermosel® for 30 minutes to reach thermal equilibrium. The samples were then preheated by putting them into the Thermosel® set to the designated temperatures for an additional 20 minutes to ensure the achievement of thermal equilibrium. The test was run and the results recorded three times at 1-minute intervals to ensure the viscosity measurements were consistent. In this study two spindles (SC4-27 and Vane Spindle V73) were used to measure the mixture viscosity.   susceptibility of the RAS-modified asphalt blends was evaluated by developing temperatureviscosity plots for the specimens prepared. If an asphalt binder has a high susceptibility to temperature, its viscosity changes rapidly as the temperature changes. Asphalts with high temperature susceptibility are undesirable as they are more prone to undergo thermal and UV oxidation (Firoozifar & Foroutan, 2011). Therefore, it is important to quantify numerically the temperature susceptibility of the binders. The following equation has been commonly used to calculate temperature susceptibility (VTS; Rasmussen, Lytton, & Chang, 2002). The magnitude of the VTS is directly proportional to the temperature susceptibility of the asphalt binder.

Shear susceptibility.
Shear susceptibility is defined as the rate of change in viscosity with the shear rate (Roberts et al., 1996). The shear susceptibility, also known as the shear index, is determined by calculating the slope of the line formed by a log of rotational speed versus the log viscosity graph using Equation 3 Prior studies showed that binder with relatively small shear susceptibility (low gains in shear susceptibility relative to the increase in viscosity) result in better overall pavement performance (Roberts et al., 1996).

Blending index.
Blending index is an indication of the degree of blending achieved between the oxidized binder in RAS and virgin binder. The blending index of the RAS-modified binder was evaluated using viscosity variation versus temperature. Using the difference between the two measurements at the same temperature and speed rate, a blending index was defined as follows: where T is the temperature of the binder at known points expressed in degrees Celsius ( o C), and ηSC27 and ηV73 are the viscosities of the binder at known points (cP).

Dynamic Shear Rheometer (DSR) Test
The dynamic shear rheometer (DSR) is an instrument used to characterize the viscous and elastic behaviors of asphalt binders at medium to high temperatures. This characterization is used in the superpave PG asphalt binder specification. Due to the viscoelastic nature of asphalt it behaves partly like an elastic solid (deformation due to loading is recoverable; is able to return to its original shape after load is removed) and partly like a viscous liquid (deformation due to loading is non-recoverable; it cannot return to its original shape after the load is removed). DSR measures an asphalt's complex shear modulus (G*) and phase angle (δ).
The complex shear modulus is the ratio of total shear stress (ηmax-ηmix) to the total shear strain (γmax-γmin) and is considered to be the asphalt's total resistance to deformation when repeatedly sheared. The phase angle is a measure of the response time between the applied shear stress and the resulting shear strain. If asphalt was purely elastic, the phase angle would be zero degrees. If asphalt was purely viscous, the phase angle would be 90 degrees. Figure 3.6 illustrates the relationship between the phase angle and time factor.  For asphalt binder to have rutting resistance, it must have high stiffness and elastic properties at high temperatures. Elasticity is defined as the property of being able to recover its original shape after being deformed by a load. The higher the G* value, the stiffer the asphalt binder is. Similarly, the lower the δ value is, the greater the elastic portion of G* is. Therefore, as part of the PG binder specification system, the parameter G*/Sin (δ) is specified to be a minimum value (1.0 kPa for un-aged binders and 2.2 kPa for RTFO-aged binders).

CHAPTER 4 Results and Discussion
In this chapter data from the experiments conducted in the large-scale viscosity tests are analyzed. First, the viscosity tests and results using spindle SC 27 will be explained. The viscosity tests and results measured using spindle V73 will then be explained and a comparison made between the rheological properties revealed with the SC27 & V73 spindles. Using the empirical relationships of the blending index, all modified binders blending indices are discussed.

