- Email: [email protected]
Aging of bituminous mixes for rap simulation
Construction and Building Materials 68 (2014) 49–54
Contents lists available at ScienceDirect
Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat
Aging of bituminous mixes for rap simulation Aybike Ongel ⇑, Martin Hugener 1 Laboratory for Road Engineering/Sealing Components, EMPA, Swiss Federal Laboratories for Materials Science and Technology, CH-8600 Duebendorf, Switzerland
h i g h l i g h t s Aging of asphalt is solely controlled by oxidation. Aging of asphalt initially increases rapidly then slows down to a constant rate. Humidity slows down the asphalt aging process. There is no difference in age hardening for different asphalt sample depths. Asphalt aging is affected by trace substances in the air.
a r t i c l e
i n f o
Article history: Received 20 March 2014 Received in revised form 16 June 2014 Accepted 18 June 2014
Keywords: Bitumen Asphalt recycling Asphalt aging Rheological properties RAP Laboratory aging tests
a b s t r a c t The use of reclaimed asphalt pavement (RAP) has gained considerable importance due to increasing environmental concerns and the search of more economic ways in road paving. Although a large amount of RAP is produced each year, only a part of it can be recycled in a suitable way. The efforts have been towards increasing the RAP content, and if possible, making complete use of RAP in road pavement construction. There have been many studies conducted on the aging of bitumen and compacted bituminous mixes, however the effects of aging conditions on the loose bituminous mixture to simulate RAP has not been examined in detail. As part of a research project on hot mix asphalt recycling, the effect of temperature, aging duration, humidity, oxygen concentration, and sample thickness on loose asphalt aging for the production of RAP was investigated. Bituminous mixture was aged in the forced draft oven in a special aging box and representative samples of asphalt were taken at defined time intervals for rheological tests. Hyperbolic-like curves were observed for rheological properties of bitumen versus aging duration. Asphalt aging was shown to be not only affected by the temperature, oxygen content, aging duration, and humidity but also the presence of trace substances in the air. No effect of layer thickness on the bitumen aging was found. Ó 2014 Elsevier Ltd. All rights reserved.
1. Introduction In the last decade, there has been growing concern about the increased demand placed on natural resources such as aggregates and crude oil for the bitumen production. The use of reclaimed asphalt pavement (RAP) has been a viable alternative for transportation agencies and asphalt producers to make more efficient use of the resources. RAP material is a reusable mixture of bitumen and aggregates that is generated from milling and/or crushing of old and damaged pavements for addition into new asphalt mixes. There is a trend towards increasing the RAP content in hot mix ⇑ Corresponding author. Permanent Address: Bahcesehir University, Ciragan Cad. No: 4, 34353 Besiktas/Istanbul, Turkey. Tel.: +90 555 690 5508. E-mail addresses: [email protected] (A. Ongel), [email protected] (M. Hugener). 1 Tel.: +41 58 765 4487. http://dx.doi.org/10.1016/j.conbuildmat.2014.06.030 0950-0618/Ó 2014 Elsevier Ltd. All rights reserved.
asphalt not only because it helps conserve non-renewable resources but also eliminates disposal problems as it reduces the need for landfills, and reduces energy consumption and costs. The construction of new asphalt plants with special drums for the addition of high percentages of RAP makes it technically possible to recycle asphalt up to 100%. However, there are concerns about the durability of pavements with high RAP content which impose restrictions in many countries on the maximum percentages of RAP incorporated into the mix with respect to rutting, fatigue life, and durability. In order to overcome the restrictions on the maximum RAP percentages, extended research activities have started all over the world. RAP generally has to be produced artificially in the laboratory, since the RAP available from the roads is not suitable to study specific conditions such as effects of repeated recycling on the asphalt properties. However, there is not yet a standardized procedure for
50
A. Ongel, M. Hugener / Construction and Building Materials 68 (2014) 49–54
the production of RAP in the laboratory. The RILEM TC-ATB-TG5 group aimed to develop an aging methodology to reproduce aging of bituminous mixtures until the end of their service life and simulate RAP in the laboratory [1]. However, before adapting a test protocol to simulate RAP, there is a need to understand the effects of temperature, aging duration, asphalt layer thickness, humidity, and oxygen concentration on the aging of the loose bituminous mix. In general, asphalt binder aging has two distinct phases: shortterm and long-term. Aging occurring during production and transportation of asphalt at elevated temperatures due to volatilization and oxidation is referred to as short-term aging. Oxidation that continues through the service life of the pavement at ambient temperatures, due to exposure to sunlight, temperature variations, rainfall and traffic loading after pavement is opened to traffic, is referred to as long-term aging. Peterson [2] summarized three factors contributing to the aging of the bituminous binder: loss of oily components by volatility or adsorption to mineral aggregates, changes in the chemical composition by reaction with atmospheric oxygen, and molecular structuring that produces thixotropic effects (steric hardening). Volatilization is the loss of lighter constituents of bitumen resulting in hardening of bitumen and is mainly controlled by temperature. Penetration graded bitumen below a penetration of about 100 0.1 mm, which are commonly used in hot and moderate climates, show little volatilization and therefore insignificant hardening [3]. Oxidation is the interaction of bitumen with oxygen from the environment resulting in a change in the composition of asphalt due to the formation of polar functional groups containing oxygen. The newly formed polar groups cause marked increases in bitumen stiffness and viscosity as they have stronger molecular interactions [2]. For all bitumen, the viscosity change with time on oxidation follows a hyperbolic curve, an initial rapid increase in viscosity followed by a leveling off to a slower and rather constant rate [4–7]. In addition to the formation of polar functional groups by oxidation, bitumen properties can also be significantly altered by isothermal and reversible molecular structuring, called steric hardening [2,8]. However, it is usually hard to quantify since structuring is destroyed during the solvent recovery of binder from the aged pavement [9]. Oxidation has been shown to be the principal factor leading to the hardening and embrittlement of bitumen [2,10]. Oxidation rate depends on bitumen composition and diffusion rate of oxygen [11,12]. Oxygen reactivity increases with increasing fraction polarity, therefore different bitumens would age differently under the same conditions [13]. In addition, the diffusion coefficient of oxygen increases with increasing temperature, and decreasing viscosity (penetration) and chemically bound oxygen in bitumen [11,12]. Temperature affects the oxidation process in terms of both reaction rate and relative amounts of the oxidation products [13]. According to Arrhenius Law, an approximately 10 °C increase in temperature would double the rate of most oxidation reactions [14]. However, Herrington et al. [15] suggested that asphalts have a limiting viscosity on oxidation for each temperature as the relative rates of oxidation products and the reaction extent would be different at different temperatures. Therefore, temperature effects on oxidation kinetics may not be demonstrated for all points on the asphalt viscosity-time curve at different temperatures. Humidity or water content was also shown to influence the age hardening of bitumen. However, the findings have been quite variable. Polymer modified binders experienced less hardening when aged in the presence of water compared to that aged under dry conditions [16]. The retarding effect of water has been attributed to the higher amount of chemically bound oxygen found in bitumen in presence of water which would reduce the diffusion of oxygen rate [16,17]. However, Hagos et al. [18] and Campbell
and Wright [19] showed increased oxidation with the presence of water during aging. Thomas [20] suggested that the effect of water during aging on asphalt hardening, as in the case of oxidation extent, depends on the fractional composition of the bitumen. The extent of aging was also shown to vary throughout the depth of the pavement. The aging of bitumen decreases with increasing distance from the surface [21,22]. Coons and Wright [23] showed that only the top 12.5 mm of the dense pavements was shown to have increased viscosity. This effect was attributed to the decreasing maximum temperatures as well as less exposure to oxygen with increasing depth in the pavement. However, the relative contribution of these factors is unknown. Aging data are generally obtained from laboratory aging tests and used to help predict the long-term performance of asphalt pavements in the field. Most of the research has focused on the factors contributing to aging of bituminous binder. However, aging of bituminous mixtures can be quite different than the bitumen itself. When bitumen is mixed with aggregates, the bitumen molecules interact with aggregates. Aggregates may smooth [24] or catalyze [25] the oxidation process, therefore there is a need to take into account bitumen-aggregate interaction effects on aging. An aging study on bitumen–filler mixes showed that both bitumen and filler influence the binder hardening [26]. The short-term and long-term oven aging procedures developed under the SHRP-A-003A [5] have been the most widely adapted aging protocols for asphalt. The SHRP short-term oven ageing (STOA) procedure requires that prior to compaction, the loose bituminous mix should be aged in a forced draft oven at 135 °C for 4 h. The SHRP long-term oven aging (LTOA) procedure is a continuation of the STOA treatment, where the compacted samples are heated in the oven at 85 °C for further 120 h. When the results were compared with field aging (mixes with typically 5% air voids), it was found that the STOA method was roughly equivalent to 0–2 years in the field, and the LTOA method to 5–15 years, depending on climate. There are other accelerated long-term aging protocols for bituminous mixes including the use of higher temperatures (100 °C and above), high oxygen/air pressures, and feeding air/oxygen or humid air into the mix. However, all these methods use compacted samples and in order to produce RAP in the laboratory, these samples should be reheated to obtain a loose mix. A study was initiated at EMPA in order to verify the earlier findings on aging and investigate the aging behavior of loose bituminous mixes. This helps determine the aging parameters relevant for the RAP simulation in the laboratory. In this study, a dense asphalt mixture, AC 11 was produced in the laboratory and aged in a special aging box in the draft force oven at different temperatures, durations, oxygen concentrations, and humidity levels. The loose bituminous mixture was sampled at different time intervals and tested for its rheological properties on the recovered binder in order to follow the aging progression. The effect of sample thickness on aging was also investigated.
