Influence of processing conditions on rheology of tyre rubber modified bitumens

Transcription

1 Influence of processing conditions on rheology of tyre rubber modified bitumens D. Lo Presti* - N.Memon* - G.Airey* * Nottingham Transportation Engineering Centre University of Nottingham, Nottingham, UK davide.lopresti@nottingham.ac.uk naeem.memon@nottingham.ac.uk gordon.airey@nottingham.ac.uk ABSTRACT. Tyre Rubber Modified Binders (TR -MBs), produced through McDonald wet process and used worldwide (e.g. asphalt rubber), have been demonstrated to provide various benefits to pavements and, moreover, they represent a good opportunity for recycling tyre rubber. However this technology is still struggling to be fully adopted in Europe, mainly because of their poor stability during high temperatures storage, which leads to high initial costs in modifying existing asphalt plants. NO-agitation TR-MBs (also known a s terminal blends) is proving to be a great option and its development could also be the key to spreading the recycling of tyre rubber in paving applications in Europe. This paper aims to enrich this field of research by presenting the results of a study focused on the optimisation of laboratory procedures to better understand the effect of varying processing conditions on the rheology TR-MBs. The experimental programme has been carried out by a preliminary selection of materials, followed by the production of the TR-MBs, via practical laboratory protocols. A continuous comparison between two commercially used SBS-MBs, with high and medium levels of modification, and the produced TR-MBs, helped to understand the effect of varying the selected processing conditions on binder properties. Results have shown that TR-MBs are a very effective alternative to commercially used SBS-MB. Although the rheology of TR- MBs is very sensitive to the varying of processing conditions, therefore an appropriate selection of materials and a superior binder design are mandatory to optimise the modification process to achieve the desired level of modification. KEYWORDS: Tyre rubber, modified bitumen, processing conditions, DMA

2 2 Lo Presti et.al - AR Introduction The method of modifying bitumen with Tyre Rubber (TR) is generally associable with the term wet process. The wet process requires mixing of the tyre rubber (TR) in hot bitumen (176ºC to 226ºC) and holding the resulting blend at elevated temperatures (150ºC to 218ºC) for a designated pe riod of time (typically 45 to 60 minutes, shorter for some variations) to permit an interaction between rubber and bitumen. The percentage of the rubber used in the blend is usually in the range of 18-22% bitumen weight [1]. As mentioned before, the wet process technology existed since the 1960s, but the nature of the mechanism by which the interaction between bitumen and TR took place has not been fully characterised. Of course the selected materials play an important role on the property of the final blend, but the nature of the reaction itself is strongly affected by the following processing conditions: - Processing Temperature - Processing Time - Processing Device (applied shear stress and pressure) 1.1 Processing TR-MB Traditionally, it is believed that the reaction occurring between tyre rubber and bitumen is made up of two simultaneous processes: partial digestion of the rubber into the bitumen on one hand and, on the other, adsorption of the aromatic oils of the bitumen, within the polymeric chains of the TR. The absorption of aromatic oils causes the rubber to swell and soften [2,3]. TR particles are swollen by the absorption of the bitumen oily phase at high temperatures ( ºC) into the polymer chains, which are the key components of the TR-MB to form a gel-like material. Thus, it is appropriate to use as base bitumen, bitumens that are particularly rich in aromatics or, if necessary, to add aromatic oils in order to create swelling of the rubber granulate. Tyre rubber reacts in a time-temperature dependent manner and during this process there is a contemporaneous reduction in the oily fraction of the bitumen and an increase of the TR particle sizes with a consequent reduction of the inter-particle distance. This implies the formation of gel structures that produce a viscosity increase up to a factor of 10 [4,5 and 6]. Final viscosity of it is also affected by the applied shear. In fact, tyre rubber particles tend to agglomerate and this can contribute to another increase of stiffness/viscosity during the modification process [7]. By using high shear mixing, it is possible to provide the energy to break particle-particle bonds and this helps in reducing agglomeration [7]. Once the polymer wets the particle, the bond fails to re-form, leading to better dispersion of the filler in the polymer matrix [8]. Therefore, high shear rate is fundamental to ensure proper dispersion of the rubber within the bitumen matrix and it has been demonstrated that this helps the storage stability of the final blend [8]. However, the