Rheological Characterization of Binders Utilizing Spindle SC27
These experiments were designed to characterize the rheological properties of RASmodified binders with or without the incorporation of amine-based modifier. viscosity of RAS-modified binder was found to be higher than the virgin binder or unmodified binder. When the temperature was increased the viscosity was decreased in all cases (either RAS-modified or non-modified). As seen in Figure 4.1, at a temperature of 105 o C the viscosity of RAS-modified binders was higher than the control sample, and at a temperature of 150 o C, the viscosity of the RAS-modified binders was still higher than the control but the viscosity value was less than the viscosity measured at a lower temperature. These results suggest that viscosity decreased when mixing temperature was increased.   Rediset® and Evotherm® are commercial modifiers and the doses used were those specified by the manufacturers. In contrast, bio-binder is a modifier produced in the lab by the author by thermo chemical liquefaction of swine manure, and doses used were those specified in past research. The rheological properties of the binders prepared by incorporation of the modifiers were characterized. In Figure 4.3, all modified binders are shown to have lower viscosities than the non-modified binders. At lower temperatures each binder had a higher viscosity value than that found at the higher temperatures. At each temperature bio-modified binder had a lower viscosity than Rediset®-modified and Evotherm®-modified. Therefore, it can be said that biobinder can effectively reduce the binder viscosity. This trend was consistent for all other speeds tested.

Rheological Characterization of Binders Utilizing Spindle V73
In this study the entire experiment was repeated for spindle V73 following the same procedures and using the same machine (Brookfield viscometer). The only difference was the spindle used.

30% RAS-modified binder with and without modifiers. Increases in the RAS
percentage caused increased viscosity in binders. Thirty percent RAS content binders had higher viscosity values than those measured below 30%. In Figures 4.11 and 4.12 it can be seen that higher viscosities were found compared to the values shown in Figures 4.9 and 4.10.
As shown in Figure 4.11 the non-modified (MB-30-N) binder has a higher viscosity than the modified binders. In addition of modifiers into mixture helps to decrease the viscosity of the mixture. In all cases the Bio-binder reduced viscosity more than the mixtures of Rediset® and Evotherm® seen in  The same trend is shown in Figure 4.12, but the values are lower than the values seen in   In Figure 4.15, it is seen that the viscosity measured from V73    Here it can be seen that the difference between the two spindle values measured decreased as the temperature was increased in steps from 105 o C to 150 o C.
The graph shown in Figure 4.17 is the viscosity measured using two different spindles for the same specimen modified by the amine-based modifier Rediset®. The results show that at all designed temperatures the vane spindle measured higher values than the smooth spindle.  Therefore, the vane spindle is more effective when evaluating the rheological properties of the binders because of it measured more significant value than smooth spindle.

Viscosity Temperature Susceptibility (VTS)
The viscosity temperature susceptibility of all modified and non-modified binders was evaluated separately from the data gathered using the two different types of spindles at 20 rpm and calculated values are summarized in    The VTSs were evaluated for all amine-based modifiers separately. All modified binders showed a similar trend in that the Bio-modified binder showed lower VTS values than the Rediset®-and Evotherm®-modified binders. In Figure 4.20, the VTSs of Bio-binder modified with three percentages of RAS is shown. The Bio-binder modified with 20% RAS was less susceptible to temperature than the mixes containing 30% and 40% RAS. Furthermore, the VTSs were lower in data obtained with the smooth spindle compared to the vane spindle measured at all temperatures. Therefore, it can be concluded that low concentrations of RAS in the mixture were less susceptible to temperature for cases with and without incorporation of amine-based modifiers.    In this study, Evotherm® made the mixture less temperature susceptible than the nonmodfied samples, Rediset® made the mixture less temperature susceptible than Evotherm®, and, finally, Bio-binder made the mixture less temperature susceptible than other binders. Therefore, it can be said that Bio-binder seemed to be a good modifier in the sense of viscosity temperature susceptibility for all percentages of recycled asphalt shingles content binders.