2. Methodology 2.1. Experimental test setup In this study, a loose asphalt sample was aged using a forced draft oven. Since the main objective was to compare different variables affecting aging over time, only long-term aging was assessed. A dense asphalt mixture, AC 11 prepared according to EN 13108-1 [27] with standard penetration grade bitumen 70/100 was used in testing. Binder content of the mix was 6.12% by mass. The aggregate gradation of the AC11 mix is given in Table 1. A batch of 150 kg of asphalt mixture was prepared in the lab at a mixing temperature of 160 °C. After production, it was partitioned and placed into boxes, each containing 6 kg of the mix, and stored at room temperature. Asphalt mix samples of 6 kg were pre-heated in an oven at 120 °C for 30 min then placed into a metal aging box with an approximately 3.5 cm thick even layer. The box had a volume of 0.01 m3 (10 28 35 cm) as shown in Fig. 1, and was
A. Ongel, M. Hugener / Construction and Building Materials 68 (2014) 49–54
51
Table 1 AC11 particle size distribution. Sieve (mm)
Percent passing
Specification requirement
0.063 0.125 0.25 0.5 1.0 2.0 2.8 4.0 5.6 8.0 11.2 16.0
7.3 9.6 13.6 19.3 26.2 34.8 43.0 56.2 70.3 87.9 98.7 100
6.4–10.4 – – – 20.4–28.4 29.9–41.9 – 48.8–62.8 – – 91.5–100 Fig. 3. Humid air generation setup with flow meter.
Opening for temperature sensor
Gas in/outlet
Fig. 1. Loose asphalt mix in the metal aging box.
sealed with silicon in order to make it airtight. The cover of the box consisted of three openings: two openings for the in- and outlet of air or nitrogen and a third one for two temperature probes that record the temperature at different points within the mix. Air and nitrogen coming from pressure cylinders were preheated by passing these gases through a copper coil prior to entering the aging box in order to prevent the cooling of the asphalt mix by the incoming gas. The incoming air temperature was verified at 100 ± 1 °C. The setup of the aging process in the forced draft oven is shown in Fig. 2. Gas flow into the aging box was set at a constant rate and monitored with an adjustable flow meter during the aging processes. Fig. 3 shows the adjustable flow meter with the humid air generator. The temperature of the asphalt sample was also recorded every minute during aging using a data
Fig. 2. Closed box aging setup.
logger and two temperature probes. Asphalt samples of 600 ± 50 g were taken every 24–48 h for testing. The initial sample, Day 0, corresponds to the mix before aging, but after 30 min of pre-heating at 120 °C. Bitumen was extracted from the asphalt by toluene using the cold extraction method. The asphalt sample was soaked in cold toluene for 15 min in order to dissolve the bitumen. Then the mixture was stirred by hand and the overlaying asphalt toluene solution without aggregates was poured through a 0.063 mm sieve into a glass container. The filler, which passed the 0.063 mm sieve, was separated from the toluene solution using a centrifuge according to EN 12697-1 [28]. Following the asphalt extraction, the bitumen was recovered from the toluene solution using a rotary evaporator at 145 °C and pressure of initially 45 kPa and then 1.9 kPa according to EN 12697-3 [29]. The extent of aging was assessed for the recovered binder through the Ring and Ball (R&B) softening point method [30] and the complex shear modulus (G*) test performed with a Physica MCR 301 dynamic shear rheometer (Anton Paar, Austria) according to EN 14770 [31]. An approximately 1 mm thick sample of bitumen was prepared in a silicon mould and placed between two 25 mm parallel plates. The gap was adjusted to 1.00 mm and the complex modulus was measured at a temperature of 30 °C and at the frequencies of 1.25 and 4.4 Hz. However, only the complex modulus values at 1.25 Hz are presented in this paper. Two replicate tests were conducted for each asphalt mixture both for softening point and complex shear modulus determination.
2.2. Aging methods In this study, four sets of aging tests were run in order to test the effects of temperature, oxygen content, humidity, and asphalt mixture layer thickness. The experimental test set up is given in Table 2. A reference test was carried out heating the asphalt mix in the closed aging box in the force-draft oven at a temperature of 100 °C and an air-flow of 0.45 l/min for 5 days. Set 1 evaluates the temperature effects on aging. In addition to the reference test conducted at the temperature of 100 °C, aging tests at 85 °C and 115 °C were performed, designated as Temp1 and Temp2 in Table 2, respectively. Since aging is faster at higher temperature, aging duration was shortened to 3 days when tested at 115 °C while extended to 8 days at 85 °C. All tests were conducted at an airflow of 0.45 l/min. Set 2 evaluates the effects of oxygen content on aging. In this set, aging tests were conducted both feeding air and nitrogen into the asphalt mix. In addition to the reference test conducted at an airflow of 0.45 l/min, aging with an airflow of 4.5 l/min and aging in the absence of oxygen, feeding 100% nitrogen at a flow rate of 0.45 l/min were performed, designated as Air1 and Air2 in Table 2, respectively. All the Set 2 tests were conducted in a closed box at a temperature of 100 °C. Set 3 evaluates the effects of humidity on aging. The influence of water vapor on the aging behavior was examined by feeding a saturated air stream into the aging box at 100 °C at a flow rate of 0.45 l/min, designated as Hum1 in Table 2. The water saturated air was produced by purging air through a fine sintered metal filter making fine air bubbles, which rapidly saturated the air with water. Due to the temperature dependency of relative humidity (r.H.), the saturated air at 25 °C corresponds to a relative humidity of approximately 4% at 100 °C. Humidity could also have been increased by purging air into hot water, however, then, the vapor content would be more difficult to control. In this set, in addition to the reference test, the standard aging procedure was also performed which is essentially aging a layer of asphalt mix in the draft forced oven, designated as Hum2 in Table 1. In the standard procedure, aging was performed in the aging box, but with the cover left open, without external air feeding, at a temperature of 100 °C. The relative humidity of air in the pressure cylinder is approximately 0.0065% whereas the relative humidity in a non-conditioned laboratory typically varies between 40% and 80% at an ambient temperature of 18–26 °C, resulting in a relative humidity of 1–3% at 100 °C. However, open box aging is not a controlled test in terms of airflow rate or air purity, therefore the differences in the aging progression cannot be fully attributed to humidity.
52
A. Ongel, M. Hugener / Construction and Building Materials 68 (2014) 49–54
Table 2 Experimental test setup. Test code
Temperature (°C)
Gas flow rate (l/min)
Gas type
Layer thickness (mm)
Aging duration (days)
Reference Temp1 Temp2 Air1 Air2 Hum1 Hum2 Layer Thickness
100 85 115 100 100 100 100 100
0.45 0.45 0.45 4.5 0.45 0.45 Open box Open box
Dry air Dry air Dry air Dry air Dry nitrogen Water saturated air with r.H. of 100% at 25 °C Ambient air with r.H. of 40–80% at 25 °C Ambient air
35 35 35 35 35 35 35 80
5 8 3 5 5 5 5 5
Set 4 evaluates the effects of layer thickness on aging. An 8-cm-thick asphalt mix was aged at 100 °C in an open container, designated as layer in Table 2. In this set, sampling was conducted only at the end of the testing period, after 5 days of aging. Aging was evaluated for the top 4 cm and bottom 4 cm of the mix separately.