3 Lo Presti et.al - AR processing device is secondary to the processing temperature. In fact, if the tyre rubber-bitumen blend is subjected to long exposure to high temperatures, the reaction process changes completely: the swelling process is replaced by a depolymerisation/devulcanisation of the tyre rubber with its dispersion into the bitumen. This happens independently from the used processing device (applied shear and pressure). Depolymerisation starts releasing rubber components back to the liquid phase causing a destruction of the binder networking [9]. From the mechanical point of view, this causes a decrease in the stiffness (G*), and so in viscosity, while the elastic properties ( ) are less affected. Then, it is evident, how depending on the chosen processing conditions the wet process can provide different products and improving one aspect (e.g. storage stability) could compromise another (e.g. rheological properties). 1.2 Aim of the study As seen above, TR-MBs are extremely dependent on the selected processing conditions. In particular, the processing temperature is the factor that most affects the whole modification process; only high processing temperature, indeed, could lead to TR-MBs which do not require agitation. Based on these findings, the authors have developed from their own experience with TR-MB manufacturing. This study presents optimised laboratory procedures to monitor the effect of varying processing conditions on the rheology of TR-MBs in order to perform a superior binder design. 2. Experimental Programme Two TR-MBs were produced by using the same materials and blending protocols but changing the processing temperature from 180 C and 210 C. Low shear protocol was firstly used to perform a design of the blends with the selected materials at both processing temperatures. The design consisted of the production of several blends with different tyre rubber content supported by a detailed rheological characterisation by means of DRS tests, to assess the optimum proportion between the tyre rubber and the bitumen. Secondly, the high shear protocol was used to increase dispersion of the rubber within the bitumen and so optimising processing time and settling properties of the final product. Binders produced in high shear, as well as samples taken during the modification process, were characterised and compared through empirical tests such as: penetration (EN 1426), softening point (EN 1427), rotational viscosity at 135 ºC and 160ºC (EN 13302) and the apparent viscosity at C as prescribed by the Asphalt Rubber standard (ASTM D6114). Furthermore a detailed rheological characterisation has been performed carrying a Dynamic Mechanical Analysis (DMA). At last, also a comparison with two Styrene- Butadiene-Styrene modified bitumens (SBS-MBs with a medium and a high level of modification) was performed in order to understand the achieved level of modification.

4 4 Lo Presti et.al - AR Materials A 50/70 pen bitumen and a 30# ambient car and trucks tyre rubber, processed at ambient temperature and free of fibres and steel, were selected as base materials for the production of two TR-MBs. The base bitumen was chosen between six available bitumens provided from various Italian plants and identified as bases for modification. In order to choose the best base, all the bitumens were chemically characterised through a SARA analysis, based on the method B of the ASTM D4124 standard. According to what has been found in literature about the production of PMBs [11], higher aromatics content and lower colloidal index/higher dispersion factor enhances the compatibility between the bitumen and its modifier. Based on these findings, the bitumen shown in the Table 1 was used for the production of the TR-MBs. Penetration 25 C (mm/10) EN dmm Softening Point ( C) EN ºC Penetration Index Rotational 100ºC (ASTM T316-04) Pa.s Rotational 135ºC (ASTM T316-04) 0.429Pa.s Rotational 160ºC (ASTM T316-04) Pa.s Saturates [%] 11.9 Aromatics [%] 61.4 Resins [%] 17.1 Asphaltenes [%] 9.6 Table 1. Physical and chemical properties of the selected base bitumen Summarising, the following binders have been incorporated into this study: neat bitumen: TR-MB HS180: TR-MB HS210: pen 50/70 bitumen Tyre Rubber-Modified Bitumen (TR -MB) produced mixing in High Shear (HS) base bitumen with 18% in weight of 30# TR at 180ºC Tyre Rubber-Modified Bitumen (TR -MB) produced mixing in High Shear (HS) base bitumen with 18% in weight of 30# TR at 210ºC SBS-MB H: SBS-MB with a high content of SBS (Hard - H) SBS-MB M: SBS-MB with a medium content of SBS (Medium - M). 2.2 Processing devices and procedures Apparent viscosity is the physical parameter that is commonly used to check the improved properties of a TR-MB during the modification process. In fact, the rubber-bitumen interaction, at a certain temperature, implies an increase in the blend viscosity up to a maximum value and a subsequent noticeable decrease with the