Shear Susceptibility
Shear susceptibility of all the binders was evaluated by utilizing the smooth spindle measured data. The shear susceptibility was plotted as log (shear rate) versus log (viscosity) at The Shear Susceptibility of the modified and non-modified binders obtained using spindle An analysis of Figure 4.24 reveals that the shear susceptibility for 40% RAS-modified binder (MB-40-N) was more consistent than other percentages of RAS-modified binders.
However, the values are in decreasing order as the shear rate increased, and none of the binders showed momentary fluctuations in value which suggests stability. A gradual decrease in shear susceptibility can occur at all other temperatures.       The increment in Bx from 105 o C to 150 o C for Rediset®-modified binder is 1.18%, in Evotherm®-modified binder it is 2.0%, and for Bio-binder modified binder it is 2.3%. The change in the Bx values in the case of Bio-binder is higher than the other two cases; therefore, comparatively, bBio-binder showed better results in this study.
A similar trend was seen in the 40% RAS content binder (see Figure 4.30).

Dynamic Shear Rheometer (DSR) Test
A dynamic shear rheometer (DSR) was utilized to measure the binders' viscoelastic properties. Twenty-percent recycled asphalt shingles, containing mixtures with or without modifiers, were used to measure viscoelastic properties. The results are plotted in Figures 4.32 through 4.34. In figure 4.32, the plot shows the changes in the complex modulus (G*) values at different frequencies. The complex modulus was increased with increases in the reduced frequency or decreases in temperature. It can be seen that the G* of all modified binders is greater than the control binder. This indicates that the RAS makes binder stiffer, as higher G* values indicate more stiffness and lower G* values indicate less stiffness.
Furthermore, incorporation of amine-based modifiers improved the softness of the binder.
As seen in Figure 4.32, at all reduced frequencies RAS-modified binders        As shown in Figure 4.34, all modified binders have lower phase angle values than the non-modified binders. This indicates that the amine-based modifiers made the binders more elastic. However, among these, the Bio-binder modified binder has the lowest phase angle value.

Complex Modulus(G*) Pa
Complex modulus of the binders at temperature 64 o C and frequency at 1.67E+00

Phase angle(δ)
Phase angle of the binders at temperature 64 o C and frequency at 1.67E+00

Summary
This research was undertaken to evaluate the rheological characteristics of mixtures prepared by the addition of various percentages of recycled asphalt shingles into virgin asphalt with or without the incorporation of amine-based modifiers. The effectiveness of the spindles used to measure the viscosity of the modified mixtures was also investigated. Three (20%, 30%, and 40%) RAS-filled viscous media were prepared with or without incorporation of amine-based modifiers. A Brookfield viscometer was utilized to measure the viscosity of these binders using two different spindles.
In total, 24 specimens were made and the viscosity of each specimen was measured at four different temperatures and six different shearing rates. In this study a dynamic shear rheometer (DSR) test was also conducted on four specimens which were prepared with or without the incorporation of three amine-based modifiers at 20% RAS-filled media. All tests were conducted at the Civil Engineering Lab at North Carolina A & T State University.
The RV test was used to measure viscosity, which is the rate of deformation due to an applied shear or tensile stress. For each sample, an RV test was run three times to ensure accuracy with a fixed temperature and a fixed shear rate. These three readings were then tabulated, and the mean and coefficient of variation were calculated. The temperature was kept constant for five different shear rates, each of which was measured three times. This test was used to determine the rheological properties (temperature susceptibility, shear susceptibility, and blending index) of the samples.
The DSR test was used to measure viscoelastic properties, shear modulus (G * ), and the phase angle (δ) of the mixtures prepared with and without amine-based modifiers. A small sample with an 8 mm diameter was prepared from each binder and placed ("sandwiched") between the two plates of the rheometer. The test specimens were kept at near constant temperatures by heating and cooling a surrounding environmental chamber. The top plate oscillated at 10 rad/sec (1.59 Hz) in a sinusoidal wave form while the equipment measured the maximum applied stress, the resulting maximum strain, and the time lag between them. The software then automatically calculated the complex modulus (G*) and phase angle (δ). Much of the procedure is automated by the test software.
An empirical relationship between viscosity and temperature was proposed to measure the blending index of the mixtures based on the measured viscosity using two different spindles at the same temperature. The results of this portion of the research study are listed below.