3. Results and discussion Data analysis showed that the softening point and complex shear modulus, G* before aging were not always the same for different batches containing 6 kg of virgin bituminous mixture. This can mainly be attributed to the storage time of the hot mix asphalt samples after production. Since bitumen testing and extraction steps are labor and time intensive, the laboratory tests had to be carried out successively and therefore storage time was not the same for all asphalt mix samples. Therefore asphalt mixtures stored longer may have aged more and got stiffer than those stored for a shorter time. Additionally, even though special effort has been given in order to follow the same procedure for each test, the handling differences during the preparation and testing of the binder, including the heating of the hot mix sample, and the extraction and rheological measurements of the binder, may have caused some measurement error. Therefore, for comparison of the aging progression, data have been normalized by dividing the softening point and complex shear modulus G* values of the aged binder by the values before aging, designated as Day 0. The ratio of aged binder value to that of unaged binder was defined as aging index. Aging index in terms of complex shear modulus was calculated using the logarithm of complex modulus G* values, log Gaged = log Gunaged , as the softening point is linearly proportional to the logarithm of the viscosity at a specific temperature. The repeatability of the R&B test according to EN 1427 [30] is given as ±1 °C for non-modified binders. DSR test allows 10% deviation from the mean value when repetitive samples are tested [31]. Therefore, aging index for softening point was given with error bars showing 1 °C deviation from the mean temperature, while aging index for complex modulus was given with error bars showing 10% deviation from the mean complex modulus value. A hyperbolic trend line was fitted to the data. Figs. 4 and 5 show the aging index for Set 1 tests for logarithm of complex modulus G* at 1.25 Hz and softening point, respectively. Asphalt aging was conducted at 85 °C, 100 °C, and 115 °C at a constant airflow of 0.45 l/min. According to Fig. 4, logarithm of complex shear modulus G* of the bitumen reached approximately 1.05, 1.07, and 1.06 times that of the unaged bitumen at the end of testing period when heated at temperatures of 85 °C, 100 °C, and 115 °C, respectively. According to Fig. 5, the softening point of the bitumen reached approximately 1.1, 1.15, and 1.2 times that of the unaged bitumen at the end of testing period for aging temperatures of 85 °C, 100 °C, and 115 °C, respectively. Higher aging indices in terms of softening point may be due to softening point test being more sensitive to temperature changes compared to shear modulus test. It can be seen from Fig. 4 that aging process is not complete yet at the end of the testing period as the complex shear modulus G* is still increasing. However, according to Fig. 5, age hardening slows
Fig. 4. Aging index for logarithm of complex modulus G* at 30 °C at 1.25 Hz for bitumen aged at different temperatures.
Fig. 5. Aging index for softening point for bitumen aged at different temperatures.
down for temperatures of 85 °C and 100 °C, and still goes on for 115 °C. This may indicate that aging rate in terms of complex shear modulus G* is slower than that of softening point. Hence, the aging indices in terms of shear modulus may reach to those of softening point when longer aging durations are applied. In general, the rate of aging goes down after the third day of the testing period. Since the age hardening is still in progress, even though at a slower rate, at the end of the testing period, it can be concluded that the aging durations chosen were too short in order to obtain the final aging index value, where there is no further age hardening. The rate of increase in age hardening goes up with increasing temperature both for complex modulus G* and softening point. An approximately doubling of oxidation rate was expected with every 10 °C increase in temperature according to the Arrhenius Law. Therefore, aging rate at 100 °C and 115 °C was expected to be three times and eight times, respectively of that at 85 °C. However, it can be seen from Figs. 4 and 5 that the aging rate increases much less than expected which may be due to asphalt-aggregate interaction slowing down the aging process as suggested by Recasens et al. [24]. It can also be seen from the figures that the aging value at 85 °C will never reach the aging indices at 100 °C or at 115 °C regardless of aging time. This indicates that different chemical reactions are responsible for the age hardening of bitumen at different temperatures as suggested by Herrington et al. [15]. Chemical reactions require specific activation energy to start. At 85 °C the activation
53
A. Ongel, M. Hugener / Construction and Building Materials 68 (2014) 49–54
energy may be too low for some chemical reactions taking part in the oxidation of bitumen, which is a complex series of chemical reactions, not only including the incorporation of oxygen but also oxygen induced or catalyzed reactions [32]. This is why the oxygen content as determined by elemental chemical analysis does not increase significantly during oxidation. Standard bitumen contain only around 0.5% (mass percentage) of oxygen and even with severe air blowing conditions using pure oxygen, only an increase to 1% has been shown [33,34]. Figs. 6 and 7 show the aging index for Set 2 and 3 tests for logarithm of complex shear modulus, G* and softening point, respectively. Asphalt samples were aged in an open box; and in a closed box feeding air with a flow rate of 0.45 l/min and 4.5 l/ min, feeding air at a flow rate of 0.45 l/min with relative humidity of 4% and feeding nitrogen at a flow rate of 0.45 l/min. All the aging tests have been carried out at 100 °C. Instead of a hyperbolic trend line, data points were connected with a straight line for the test in the open box, since aging process stopped/slowed down after 2 days. It can be seen from Figs. 6 and 7 that age hardening stopped or slowed down to a constant rate at the end of the testing period. When pure nitrogen was fed through the asphalt sample, the aging index fluctuated around 1, showing no age hardening. It can be concluded that nitrogen inhibits the aging efficiently even at 100 °C, indicating the necessity of oxygen to be present for age hardening. There may be chemical reactions taking place in the presence of nitrogen, however they do not affect the viscosity of the binder significantly. Besides, this experiment shows that there is no leakage of oxygen into the aging box. When the airflow was increased from 0.45 l/min to 4.5 l/min, a slightly higher aging index compared to the reference procedure was obtained. When humid air was used instead of dry air, there was a reduction in the aging index both for complex modulus and softening point. Since the water was in vapor state and relative
Table 3 Complex modulus and softening point for the top and bottom samples. Complex modulus G*, Pa Top Bottom Ratio: Top/bottom
6
1.27 ⁄ 10 1.21 ⁄ 106 1.05
Softening point, °C 63.3 63.5 0.99
humidity was only around 4% at 100 °C, it was not possible that the binder was protected by a water film against oxygen molecules as suggested by Negulescu et. al. and Daranga [16,17]. The reduction in aging index may be due to the water molecules scavenging reactive compounds which have been produced by the reaction with oxygen or even oxygen radicals. Aging rate was the highest when asphalt was aged in the open box, which represents the traditional oven aging. After the second day of aging process, there was no significant increase in age hardening in terms of complex modulus or softening point. However, after 6 days of aging at 100 °C, the aging index derived from the softening point reached nearly the same value as when aged in the closed box at an airflow of 4.5 l/min. The faster rate of age hardening in the open box can be explained with the presence of reactive trace substances. Since the air in the lab is not purified like the air coming from the pressure cylinder, trace substances like ozone, nitro- and sulfur oxides could be present which may affect the aging behavior of bitumen. This indicates that oxygen concentration in terms of both airflow rate and air purity control the aging rate. But further research is necessary to find the relevant trace substances which are responsible for the accelerated aging. In the Set 4 testing, the complex modulus G* at 30 °C at 1.25 Hz and softening point of the top 4 cm and bottom 4 cm of the 8-cmthick asphalt mix were evaluated as shown in Table 3. The complex modulus G* and softening point of the top and bottom part are not significantly different from each other, within the standard error of the tests. Therefore, it can be concluded that there is no significant difference in aging between the top and bottom sections. This does not confirm the earlier findings that aging goes down with increasing depth. However, there may be homogenous temperature distribution and exposure to oxygen throughout the loose mix, different than that of the pavements in the field. It could be as well due to the fact that aging takes place only in a very thin layer of the surface mix with negligible impact on the whole layer as suggested by Coons and Wright [23]. 4. Summary and conclusions
Fig. 6. Aging index for logarithm of complex modulus G* at 30 °C at 1.25 Hz for bitumen aged at 100 °C.
Fig. 7. Aging index for softening point for bitumen aged at 100 °C.