5 Lo Presti et.al - AR reaction time [12,13]. Furthermore, performing a design of a TR-MB could be very much time-consuming and it could lead to producing a very high number of blends which results in long procedures and a very high consumption of material. For this reason the design of the binders has been done through the low shear protocol which reduces time and material consumption. Low shear protocol The low shear protocol, has been preliminary used to perform a binder design aiming to assess two pieces of information: first, the minimum rubber content to be added to the selected bitumen to get relevant improvements in terms of stiffness and elasticity; secondly to establish the optimum content of tyre rubber for the selected bitumen in terms of properties improvements ratio. This preliminary study has been performed by adapting the Brookfield viscometer as a mixer. Rotational viscometer, indeed, offers the possibility of mixing bitumen with fine polymers at a low shear rate (max 250 rpm), with an accurate control of the temperature and allows for constantly taking apparent viscosity measurements. This procedure helped to perform a monitoring of the modification process based on the trend of the viscosities values (Fig.2). Although, considering the high heterogeneity of the material and the possible effects on the registered viscosity values due to the small gap between the spindle and the walls of the container could, the optimum binder design was based on a rheological characterisation of the blend with further Dynamic Mechanical Analysis (DMA). DMA has been chosen also because Brookfield Viscometer allows producing grams of material which are not enough to perform conventional tests like penetration, but it is sufficient to perform tests with a DSR. High shear protocol Previous studies [14] have highlighted that high shear stress increases the dispersion of TR within the bitumen and optimises the blending time and the storage stability. Based on this information the binders designed in low shear have been subsequently produced in high shear. High shear mixing has been performed by using a Silverson L4 mixer equipped with a duplex head to reduce and disperse TR within the bitumen matrix. A constant high shear, rpm, has been applied for two hours, The proportions between the bitumen and the tyre rubber (TR), determined with the preliminary study in low shear, were kept constant while two different blending temperatures were used: 180 C and 210 C. The laboratory mill and a thermostatic bath, used as a processing vessel, allowed to perform the blending with a maximum temperature variability of ± 1ºC within the whole blend. Nevertheless, considering the capacity of the processing vessel (4 litres) high amounts of materials were needed (~ 3 Kg of