Observation and Conclusions
The purpose of this study was to evaluate the effect of specific amine-based modifiers in partially filled viscous media on the basis of changes in rheological properties. Based on the test results for amine-based modifiers modified asphalt, the following conclusions can be made:  Viscosity increased with the addition of recycled asphalt shingles into virgin asphalt binder (PG 6.4-22) and the increasing viscosity correlated with increases in the percentage of recycled asphalt shingles added.
 The viscosity of the binder was decreased with increases in the mixing temperature and increases in the shear rate. Furthermore, the viscosities were decreased with incorporation of the amine-based modifiers into RAS-filled mixtures.  In all modified and non-modified binders, use of the vane spindle (V73) resulted in higher measured viscosities than those measured using the smooth spindle (SC27).
 The coefficient of variation of the measured viscosities was significantly lower in the case of the vane spindle versus the smooth spindle, indicating that the vane spindle was more appropriate for measuring the viscosity of the mixtures/binders.
 The viscosity temperature susceptibility (VTS) of the binder was increased by increasing the percentage of RAS added to the virgin asphalt. Using either spindle, 20% RASmodified binder was less temperature susceptible than 30% and 40% RAS-modified binders at all temperatures and rotational speeds.
 Rediset®, Evotherm®, and Bio-binder modifiers reduced the VTS of the binders. Among them, Bio-binder reduced the VTS effectively when using either spindle. Overall, use of the vane spindle resulted in higher measured VTS values than use of the smooth spindle.
 The shear susceptibility for 40% RAS-modified mixture was more consistent than the 30% and 20% RAS-modified mixtures.
 The shear susceptibility of the Bio-modified binder was found to be more consistent in all percentages of RAS compared to the Rediset®-and Evotherm®-modified binders.
 The blending index was measured by using an empirical relation. Results indicated that the blending index increased as temperature increased. All modified binders showed higher blending indices at a temperature of 150 o C compared to samples measured at 135 o C, 120 o C, and 105 o C.
 Overall blending index was higher in amine-based modifier's modified binder compared to only RAS-modified binders at all temperatures.
 Comparing the blending index of the Redist®-, Evotherm®-, and Bio-binder modified binders, the Bio-binder showed the best results at all temperature tested at a rotational speed of 20 rpm.
 Among the blending indices evaluated at 135 o C and 20 rpm for 20%, 30%, and 40% RAS-filled medium, the highest value was found in Bio-binder modified mixtures.
Additionally, 20% RAS-filled media showed higher results than 30% and 40% RASfilled media (mixtures) indicating that Bio-binder most effectively increases mixing between aged and unaged asphalt in the mixture.
 The dynamic shear rheometer test was conducted for all modified and unmodified binders at 20% RAS content viscous media. The complex moduli (G*) for modified binders were higher than the control (PG 64-22) binder. Furthermore, incorporation of amine-based modifiers into the control decreased the G*.
 The phase angle (δ) was found to be lower in modified binders compared to the nonmodifiers content binder indicating that the amine-based modifiers make binder more elastic, which is only possible when thorough mixing of the ingredients occurs in the mixture.
 At higher temperatures (lower frequencies) Bio-modified binders show higher values of G* than the others but at lower temperatures (higher frequencies) it showed lower G* than the other binders. This finding indicates that incorporation of Bio-binder into the RAS-modified mixture at lower temperatures is more beneficial in terms of reduction of the mixture stiffness.
 Among the three modifiers, the Bio-binder reduced the G* and δ in the mixture effectively and enhanced the mixing between RAS and virgin asphalt in the mixture.
In summary, the addition of Bio-binder to partially RAS-filled viscous mixtures reduces the viscosity, temperature susceptibility, shear susceptibility, complex modulus, and phase angle and enhances the blending index of the asphalt binders tested.

Future Research
This study focused primarily on three amine-based modifiers and their application to enhance rheological characteristics of asphalt binder. Further research is needed to specify interaction mechanisms between each of these modifiers and asphalt molecules. In addition, determining the optimum percentage of each additive should be determined in order to maximize the blending of the modified binder. As such, the following recommendations are made for future studies:  study molecular interactions between modifiers and asphalt (aged as well unaged),