In order to verify the earlier findings and simulate RAP in the laboratory, the effect of temperature, aging duration, humidity, oxygen concentration, and layer thickness on the age hardening of loose bituminous mixes was evaluated. Aging methods used in this study included: heating the loose asphalt at 85 °C, 100 °C, and 115 °C for 3 days, 5 days, and 8 days, respectively, feeding air at a flow rate of 0.45 l/min; heating the loose asphalt in an open box, and in a closed box feeding air at a flow rate of 4.5 l/min, feeding humid air at a flow rate of 0.45 l/min, and feeding nitrogen at a flow rate of 0.45 l/min at 100 °C for 5 days; and heating 8-cm-thick asphalt mix in an open pan at 100 °C for 5 days. There was no aging observed when nitrogen instead of air was fed through the sample, indicating aging of the bituminous mix was solely controlled by oxidation. Aging index was found to increase with increasing airflow rate and hence oxygen concentration at a given temperature. Aging rate also increases with increasing temperature. Doubling in oxidation rate was expected with 10 °C increase in temperature according to Arrhenius Law. However, aging rate increase
54
A. Ongel, M. Hugener / Construction and Building Materials 68 (2014) 49–54
in terms of both complex modulus and softening point was shown to be slower which may be due to the smoothing effect of aggregate–bitumen interaction. Aging usually slows down after the third day of testing. Hyperbolic-like curves, an initial rapid increase in aging indices followed by a slower rate, were observed for tests conducted at 85 °C and 100 °C. Age hardening still goes on at a rapid rate after 3 days for testing at 115 °C. Therefore, aging tests should be conducted for longer durations in order to observe the final aging index value. Final aging indices are different at the temperatures 85, 100 and 115 °C. This leads to the conclusion that the chemical reactions change with temperature as indicated by earlier findings. Humidity was found to slow down the aging process. In contrast to earlier findings, no difference in age hardening was observed for different sample depths. Hardening was the same for the top and bottom layer of the bituminous mixture. R&B softening point test was found to be more sensitive to temperature changes compared to DSR measurements since the rate of change in aging index was slower for the dynamic complex modulus. There are indications that certain trace substances in the air of the laboratory promote the aging of the asphalt mix. Further experiments are necessary to identify the responsible trace substances.
Based on the findings, the testing protocol for RAP simulation should be conducted in a closed box controlling the oxygen content, since trace substances affect the aging of the bitumen. Bituminous mixture should be used in the aging process rather than the bitumen, due to the fact that aggregates may affect the aging rate of the bitumen. Aging temperature should be selected based on the rheological properties of the bitumen extracted from the recycled pavement since the limit for age hardening varies with aging temperature. Age hardening was not significantly different for the top and bottom layer of the bituminous mix, therefore there is no need to stir the loose mix during the aging process for RAP simulation. In the future, the results should also be correlated with the values obtained from RAP from the field.
References [1] De la Roche C, Van de Ven M, Van den Bergh W, Gabet T, Dubois V, Grenfell J, et al. Development of a laboratory bituminous mixture ageing protocol. Advanced testing and characterization of bituminous materials. Taylor and Francis Group; 2009. [2] Peterson JC. Chemical composition of asphalt as related to asphalt durabilitystate of the art. Transp Res Rec 1984;999:13–30. [3] Zoorop SE, Collop AC, Brown SF. Performance of bituminous and hydraulic materials in pavements. In: Proceedings of the fourth European symposium on performance of bituminous and hydraulic materials in pavements. Bitmat 4, Nottingham, United Kingdom; 2002. [4] Petersen JC, Harnsberger PM, Robertson RE. Factors affecting the kinetics and mechanisms of asphalt oxidation and the relative effects of oxidation products on age hardening. Prepr Pap Am Chem Soc Div Fuel Chem 1996;41(4):1232–44. [5] Bell CA, Sosnovske D. Validation of the A-002A hypothesis for ageing. TM-OSUA- 003A-92-24, final summary report to SHRP: 92-2, Strategic Highway Research Program, National Research Council, Washington, D.C.; 1992.
[6] Von Quintas H, Scherocman J, Kennedy T, Hughes CS. Asphalt aggregate mixture analysis system. NCHRP report No: 338, National Highway Research Program, Transportation Research Board, National Research Council, Washington, D.C.; 1988. [7] Bell CA, Sosnovske D. Aging: binder validation. SHRP-A-384, Strategic Highway Research Program, National Research Council, Washington, D.C.; 1994. [8] Huang SC, Robertson RE, Branthaver JF, McKay JF. Study of steric hardening effect of thin asphalt films in presence of aggregate surface. Transp Res Rec 1999;1661:15–21. [9] Petersen, JC. Chemical composition of asphalt as related to asphalt durability in developments. Petroleum Science, 40: Part B: 363–395, Asphaltenes and Asphalts, 2; 2000. [10] Griffin RL, Miles TK, Penther CJ. Microfilm durability test for asphalt. Proc Assoc Asphalt Paving Techonol 1955;24:31–62. [11] Van Oort WP. Durability of asphalt: its aging in dark. Ind Eng Chem 1956;48(7):1196–201. [12] Negulescu I, Mohammad L, Daly W, Abadie C, Cueto R, Daranga C, et al. Chemical and rheological characterization of wet and dry aging of SBS copolymer modified asphalt cements: laboratory and field evaluation. J Assoc Asphalt Paving Technol 2006;75:267–96. [13] Petersen JC. A Review of the Fundamentals of asphalt oxidation: chemical, physicochemical, physical property, and durability relationships. Transportation Research Circular E-C140; 2009. [14] Gawel I, Baginska K. Effect of the chemical nature on susceptibility of asphalt to aging. Pet Sci Technol 2004;22:1261–71. [15] Herrington PR, Patrick JE, Ball GFA. Oxidation of roading asphalts. Ind Eng Chem Res 1994;33:2801–9. [16] Negulescu I, Mohammad L, Daly W, Chiparus I. The use of DMA to characterize the aging of asphalt binders. FHWA/LA.07/461; 2010. [17] Daranga C. Characterization of aged polymer modified asphalt cements for recycling purposes. Dissertation, Graduate Faculty of the Louisiana State University and Agricultural and Mechanical College; 2005. [18] Hagos ET, Molenaar AAA, Van de Ven MFC. Chemical characterization of laboratory and field bitumen aging in porous asphalt concrete. In: Advanced testing and characterization of bituminous materials, two volume set. CRC Press; 2009. [19] Campbell PG, Wright JR. The effect of temperature and humidity on the oxidation of airblown asphalts. Am Chem Soc, Div Pet Chem Prepr 7 1962;3:123. [20] Thomas KP. Impact of water during the laboratory aging of asphalt. Road Mater Pavement Des 2002;3(3):299–315. [21] Al-Khateeb G, Qi X, Shenoy A, Stuart K, Mitchell T. Assessment of aging at FHWA’s pavement testing facility. Transp Res Rec 2005;1940:146–55. [22] Li NL, Zhao XP, Du C, Zhang CL, Xiao QY. Effect of aging on urban asphalt pavement. Adv Mater Res 2013;1287:671–4. [23] Coons RF, Wright PH. An investigation of the hardening of asphalt recovered from pavements of various ages. Proc Assoc Asphalt Paving Technol 1967;37:510–28. [24] Recasens RM, Martinez A, Jimenez FP, Bianchetto H. Effect of filler on the aging potential of asphalt mixtures. Transp Res Rec 2005;1901:10–7. [25] Petersen JC, Barbour FA, Dorrence SM. Catalysis of asphalt oxidation by mineral aggregate surface and asphalt components. Proc Assoc Asphalt Paving Technol 1974;43:162–77. [26] Gubler R. Einfluss schweizerischer filler auf die alterung von bituminösen bindemitteln und die rissbildung im belag. Eidgenössisches Departement für Umwelt, Verkehr, Energie und Kommunikation – Bundesamt für Strassen 2005; 1138. [27] EN 13108-1:2006 Bituminous mixtures – material specifications – Part 1: asphalt concrete. [28] EN 12697-1:2005. Bituminous mixtures – test methods for hot mix asphalt – Part 1: soluble binder content. [29] EN 12697-3:2005. Bituminous mixtures – test methods for hot mix asphalt – bitumen recovery. Rotary evaporator. [30] EN 1427:2007. Bitumen and bituminous binders – determination of the softening point. Ring and Ball method. [31] EN 14770:2005. Bitumen and bituminous binders – determination of complex shear modulus and phase angle – Dynamic Shear Rheometer (DSR). [32] Moschopedis S, Speight J. The effect of air blowing on the properties and constitution of a natural bitumen. J Mater Sci 1977;12(5):990–8. [33] Read J, Whiteoak D. The Shell bitumen handbook. London: Thomas Telford Publishing; 2003. [34] Campbell PG, Wright JR. Ozonation of asphalt flux. Ind Eng Chem Prod Res Dev 1964;3(3):186–94.