6 6 Lo Presti et.al - AR2012 bitumen). Although, such big amount of material allows for performing several sampling during the mixing. Samples were further analysed by performing, viscosity measurements and DMA. 2.3 Rheological analysis Optimisation of the TR-MBs production and properties comparisons were mainly based on the evaluation of the rheological properties. Therefore, in order to obtain a complete rheological characterisation and since it has been proved that SUPERPAVE DSR protocol can also be applied to rubber modified binders [15], dynamic mechanical analyses (DMA) have been performed. DMA have been carried out by making frequency sweep tests over a wide range of temperatures with an Antoon Paar Physica MCR 101 dynamic shear rheometer (DSR). The tests were performed under the following conditions: - Loading Mode: controlled-strain - Temperatures: 0ºC to 80 ºC with 5ºC intervals (neat bitumen) 30ºC to 80 ºC with 5 or 10 ºC intervals (TR-MB) - Frequencies: 0.10, 0.16, 0.25, 0.40, 0.63, 1, 1.6, 2.5, 4, 6.3 and 10 Hz - Plate geometries: 8 mm and 2 mm gap (0-50ºC) (only neat bitumen) 25 mm and 1 (only neat bitumen) to 2 mm gap (30-80ºC) - Strain amplitude: 0.5% with 8 mm (within LVE resp. dependent on G*) 0.5% and 2%-12%@80 C with 25 mm plates (within LVE response dependent on G*) For each test, samples were prepared by means of a hot pour method, based on Alternative 1 of the AASHTO TP5 Standard (AASHTO, 1998). The gap between the upper and lower plates of the DSR was chosen such that the rheological properties taken at wider gap widths (1.5-2 mm for 25mm) were independent of the gap and that sample geometry was maintained also at high temperatures. The rheological properties of the binders were measured in terms of their complex (shear) modulus, G*; and phase angle,. Once the raw data were obtained, Time Temperature Superposition Principle was applied in order to produce master curves at 30 C, isochronal plots and shift factor curves. Black diagrams, Isochronal plots, master curves and binder rutting parameters (SHRP and Shenoy s parameters ) were used as the basis of all the rheological analyses in this paper.

7 Lo Presti et.al - AR Results and discussions 3.1 TR-MB design in low shear As a matter of fact, adding tyre rubber to bitumen greatly increases the high temperature end of the paving grade, leaving the low temperature slightly better. More specifically, a rule of thumb says that for every 1% of added polymer, 2 C in high temperature PG are typically gained. On the low temperature end, the rule becomes for every 1% of added polymer, there is a 1 C reduction in low temperature PG. Since the PG classes are based on 6 C steps, the typical 3% polymer-content modification generally allows gaining one high-temperature class leaving the low temperature sometimes unchanged [16]. A total of 32 blends were made to investigate a range of rubber percentages from 3% to 24%, with 3% steps by weight of bitumen, for both mixing temperatures: 180 C and 210 C. Table 2 summarises the low shear blending procedure, while table 3 and figure 2 show the monitoring of the properties variation during the mixing of the bitumen and tyre rubber. bitumen rubber mass rubber total weight mixing mixing mixing mass (3-24% of size (each blend) time speed temp. ( each blend) bitumen) (30 #) g g mm g m rpm C optimised Table 2. Low shear blending protocol Monitoring the modification process As a result apparent viscosity increases with time until it reaches a plateau which indicates the maximum level of modification (peak viscosity). At this point the mixing has been stopped. This allows a first important result to be obtained represented by the maximum blending time, for each percentage of rubber, in order to get the best properties at low shear (Fig.1). Another important result, shown in Table 4, indicates that increasing the processing temperature to 210 C gives a significant reduction of both peak viscosity and optimum blending time (almost half time). Both the effects are probably a consequence of the partial depolymerisation of the vulcanised tyre rubber that implies a loss in mechanical properties. Furthermore, the blending procedure of the low shear protocol helped to verify that the TR-MB LS210 resulted in a less heterogeneous mix than TR-MB LS180. Indeed, during the blending, the viscosity of the binder was similar in both the top and bottom sections of the sample.

8 8 Lo Presti et.al - AR C 210 C Viscosity Optimum Time Viscosity Optimum Time % TR mpa = cp min mpa = cp min Table 4. Monitoring of the Low Shear protocol at 180 C and 210 C Figure 1. Example of monitoring the modification process of TR-MBs with low shear protocol Estimating the optimum tyre rubber content After the mixing, binders have been tested through a Dynamic Mechanical Analysis (DMA) to assess the optimum proportions between the TR and the bitumen at both temperature of 180 C and 210 C. Black diagrams (Figs. 2, 3) and isochronal plots at very high temperatures (Figs. 4,5) have been used as the base of this estimation. In fact, after experiencing several mixing processes and as a result of previous comparisons with other commercially-used modified bitumen, the authors have chosen the following conditions as parameters to assess the achievement of an acceptable level of modification and also to establish the optimum TR content:

9 Lo Presti et.al - AR ) Phase angle constantly lower then 70. 2) Isochronal plots of phase angle achieve a plateau or a downward trend at low frequencies (high temperatures). 3) Slight variation of the Black diagrams with increasing the tyre rubber content. Figure 2. Black diagrams of low shear TR-MBs LS 180 and neat bitumen Figure 3. Black diagrams of low shear TR-MBs LS 210 and neat bitumen

10 10 Lo Presti et.al - AR2012 Figure 4. Isochronal plots of G*(a) and (b) of low shear TR-MBs LS 180 and neat bitumen Figure 5. Isochronal plots of G*(a) and (b) of low shear TR-MBs LS 210 and neat Results of the comparison with the neat bitumen, show that at low frequencies, which corresponds to high temperatures, modification with tyre rubber makes the binder stiffer (i ncrease of the complex modulus) (Figs. 4a, 5a) and infers a better elastic behaviour (decrease in the phase angle) (Figs. 4b, 5b). This phenomenon is a clear effect of the formation of a polymer network, in which the elastic characteristics of the modifier agent (rubber) prevail over the typical viscous behaviour of the bitumen at high temperatures. This trend has a direct correlation with the rubber percentage: by increasing the concentration of rubber the binder became stiffer and more elastic (Fig. 2,3). Furthermore, isochronal plots show that the rubber provides a clear improvements of the thermal susceptibility and of the elastic properties of both TR-MBs (Figs. 4,5). Comparing the processed TR-MBs, it has to be highlighted that the TR-MB LS210 shows smoother curves with a more stable trend. This phenomenon confirms the achievement of a more homogeneus blend, due to the higher processing

11 Lo Presti et.al - AR temperature, but highlights also that higher depolymerisation rate of rubber does not prevent to have a tyre rubber-bitumen blend that still show good rheological properties. All the diagrams highlight one fundamental datum which confirm that even at 210ºC, 15% rubber content seems to be the minimum to obtain a TR-MB with an acceptable level of modification. Furthermore, from the rheological analyses it appears that it is not possible to gain any further significant improvement in terms of stiffness, elastic behaviour, thermal susceptibility and rutting resistance when the rubber content is higher than 18% - 21%. For all these reasons and to help the storage stability, 18% of rubber was fixed as rubber content to be used for optimising the TR-MBs by producing the high shear blends. 3.2 TR-MBs optimisation in high shear Based on the preliminary design in low shear the TR-MBs were blended by using the high shear blending protocol (Table 3). Every twenty minutes 100 g of materials were sampled and materials were further tested to assess the changes during the modification process. Apparent viscosity at 177.5ºC, and rheological parameters were monitored in order to find out the optimum processing time at both temperatures. Figure 6 shows that the optimum conditions provided by the viscosity peaks [13], were achieved between mins at 180 C and between mins at 210 C. As a first result, the high shear mixing has decreased the processing time to about the half of the time necessary to achieve peaks in low shear (Table 4). mass of Mass of TR rubber total mixing mixing mixing time bitumen (18%) size weight speed temp. g g mm g min rpm C or 210 Table 3. High shear blending protocol Figure 6. Apparent 177.5ºC of the TR-MBs at different processing time