Contents lists available at ScienceDirect
Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat
Aging of bituminous mixes for rap simulation Aybike Ongel ⇑, Martin Hugener 1 Laboratory for Road Engineering/Sealing Components, EMPA, Swiss Federal Laboratories for Materials Science and Technology, CH-8600 Duebendorf, Switzerland
h i g h l i g h t s Aging of asphalt is solely controlled by oxidation. Aging of asphalt initially increases rapidly then slows down to a constant rate. Humidity slows down the asphalt aging process. There is no difference in age hardening for different asphalt sample depths. Asphalt aging is affected by trace substances in the air.
a r t i c l e
i n f o
Article history: Received 20 March 2014 Received in revised form 16 June 2014 Accepted 18 June 2014
Keywords: Bitumen Asphalt recycling Asphalt aging Rheological properties RAP Laboratory aging tests
a b s t r a c t The use of reclaimed asphalt pavement (RAP) has gained considerable importance due to increasing environmental concerns and the search of more economic ways in road paving. Although a large amount of RAP is produced each year, only a part of it can be recycled in a suitable way. The efforts have been towards increasing the RAP content, and if possible, making complete use of RAP in road pavement construction. There have been many studies conducted on the aging of bitumen and compacted bituminous mixes, however the effects of aging conditions on the loose bituminous mixture to simulate RAP has not been examined in detail. As part of a research project on hot mix asphalt recycling, the effect of temperature, aging duration, humidity, oxygen concentration, and sample thickness on loose asphalt aging for the production of RAP was investigated. Bituminous mixture was aged in the forced draft oven in a special aging box and representative samples of asphalt were taken at defined time intervals for rheological tests. Hyperbolic-like curves were observed for rheological properties of bitumen versus aging duration. Asphalt aging was shown to be not only affected by the temperature, oxygen content, aging duration, and humidity but also the presence of trace substances in the air. No effect of layer thickness on the bitumen aging was found. Ó 2014 Elsevier Ltd. All rights reserved.
1. Introduction In the last decade, there has been growing concern about the increased demand placed on natural resources such as aggregates and crude oil for the bitumen production. The use of reclaimed asphalt pavement (RAP) has been a viable alternative for transportation agencies and asphalt producers to make more efficient use of the resources. RAP material is a reusable mixture of bitumen and aggregates that is generated from milling and/or crushing of old and damaged pavements for addition into new asphalt mixes. There is a trend towards increasing the RAP content in hot mix ⇑ Corresponding author. Permanent Address: Bahcesehir University, Ciragan Cad. No: 4, 34353 Besiktas/Istanbul, Turkey. Tel.: +90 555 690 5508. E-mail addresses: [email protected] (A. Ongel), [email protected] (M. Hugener). 1 Tel.: +41 58 765 4487. http://dx.doi.org/10.1016/j.conbuildmat.2014.06.030 0950-0618/Ó 2014 Elsevier Ltd. All rights reserved.
asphalt not only because it helps conserve non-renewable resources but also eliminates disposal problems as it reduces the need for landfills, and reduces energy consumption and costs. The construction of new asphalt plants with special drums for the addition of high percentages of RAP makes it technically possible to recycle asphalt up to 100%. However, there are concerns about the durability of pavements with high RAP content which impose restrictions in many countries on the maximum percentages of RAP incorporated into the mix with respect to rutting, fatigue life, and durability. In order to overcome the restrictions on the maximum RAP percentages, extended research activities have started all over the world. RAP generally has to be produced artificially in the laboratory, since the RAP available from the roads is not suitable to study specific conditions such as effects of repeated recycling on the asphalt properties. However, there is not yet a standardized procedure for
50
A. Ongel, M. Hugener / Construction and Building Materials 68 (2014) 49–54
the production of RAP in the laboratory. The RILEM TC-ATB-TG5 group aimed to develop an aging methodology to reproduce aging of bituminous mixtures until the end of their service life and simulate RAP in the laboratory [1]. However, before adapting a test protocol to simulate RAP, there is a need to understand the effects of temperature, aging duration, asphalt layer thickness, humidity, and oxygen concentration on the aging of the loose bituminous mix. In general, asphalt binder aging has two distinct phases: shortterm and long-term. Aging occurring during production and transportation of asphalt at elevated temperatures due to volatilization and oxidation is referred to as short-term aging. Oxidation that continues through the service life of the pavement at ambient temperatures, due to exposure to sunlight, temperature variations, rainfall and traffic loading after pavement is opened to traffic, is referred to as long-term aging. Peterson [2] summarized three factors contributing to the aging of the bituminous binder: loss of oily components by volatility or adsorption to mineral aggregates, changes in the chemical composition by reaction with atmospheric oxygen, and molecular structuring that produces thixotropic effects (steric hardening). Volatilization is the loss of lighter constituents of bitumen resulting in hardening of bitumen and is mainly controlled by temperature. Penetration graded bitumen below a penetration of about 100 0.1 mm, which are commonly used in hot and moderate climates, show little volatilization and therefore insignificant hardening [3]. Oxidation is the interaction of bitumen with oxygen from the environment resulting in a change in the composition of asphalt due to the formation of polar functional groups containing oxygen. The newly formed polar groups cause marked increases in bitumen stiffness and viscosity as they have stronger molecular interactions [2]. For all bitumen, the viscosity change with time on oxidation follows a hyperbolic curve, an initial rapid increase in viscosity followed by a leveling off to a slower and rather constant rate [4–7]. In addition to the formation of polar functional groups by oxidation, bitumen properties can also be significantly altered by isothermal and reversible molecular structuring, called steric hardening [2,8]. However, it is usually hard to quantify since structuring is destroyed during the solvent recovery of binder from the aged pavement [9]. Oxidation has been shown to be the principal factor leading to the hardening and embrittlement of bitumen [2,10]. Oxidation rate depends on bitumen composition and diffusion rate of oxygen [11,12]. Oxygen reactivity increases with increasing fraction polarity, therefore different bitumens would age differently under the same conditions [13]. In addition, the diffusion coefficient of oxygen increases with increasing temperature, and decreasing viscosity (penetration) and chemically bound oxygen in bitumen [11,12]. Temperature affects the oxidation process in terms of both reaction rate and relative amounts of the oxidation products [13]. According to Arrhenius Law, an approximately 10 °C increase in temperature would double the rate of most oxidation reactions [14]. However, Herrington et al. [15] suggested that asphalts have a limiting viscosity on oxidation for each temperature as the relative rates of oxidation products and the reaction extent would be different at different temperatures. Therefore, temperature effects on oxidation kinetics may not be demonstrated for all points on the asphalt viscosity-time curve at different temperatures. Humidity or water content was also shown to influence the age hardening of bitumen. However, the findings have been quite variable. Polymer modified binders experienced less hardening when aged in the presence of water compared to that aged under dry conditions [16]. The retarding effect of water has been attributed to the higher amount of chemically bound oxygen found in bitumen in presence of water which would reduce the diffusion of oxygen rate [16,17]. However, Hagos et al. [18] and Campbell
and Wright [19] showed increased oxidation with the presence of water during aging. Thomas [20] suggested that the effect of water during aging on asphalt hardening, as in the case of oxidation extent, depends on the fractional composition of the bitumen. The extent of aging was also shown to vary throughout the depth of the pavement. The aging of bitumen decreases with increasing distance from the surface [21,22]. Coons and Wright [23] showed that only the top 12.5 mm of the dense pavements was shown to have increased viscosity. This effect was attributed to the decreasing maximum temperatures as well as less exposure to oxygen with increasing depth in the pavement. However, the relative contribution of these factors is unknown. Aging data are generally obtained from laboratory aging tests and used to help predict the long-term performance of asphalt pavements in the field. Most of the research has focused on the factors contributing to aging of bituminous binder. However, aging of bituminous mixtures can be quite different than the bitumen itself. When bitumen is mixed with aggregates, the bitumen molecules interact with aggregates. Aggregates may smooth [24] or catalyze [25] the oxidation process, therefore there is a need to take into account bitumen-aggregate interaction effects on aging. An aging study on bitumen–filler mixes showed that both bitumen and filler influence the binder hardening [26]. The short-term and long-term oven aging procedures developed under the SHRP-A-003A [5] have been the most widely adapted aging protocols for asphalt. The SHRP short-term oven ageing (STOA) procedure requires that prior to compaction, the loose bituminous mix should be aged in a forced draft oven at 135 °C for 4 h. The SHRP long-term oven aging (LTOA) procedure is a continuation of the STOA treatment, where the compacted samples are heated in the oven at 85 °C for further 120 h. When the results were compared with field aging (mixes with typically 5% air voids), it was found that the STOA method was roughly equivalent to 0–2 years in the field, and the LTOA method to 5–15 years, depending on climate. There are other accelerated long-term aging protocols for bituminous mixes including the use of higher temperatures (100 °C and above), high oxygen/air pressures, and feeding air/oxygen or humid air into the mix. However, all these methods use compacted samples and in order to produce RAP in the laboratory, these samples should be reheated to obtain a loose mix. A study was initiated at EMPA in order to verify the earlier findings on aging and investigate the aging behavior of loose bituminous mixes. This helps determine the aging parameters relevant for the RAP simulation in the laboratory. In this study, a dense asphalt mixture, AC 11 was produced in the laboratory and aged in a special aging box in the draft force oven at different temperatures, durations, oxygen concentrations, and humidity levels. The loose bituminous mixture was sampled at different time intervals and tested for its rheological properties on the recovered binder in order to follow the aging progression. The effect of sample thickness on aging was also investigated.