12 12 Lo Presti et.al - AR2012 Achieved level of modification in high shear In order to assess the achieved level of modification of the processed binders at both temperatures, the TR-MBs were compared with the neat bitumen and the two SBS polymer modified bitumens. Comparisons were made in terms of conventional and rheological properties. From the conventional tests (Tab. 5), it can be concluded that both TR-MBs are harder than their base bitumen (neat) and have higher SP and Penetration Index. Furthermore, both TR-MBs have the characteristics (only in terms of Pen, SP and Apparent Viscosity) to be classified as Asphalt Rubber type I or II (ASTM D 6114, 2002). Nevertheless, considering PMBs as the target, the TR-MB HS180 shows the best improvements, but it is stiffer even than the SBS-MB Hard and too viscous for the requirements of the ASTM Standard specifications for modified bitumens (ASTM D5892, 2000). According to this standard, indeed the viscosity at 135ºC can be a maximum of 3000 Pa.s. TR-MB HS 210 has got a much lower viscosity but it is still to be considered a high-viscosity TR-MB. It is also stiffer than PMBs, but has poor improvement in terms of its softening point. Furthermore, both TR-MBs cannot be classified as PMBs within the European Framework also due to their combination of Penetration and Softening points (PI), Specification (EN 14023, 2005). Conventional properties neat TR-MB HS180 TR-MB HS210 SBS-MB M SBS-MB H Penetration 25 C (mm/10) (EN 1426) Softening Point ( C) (EN 1427) Penetration Index Apparent Viscosity at 177 C (Pa s) (ASTM D6114) Viscosity at 135 C, (Pa s) (EN 13302) Viscosity at 160 C (Pa s) (EN 13302) Table 5. Results of the empirical tests and viscosity measurements Supplementary information have been assessed by performing a detatiled DMA. An overall analysis of the results, here presented in form of Black diagrams and rutting parameters curves (Figs. 7,8), demonstrates that TR-MB HS180 is stiffer and more elastic at high service temperatures and it has got a rheological behavior almost comparable with the SBS-MB H. The TR-MB HS210 is softer and less elastic than the TR-MB HS180 and thus, it has rheological properties comparable only with the PMB with a medium level of modification (Fig.7).

13 Lo Presti et.al - AR Figure 7. Black diagrams of neat bitumen, PMBs and TR-MBs HS produced at 180ºC and 210ºC Figure 8. Performance comparison of TR-MBs HS, PMBs and neat by Shenoy and SHRP parameters These results are confirmed by the G* /sin and the Shenoy s parameter G* /(1- (1/sin tan [7] defined to predict the rutting behaviour of the relative mixtures (Fig 8). From the comparison between the two parameters it is possible to notice that for both modified binders, the high critical temperature (higher value of the SHRP Performance Grade) of the not processed neat bitumen increases from 66 C to more than 80 C when at least 15% of rubber is added. Shenoy s seems to better highlights the differences between the modified binders due to the higher influence of the phase angle variability.

14 14 Lo Presti et.al - AR2012 Figure 9. Black diagrams of neat bitumen, PMBs and TR-MB HS180 after 40 and 120 mins of high shear mixing Figure 10. Black diagrams of neat bitumen, PMBs and TR-MB HS210 after 40 and 120 mins of high shear mixing The comparison was extended also to the best-performing TR-MBs which were those obtained at the optimum processing time identified previously (Fig.9,10). This allowed assessing the level of modification achieved by using the optimum processing time at the selected temperatures, but also to assess the loss of properties

15 Lo Presti et.al - AR of the over-processed blends. Results confirm that processing TR-MB in high shear at 180 C for 120 min, instead of its optimum time (60-80 minutes), lead to a loss of performance in terms of G* and at high service temperatures, of about 10-20%. Instead, the TR-MB processed at higher temperature of 210 C reaches its peak performance in a shorter time (20-40 minutes) and after 120 minutes of high shear mixing the loss of performance is about 50-60% less of the peak. A confirmation of the loss of performance is highlighted in figures 9 and 10 where the Black diagrams of both the TR-MBs at 40 mins of high shear mixing are compared with the blend obtained at 120 mins and with the commercially used PMBs. 4. Conclusions From these results, it is possible to conclude that raising the processing temperature from 180 C to 210 C, as well as using high shear mixing, significantly speeds up the reaction between TR and bitumen and reduces the peak apparent viscosity. Nevertheless, by choosing the correct TR content and the appropriate mixing time, also at high processing temperature, 210ºC, it is possible to produce TR-MBs with a high level of modification. Furthermore, TR-MBs processed at 210ºC were easier to handle and less shear susceptible than TR-MB processed at 180ºC. The results confirm that the reaction process between TR and bitumen is sped up by the higher processing temperature but highlights that only with a preliminary design of the blends it is possible to obtain TR-MB with an acceptable level of modification also at very high processing temperature. This study demonstrates that TR-MBs are a very effective alternative to commercially used SBS-MB. Although the rheology of TR-MBs is very sensitive to the varying of processing conditions, therefore an appropriate selection of materials and a superior binder design are mandatory to optimise the modification process to achieve the desired level of modification. 5. Acknowledgements This study has been developed within a common project between the University of Nottingham (UK) and the University of Palermo (IT). The authors wish to acknowledge the University of Palermo for the significant contribution to the experimental programme. 6. References [1] R.G. Hicks, Asphalt Rubber Design and Construction Guidelines Vol.1 Design Guidelines. NCRACTC and CIWMB, [2] J.G. Chehovits, R.L. Dunning, G.R. Morris, Characteristics of asphalt-rubber by the

16 16 Lo Presti et.al - AR2012 slide plate microviscometer, Association of Asphalt Paving Technologists, Vol. 51, pp , [3] J.W.H Oliver, Optimizing the improvements obtained by the digestion of comminuted scrap rubbers in paving asphalts, Association of Asphalt Paving Technologists, Vol. 51, pp , [4] J.W.H. Oliver, Modification of Paving Asphalts by Digestion with Scrap Rubber, Transportation Research Record 821, Transportation Research Board, [5] Design and Construction of Asphalt Paving Materials with Crumb Rubber Modifier, Transportation M. Heitzman, Research Record [6] H.U. Bahia, R.Davis, Effect of Crumb Rubber Modifiers (CRM) on Performance Related Properties of Asphalt Binders, J. Assoc. Asph. Paving Technol., 1994, 63: [7] Shenoy, A., Rheology of filled polymer systems, Netherlands: Kluwer Academic Publishers, 1999 [8] Attia M., Abdelrahman M. Enhancing the performance of crumb rubber-modified binders through varying the interaction conditions. Int. J. Pavement Eng, 2009,10 (6): [9] S.H. Carpenter, Controlling low temperature properties of asphalt rubber binders, M.A. Abdelrahman, In: 2nd transportation specialty conference, Canadian Society of Civil Engineering, Halifax, Nova Scotia, Canada, [10] Asphalt Rubber Usage Guide, CalTrans, State of California Department of Transportation, Materials Engineering and Testing Services, 2006 [11] G.D. Airey Rheological properties of styrene butadiene styrene polymer modified road bitumens, Fuel, Vol. 82, No. 14, 2003, pp [12] P.S. Kandhal, Quality Control Requirements for Using Crumb Rubber Modified Bitumen (CRMB). J. Indian Road Congress, 2006, 67(1) paper n [13] H.B. Takallou, A. Sainton, Advances in technology of asphalt paving materials containing used tyre rubber. Transport. Res. Rec., 1339: 23-29, [14] Navarro F. J., Partal P., Martinez-Boza F., Gallegos C., Thermo-rheological behaviour and storage stability of ground tyre rubber-modified bitumens, Fuel, Volume 83, Issues 14-15, East Meets West on Heavy Oil Technology Symposium, October 2004 [15] Tayebali, A.A., Vyas, B.B., Malpass, G.A. Effect of crumb rubber particle size and concentration on performance grading of rubber modified asphalt binders. In: Jester RN editor, Progress of SUPERPAVE: Evaluation and Implementation ASTM STP 1322, p , [16] Lesueur, D., The colloidal structure of bitumen: Consequences on the rheology and on the mechanisms of bitumen modification, Advances in Colloid and Interface Science, Volume 145, Issues 1-2, Pages 42-82, 2009.