2. Methodology 2.1. Experimental test setup In this study, a loose asphalt sample was aged using a forced draft oven. Since the main objective was to compare different variables affecting aging over time, only long-term aging was assessed. A dense asphalt mixture, AC 11 prepared according to EN 13108-1 [27] with standard penetration grade bitumen 70/100 was used in testing. Binder content of the mix was 6.12% by mass. The aggregate gradation of the AC11 mix is given in Table 1. A batch of 150 kg of asphalt mixture was prepared in the lab at a mixing temperature of 160 °C. After production, it was partitioned and placed into boxes, each containing 6 kg of the mix, and stored at room temperature. Asphalt mix samples of 6 kg were pre-heated in an oven at 120 °C for 30 min then placed into a metal aging box with an approximately 3.5 cm thick even layer. The box had a volume of 0.01 m3 (10 28 35 cm) as shown in Fig. 1, and was
A. Ongel, M. Hugener / Construction and Building Materials 68 (2014) 49–54
51
Table 1 AC11 particle size distribution. Sieve (mm)
Percent passing
Specification requirement
0.063 0.125 0.25 0.5 1.0 2.0 2.8 4.0 5.6 8.0 11.2 16.0
7.3 9.6 13.6 19.3 26.2 34.8 43.0 56.2 70.3 87.9 98.7 100
6.4–10.4 – – – 20.4–28.4 29.9–41.9 – 48.8–62.8 – – 91.5–100 Fig. 3. Humid air generation setup with flow meter.
Opening for temperature sensor
Gas in/outlet
Fig. 1. Loose asphalt mix in the metal aging box.
sealed with silicon in order to make it airtight. The cover of the box consisted of three openings: two openings for the in- and outlet of air or nitrogen and a third one for two temperature probes that record the temperature at different points within the mix. Air and nitrogen coming from pressure cylinders were preheated by passing these gases through a copper coil prior to entering the aging box in order to prevent the cooling of the asphalt mix by the incoming gas. The incoming air temperature was verified at 100 ± 1 °C. The setup of the aging process in the forced draft oven is shown in Fig. 2. Gas flow into the aging box was set at a constant rate and monitored with an adjustable flow meter during the aging processes. Fig. 3 shows the adjustable flow meter with the humid air generator. The temperature of the asphalt sample was also recorded every minute during aging using a data
Fig. 2. Closed box aging setup.
logger and two temperature probes. Asphalt samples of 600 ± 50 g were taken every 24–48 h for testing. The initial sample, Day 0, corresponds to the mix before aging, but after 30 min of pre-heating at 120 °C. Bitumen was extracted from the asphalt by toluene using the cold extraction method. The asphalt sample was soaked in cold toluene for 15 min in order to dissolve the bitumen. Then the mixture was stirred by hand and the overlaying asphalt toluene solution without aggregates was poured through a 0.063 mm sieve into a glass container. The filler, which passed the 0.063 mm sieve, was separated from the toluene solution using a centrifuge according to EN 12697-1 [28]. Following the asphalt extraction, the bitumen was recovered from the toluene solution using a rotary evaporator at 145 °C and pressure of initially 45 kPa and then 1.9 kPa according to EN 12697-3 [29]. The extent of aging was assessed for the recovered binder through the Ring and Ball (R&B) softening point method [30] and the complex shear modulus (G*) test performed with a Physica MCR 301 dynamic shear rheometer (Anton Paar, Austria) according to EN 14770 [31]. An approximately 1 mm thick sample of bitumen was prepared in a silicon mould and placed between two 25 mm parallel plates. The gap was adjusted to 1.00 mm and the complex modulus was measured at a temperature of 30 °C and at the frequencies of 1.25 and 4.4 Hz. However, only the complex modulus values at 1.25 Hz are presented in this paper. Two replicate tests were conducted for each asphalt mixture both for softening point and complex shear modulus determination.
2.2. Aging methods In this study, four sets of aging tests were run in order to test the effects of temperature, oxygen content, humidity, and asphalt mixture layer thickness. The experimental test set up is given in Table 2. A reference test was carried out heating the asphalt mix in the closed aging box in the force-draft oven at a temperature of 100 °C and an air-flow of 0.45 l/min for 5 days. Set 1 evaluates the temperature effects on aging. In addition to the reference test conducted at the temperature of 100 °C, aging tests at 85 °C and 115 °C were performed, designated as Temp1 and Temp2 in Table 2, respectively. Since aging is faster at higher temperature, aging duration was shortened to 3 days when tested at 115 °C while extended to 8 days at 85 °C. All tests were conducted at an airflow of 0.45 l/min. Set 2 evaluates the effects of oxygen content on aging. In this set, aging tests were conducted both feeding air and nitrogen into the asphalt mix. In addition to the reference test conducted at an airflow of 0.45 l/min, aging with an airflow of 4.5 l/min and aging in the absence of oxygen, feeding 100% nitrogen at a flow rate of 0.45 l/min were performed, designated as Air1 and Air2 in Table 2, respectively. All the Set 2 tests were conducted in a closed box at a temperature of 100 °C. Set 3 evaluates the effects of humidity on aging. The influence of water vapor on the aging behavior was examined by feeding a saturated air stream into the aging box at 100 °C at a flow rate of 0.45 l/min, designated as Hum1 in Table 2. The water saturated air was produced by purging air through a fine sintered metal filter making fine air bubbles, which rapidly saturated the air with water. Due to the temperature dependency of relative humidity (r.H.), the saturated air at 25 °C corresponds to a relative humidity of approximately 4% at 100 °C. Humidity could also have been increased by purging air into hot water, however, then, the vapor content would be more difficult to control. In this set, in addition to the reference test, the standard aging procedure was also performed which is essentially aging a layer of asphalt mix in the draft forced oven, designated as Hum2 in Table 1. In the standard procedure, aging was performed in the aging box, but with the cover left open, without external air feeding, at a temperature of 100 °C. The relative humidity of air in the pressure cylinder is approximately 0.0065% whereas the relative humidity in a non-conditioned laboratory typically varies between 40% and 80% at an ambient temperature of 18–26 °C, resulting in a relative humidity of 1–3% at 100 °C. However, open box aging is not a controlled test in terms of airflow rate or air purity, therefore the differences in the aging progression cannot be fully attributed to humidity.
52
A. Ongel, M. Hugener / Construction and Building Materials 68 (2014) 49–54
Table 2 Experimental test setup. Test code
Temperature (°C)
Gas flow rate (l/min)
Gas type
Layer thickness (mm)
Aging duration (days)
Reference Temp1 Temp2 Air1 Air2 Hum1 Hum2 Layer Thickness
100 85 115 100 100 100 100 100
0.45 0.45 0.45 4.5 0.45 0.45 Open box Open box
Dry air Dry air Dry air Dry air Dry nitrogen Water saturated air with r.H. of 100% at 25 °C Ambient air with r.H. of 40–80% at 25 °C Ambient air
35 35 35 35 35 35 35 80
5 8 3 5 5 5 5 5
Set 4 evaluates the effects of layer thickness on aging. An 8-cm-thick asphalt mix was aged at 100 °C in an open container, designated as layer in Table 2. In this set, sampling was conducted only at the end of the testing period, after 5 days of aging. Aging was evaluated for the top 4 cm and bottom 4 cm of the mix separately.
3. Results and discussion Data analysis showed that the softening point and complex shear modulus, G* before aging were not always the same for different batches containing 6 kg of virgin bituminous mixture. This can mainly be attributed to the storage time of the hot mix asphalt samples after production. Since bitumen testing and extraction steps are labor and time intensive, the laboratory tests had to be carried out successively and therefore storage time was not the same for all asphalt mix samples. Therefore asphalt mixtures stored longer may have aged more and got stiffer than those stored for a shorter time. Additionally, even though special effort has been given in order to follow the same procedure for each test, the handling differences during the preparation and testing of the binder, including the heating of the hot mix sample, and the extraction and rheological measurements of the binder, may have caused some measurement error. Therefore, for comparison of the aging progression, data have been normalized by dividing the softening point and complex shear modulus G* values of the aged binder by the values before aging, designated as Day 0. The ratio of aged binder value to that of unaged binder was defined as aging index. Aging index in terms of complex shear modulus was calculated using the logarithm of complex modulus G* values, log Gaged = log Gunaged , as the softening point is linearly proportional to the logarithm of the viscosity at a specific temperature. The repeatability of the R&B test according to EN 1427 [30] is given as ±1 °C for non-modified binders. DSR test allows 10% deviation from the mean value when repetitive samples are tested [31]. Therefore, aging index for softening point was given with error bars showing 1 °C deviation from the mean temperature, while aging index for complex modulus was given with error bars showing 10% deviation from the mean complex modulus value. A hyperbolic trend line was fitted to the data. Figs. 4 and 5 show the aging index for Set 1 tests for logarithm of complex modulus G* at 1.25 Hz and softening point, respectively. Asphalt aging was conducted at 85 °C, 100 °C, and 115 °C at a constant airflow of 0.45 l/min. According to Fig. 4, logarithm of complex shear modulus G* of the bitumen reached approximately 1.05, 1.07, and 1.06 times that of the unaged bitumen at the end of testing period when heated at temperatures of 85 °C, 100 °C, and 115 °C, respectively. According to Fig. 5, the softening point of the bitumen reached approximately 1.1, 1.15, and 1.2 times that of the unaged bitumen at the end of testing period for aging temperatures of 85 °C, 100 °C, and 115 °C, respectively. Higher aging indices in terms of softening point may be due to softening point test being more sensitive to temperature changes compared to shear modulus test. It can be seen from Fig. 4 that aging process is not complete yet at the end of the testing period as the complex shear modulus G* is still increasing. However, according to Fig. 5, age hardening slows
Fig. 4. Aging index for logarithm of complex modulus G* at 30 °C at 1.25 Hz for bitumen aged at different temperatures.
Fig. 5. Aging index for softening point for bitumen aged at different temperatures.
down for temperatures of 85 °C and 100 °C, and still goes on for 115 °C. This may indicate that aging rate in terms of complex shear modulus G* is slower than that of softening point. Hence, the aging indices in terms of shear modulus may reach to those of softening point when longer aging durations are applied. In general, the rate of aging goes down after the third day of the testing period. Since the age hardening is still in progress, even though at a slower rate, at the end of the testing period, it can be concluded that the aging durations chosen were too short in order to obtain the final aging index value, where there is no further age hardening. The rate of increase in age hardening goes up with increasing temperature both for complex modulus G* and softening point. An approximately doubling of oxidation rate was expected with every 10 °C increase in temperature according to the Arrhenius Law. Therefore, aging rate at 100 °C and 115 °C was expected to be three times and eight times, respectively of that at 85 °C. However, it can be seen from Figs. 4 and 5 that the aging rate increases much less than expected which may be due to asphalt-aggregate interaction slowing down the aging process as suggested by Recasens et al. [24]. It can also be seen from the figures that the aging value at 85 °C will never reach the aging indices at 100 °C or at 115 °C regardless of aging time. This indicates that different chemical reactions are responsible for the age hardening of bitumen at different temperatures as suggested by Herrington et al. [15]. Chemical reactions require specific activation energy to start. At 85 °C the activation
53
A. Ongel, M. Hugener / Construction and Building Materials 68 (2014) 49–54
energy may be too low for some chemical reactions taking part in the oxidation of bitumen, which is a complex series of chemical reactions, not only including the incorporation of oxygen but also oxygen induced or catalyzed reactions [32]. This is why the oxygen content as determined by elemental chemical analysis does not increase significantly during oxidation. Standard bitumen contain only around 0.5% (mass percentage) of oxygen and even with severe air blowing conditions using pure oxygen, only an increase to 1% has been shown [33,34]. Figs. 6 and 7 show the aging index for Set 2 and 3 tests for logarithm of complex shear modulus, G* and softening point, respectively. Asphalt samples were aged in an open box; and in a closed box feeding air with a flow rate of 0.45 l/min and 4.5 l/ min, feeding air at a flow rate of 0.45 l/min with relative humidity of 4% and feeding nitrogen at a flow rate of 0.45 l/min. All the aging tests have been carried out at 100 °C. Instead of a hyperbolic trend line, data points were connected with a straight line for the test in the open box, since aging process stopped/slowed down after 2 days. It can be seen from Figs. 6 and 7 that age hardening stopped or slowed down to a constant rate at the end of the testing period. When pure nitrogen was fed through the asphalt sample, the aging index fluctuated around 1, showing no age hardening. It can be concluded that nitrogen inhibits the aging efficiently even at 100 °C, indicating the necessity of oxygen to be present for age hardening. There may be chemical reactions taking place in the presence of nitrogen, however they do not affect the viscosity of the binder significantly. Besides, this experiment shows that there is no leakage of oxygen into the aging box. When the airflow was increased from 0.45 l/min to 4.5 l/min, a slightly higher aging index compared to the reference procedure was obtained. When humid air was used instead of dry air, there was a reduction in the aging index both for complex modulus and softening point. Since the water was in vapor state and relative
Table 3 Complex modulus and softening point for the top and bottom samples. Complex modulus G*, Pa Top Bottom Ratio: Top/bottom
6
1.27 ⁄ 10 1.21 ⁄ 106 1.05
Softening point, °C 63.3 63.5 0.99
humidity was only around 4% at 100 °C, it was not possible that the binder was protected by a water film against oxygen molecules as suggested by Negulescu et. al. and Daranga [16,17]. The reduction in aging index may be due to the water molecules scavenging reactive compounds which have been produced by the reaction with oxygen or even oxygen radicals. Aging rate was the highest when asphalt was aged in the open box, which represents the traditional oven aging. After the second day of aging process, there was no significant increase in age hardening in terms of complex modulus or softening point. However, after 6 days of aging at 100 °C, the aging index derived from the softening point reached nearly the same value as when aged in the closed box at an airflow of 4.5 l/min. The faster rate of age hardening in the open box can be explained with the presence of reactive trace substances. Since the air in the lab is not purified like the air coming from the pressure cylinder, trace substances like ozone, nitro- and sulfur oxides could be present which may affect the aging behavior of bitumen. This indicates that oxygen concentration in terms of both airflow rate and air purity control the aging rate. But further research is necessary to find the relevant trace substances which are responsible for the accelerated aging. In the Set 4 testing, the complex modulus G* at 30 °C at 1.25 Hz and softening point of the top 4 cm and bottom 4 cm of the 8-cmthick asphalt mix were evaluated as shown in Table 3. The complex modulus G* and softening point of the top and bottom part are not significantly different from each other, within the standard error of the tests. Therefore, it can be concluded that there is no significant difference in aging between the top and bottom sections. This does not confirm the earlier findings that aging goes down with increasing depth. However, there may be homogenous temperature distribution and exposure to oxygen throughout the loose mix, different than that of the pavements in the field. It could be as well due to the fact that aging takes place only in a very thin layer of the surface mix with negligible impact on the whole layer as suggested by Coons and Wright [23]. 4. Summary and conclusions
Fig. 6. Aging index for logarithm of complex modulus G* at 30 °C at 1.25 Hz for bitumen aged at 100 °C.
Fig. 7. Aging index for softening point for bitumen aged at 100 °C.
In order to verify the earlier findings and simulate RAP in the laboratory, the effect of temperature, aging duration, humidity, oxygen concentration, and layer thickness on the age hardening of loose bituminous mixes was evaluated. Aging methods used in this study included: heating the loose asphalt at 85 °C, 100 °C, and 115 °C for 3 days, 5 days, and 8 days, respectively, feeding air at a flow rate of 0.45 l/min; heating the loose asphalt in an open box, and in a closed box feeding air at a flow rate of 4.5 l/min, feeding humid air at a flow rate of 0.45 l/min, and feeding nitrogen at a flow rate of 0.45 l/min at 100 °C for 5 days; and heating 8-cm-thick asphalt mix in an open pan at 100 °C for 5 days. There was no aging observed when nitrogen instead of air was fed through the sample, indicating aging of the bituminous mix was solely controlled by oxidation. Aging index was found to increase with increasing airflow rate and hence oxygen concentration at a given temperature. Aging rate also increases with increasing temperature. Doubling in oxidation rate was expected with 10 °C increase in temperature according to Arrhenius Law. However, aging rate increase
54
A. Ongel, M. Hugener / Construction and Building Materials 68 (2014) 49–54
in terms of both complex modulus and softening point was shown to be slower which may be due to the smoothing effect of aggregate–bitumen interaction. Aging usually slows down after the third day of testing. Hyperbolic-like curves, an initial rapid increase in aging indices followed by a slower rate, were observed for tests conducted at 85 °C and 100 °C. Age hardening still goes on at a rapid rate after 3 days for testing at 115 °C. Therefore, aging tests should be conducted for longer durations in order to observe the final aging index value. Final aging indices are different at the temperatures 85, 100 and 115 °C. This leads to the conclusion that the chemical reactions change with temperature as indicated by earlier findings. Humidity was found to slow down the aging process. In contrast to earlier findings, no difference in age hardening was observed for different sample depths. Hardening was the same for the top and bottom layer of the bituminous mixture. R&B softening point test was found to be more sensitive to temperature changes compared to DSR measurements since the rate of change in aging index was slower for the dynamic complex modulus. There are indications that certain trace substances in the air of the laboratory promote the aging of the asphalt mix. Further experiments are necessary to identify the responsible trace substances.
Based on the findings, the testing protocol for RAP simulation should be conducted in a closed box controlling the oxygen content, since trace substances affect the aging of the bitumen. Bituminous mixture should be used in the aging process rather than the bitumen, due to the fact that aggregates may affect the aging rate of the bitumen. Aging temperature should be selected based on the rheological properties of the bitumen extracted from the recycled pavement since the limit for age hardening varies with aging temperature. Age hardening was not significantly different for the top and bottom layer of the bituminous mix, therefore there is no need to stir the loose mix during the aging process for RAP simulation. In the future, the results should also be correlated with the values obtained from RAP from the field.
References [1] De la Roche C, Van de Ven M, Van den Bergh W, Gabet T, Dubois V, Grenfell J, et al. Development of a laboratory bituminous mixture ageing protocol. Advanced testing and characterization of bituminous materials. Taylor and Francis Group; 2009. [2] Peterson JC. Chemical composition of asphalt as related to asphalt durabilitystate of the art. Transp Res Rec 1984;999:13–30. [3] Zoorop SE, Collop AC, Brown SF. Performance of bituminous and hydraulic materials in pavements. In: Proceedings of the fourth European symposium on performance of bituminous and hydraulic materials in pavements. Bitmat 4, Nottingham, United Kingdom; 2002. [4] Petersen JC, Harnsberger PM, Robertson RE. Factors affecting the kinetics and mechanisms of asphalt oxidation and the relative effects of oxidation products on age hardening. Prepr Pap Am Chem Soc Div Fuel Chem 1996;41(4):1232–44. [5] Bell CA, Sosnovske D. Validation of the A-002A hypothesis for ageing. TM-OSUA- 003A-92-24, final summary report to SHRP: 92-2, Strategic Highway Research Program, National Research Council, Washington, D.C.; 1992.
[6] Von Quintas H, Scherocman J, Kennedy T, Hughes CS. Asphalt aggregate mixture analysis system. NCHRP report No: 338, National Highway Research Program, Transportation Research Board, National Research Council, Washington, D.C.; 1988. [7] Bell CA, Sosnovske D. Aging: binder validation. SHRP-A-384, Strategic Highway Research Program, National Research Council, Washington, D.C.; 1994. [8] Huang SC, Robertson RE, Branthaver JF, McKay JF. Study of steric hardening effect of thin asphalt films in presence of aggregate surface. Transp Res Rec 1999;1661:15–21. [9] Petersen, JC. Chemical composition of asphalt as related to asphalt durability in developments. Petroleum Science, 40: Part B: 363–395, Asphaltenes and Asphalts, 2; 2000. [10] Griffin RL, Miles TK, Penther CJ. Microfilm durability test for asphalt. Proc Assoc Asphalt Paving Techonol 1955;24:31–62. [11] Van Oort WP. Durability of asphalt: its aging in dark. Ind Eng Chem 1956;48(7):1196–201. [12] Negulescu I, Mohammad L, Daly W, Abadie C, Cueto R, Daranga C, et al. Chemical and rheological characterization of wet and dry aging of SBS copolymer modified asphalt cements: laboratory and field evaluation. J Assoc Asphalt Paving Technol 2006;75:267–96. [13] Petersen JC. A Review of the Fundamentals of asphalt oxidation: chemical, physicochemical, physical property, and durability relationships. Transportation Research Circular E-C140; 2009. [14] Gawel I, Baginska K. Effect of the chemical nature on susceptibility of asphalt to aging. Pet Sci Technol 2004;22:1261–71. [15] Herrington PR, Patrick JE, Ball GFA. Oxidation of roading asphalts. Ind Eng Chem Res 1994;33:2801–9. [16] Negulescu I, Mohammad L, Daly W, Chiparus I. The use of DMA to characterize the aging of asphalt binders. FHWA/LA.07/461; 2010. [17] Daranga C. Characterization of aged polymer modified asphalt cements for recycling purposes. Dissertation, Graduate Faculty of the Louisiana State University and Agricultural and Mechanical College; 2005. [18] Hagos ET, Molenaar AAA, Van de Ven MFC. Chemical characterization of laboratory and field bitumen aging in porous asphalt concrete. In: Advanced testing and characterization of bituminous materials, two volume set. CRC Press; 2009. [19] Campbell PG, Wright JR. The effect of temperature and humidity on the oxidation of airblown asphalts. Am Chem Soc, Div Pet Chem Prepr 7 1962;3:123. [20] Thomas KP. Impact of water during the laboratory aging of asphalt. Road Mater Pavement Des 2002;3(3):299–315. [21] Al-Khateeb G, Qi X, Shenoy A, Stuart K, Mitchell T. Assessment of aging at FHWA’s pavement testing facility. Transp Res Rec 2005;1940:146–55. [22] Li NL, Zhao XP, Du C, Zhang CL, Xiao QY. Effect of aging on urban asphalt pavement. Adv Mater Res 2013;1287:671–4. [23] Coons RF, Wright PH. An investigation of the hardening of asphalt recovered from pavements of various ages. Proc Assoc Asphalt Paving Technol 1967;37:510–28. [24] Recasens RM, Martinez A, Jimenez FP, Bianchetto H. Effect of filler on the aging potential of asphalt mixtures. Transp Res Rec 2005;1901:10–7. [25] Petersen JC, Barbour FA, Dorrence SM. Catalysis of asphalt oxidation by mineral aggregate surface and asphalt components. Proc Assoc Asphalt Paving Technol 1974;43:162–77. [26] Gubler R. Einfluss schweizerischer filler auf die alterung von bituminösen bindemitteln und die rissbildung im belag. Eidgenössisches Departement für Umwelt, Verkehr, Energie und Kommunikation – Bundesamt für Strassen 2005; 1138. [27] EN 13108-1:2006 Bituminous mixtures – material specifications – Part 1: asphalt concrete. [28] EN 12697-1:2005. Bituminous mixtures – test methods for hot mix asphalt – Part 1: soluble binder content. [29] EN 12697-3:2005. Bituminous mixtures – test methods for hot mix asphalt – bitumen recovery. Rotary evaporator. [30] EN 1427:2007. Bitumen and bituminous binders – determination of the softening point. Ring and Ball method. [31] EN 14770:2005. Bitumen and bituminous binders – determination of complex shear modulus and phase angle – Dynamic Shear Rheometer (DSR). [32] Moschopedis S, Speight J. The effect of air blowing on the properties and constitution of a natural bitumen. J Mater Sci 1977;12(5):990–8. [33] Read J, Whiteoak D. The Shell bitumen handbook. London: Thomas Telford Publishing; 2003. [34] Campbell PG, Wright JR. Ozonation of asphalt flux. Ind Eng Chem Prod Res Dev 1964;3(3):186–94.