STRENGTH CHARACTERISTICS OF POLYMER MODIFIED ASPHALT BINDERS AND MIXES

Abstract

With the ever increasing demands of traffic and temperature imposed on highways, it has become mandatory to find alternative means, either in terms of design or use of better materials, to tackle these issues. Modification of bitumen is one of the simplest and most effective techniques which has been used over years to increase the strength and life of pavements. Out of various forms of modification, polymer modification has gained the most importance due to its capability to improve the viscoelastic response of the bitumen. It increases the stiffness of asphalt binder at higher temperature while maintains the flexibility at lower temperatures, thereby improving both the rutting and fatigue resistance of bitumen. The study of viscoelastic response of bitumen is usually done using dynamic shear rheometer (DSR). DSR can capture various rheological aspects of bitumen at a wide range of frequency and temperature. The traditional Superpave parameters for characterization of rutting and fatigue resistance of bitumen (G*/sinδ and G*.sinδ) has been criticized in many literatures and has been found inadequate, specially for modified binders. Improvement in the test methods came out in form of multiple stress creep and recovery (MSCR) and linear amplitude sweep (LAS) test for quantifying the rutting and fatigue characteristics of bitumen. The study of binder alone is not sufficient for judging the performance of the pavement. Aggregate gradation also plays a vital role in describing the overall performance of the asphalt mixture. Asphalt mixture are usually dense graded, gap graded or open graded. The volumetrics of these different aggregate gradation along with the properties and amount of asphalt binder is responsible for total characterization of the asphalt mixture. In India, Marshall mix design is used for evaluation of optimum binder content. In order to evaluate the permanent deformation and fatigue characteristics of asphalt mixtures, tests like wheel rut test and four point beam bending test are the most common. Correlation between the binder and mix attributes is another important aspect for judging field performance of the asphalt mix. This research outlines the study conducted on conventional and polymer modified asphalt binders and mixes. The conventional and rheological properties of the asphalt binders iv are studied in detail at a temperature range of 10-70 ˚C. The performance of these binders with respect to fatigue and rutting is evaluated and the suitability of binder for different practical conditions is explored. The study also evaluates the performance of different aggregate gradations with use of tests like Marshall stability and flow, indirect tension test, wheel rut test and four point beam bending test. The correlation of binder properties with mix results are also taken up as a part of this research. The study begins with finding the optimum blending requirements and optimum modifier content of two polymers i.e ethylene vinyl acetate (EVA) and styrene butadiene styrene (SBS). SBS was incorporated in VG 10 at five different modifier content of 1-5% while EVA was blended at 1-7%. The blending temperature, shear rate and time was varied from 160-200 ˚C, 300-1500 s-1 and 20-60 minutes respectively. Conventional binder like VG 10 and VG 30 were also used for comparison. Storage stability and fluorescence microscopy were used to find the optimum blending requirements and modifier content. From the analysis it was found that SBS could be incorporated in the base binder at a temperature of 180 °C using a high shear mixture operated at 1500 rpm for 60 minutes. The corresponding temperature, shear rate and time for EVA were 190 °C, 600 rpm and 30 minutes. Storage stability test showed that optimum modifier content for EVA is 5% while for SBS is 3%. This result was also validated by the study of morphology using Fluorescence microscopy. An interlocked bitumen-polymer phase is the most desirable for obtaining a homogenous blend. The flow properties of the binders were evaluated using steady shear method. The viscosity versus shear rate for the binders were measured using DSR at a temperature range of 40-80 ˚C. Carreau-Yasuda (C-Y) equation was combined with the concept of rheogram to simulate the variation of viscosity versus shear rate. A new model was proposed which could be used to evaluate the mixing temperature of asphalt binder corresponding to the viscosity at any desired shear rate. The model was found to be critical to the value of zero shear viscosity (ZSV). At higher shear rates the viscosity of the modified binders were found to be even lower than the conventional binders owing to the shear thinning behavior, which indicated that lower mixing temperatures could be used for modified binders at practical shear rates. v The rheological measurements for all the binders were done using frequency sweep test at a temperature range of 10-70 ˚C. The linear viscoelastic (LVE) limits were evaluated as the initial part of the study. The effect of spindle geometry and plate gap was also evaluated. 25 mm and 8 mm diameter spindle were used. For 25 mm, three spindle gaps viz. 1 mm, 2 mm and 3 mm were used, while for 8 mm diameter 1 mm and 2 mm gap was employed. Master curves were constructed at three different reference temperatures of 20, 40 and 60 ˚C. Various shift factor methods were analyzed and a new method was proposed for the construction of master curves. The method was named as ‘Equivalent Slope Method’. The complex modulus and phase angle master curves were modelled using the concept of rheogram and C-Y equation. At higher temperatures spindle diameter 8 mm gave higher values for complex modulus and phase angle of the asphalt binders as compared to 25 mm diameter spindle. At higher temperatures higher plate gap gave lower values of complex modulus as compared to lower gap width. The difference in the measurements decreased with decrease in temperature with similar values at intermediate temperatures. It was found that spindle geometry plays a crucial role in determination of rheological properties of both conventional and modified binders. The shift factor obtained by the equivalent slope method which was developed in the study gave the better results in plotting master curves as compared to WLF and Arrhenius equation. It was found that Cox-Merz rule can be successfully applied in the zero shear viscosity domain. The simple C-Y model was found to be successfully applicable in modelling the rheological properties of both conventional and modified bitumen. Phase angle was found to be sensitive to the type and chemical nature of bitumen. The performance of the binder with respect to rutting and fatigue were evaluated using MSCR and LAS test. Multiple Stress Creep and Recovery (MSCR) test was conducted at three different temperatures (40, 50 and 60 °C) using Dynamic Shear Rheometer (DSR) operated in constant stress mode. Four different stress levels were chosen, viz. 100, 3200, 5000 and 10,000 Pa. The test was done using 25 mm sample geometry with 1 mm gap between the spindle and the base plate on RTFO aged samples. The strain response was modelled using Burgers four element model and power law model. Applicability of Boltzmann superposition principle was also checked in the study. Burger model was not able to model the delayed elastic response of asphalt binders. The power law model was modified vi for the recovery domain to incorporate the effect of delayed elastic response. The model parameters were analyzed and were correlated with the conventional properties like softening point and ZSV of the binders. LAS test was conducted at 10, 20 and 30 ˚C to determine the fatigue lives of the asphalt binders. The parameters A and B were evaluated at different temperatures. LAS test was found to be more practical than the existing intermediate performance criteria of G*.sinδ. By LAS test it was possible to evaluate the complex behavior of the binder at a wide range of loading level. PMB (S) gave the best overall performance in both the test methods. PMB (E), though performs well at higher temperature, but at intermediate pavement temperature it may be susceptible to fatigue cracking attributed to higher sensitivity to strain amplitudes. Among the conventional binders both VG 10 and VG 30 were found to be suitable for resisting fatigue cracking at intermediate pavement temperatures. Nevertheless, at higher temperatures both VG 10 and VG 30 showed poor performance. Marshall mix design was used for the preparation of asphalt mixtures with three different aggregate gradation viz. bituminous concrete (BC), dense bituminous macadam (DBM) and stone mastic asphalt (SMA). All the mixes were prepared at 4% target air void content. The moisture susceptibility of the mixtures were also evaluated using retained Marshall stability test and tensile strength ratio. The film thickness for these mixtures were also calculated. Modified mixtures were found to have higher strength than the mixes prepared with conventional binders. The Marshall stability and indirect tension strength of SMA was found to be lower than BC and DBM, however their retained Marshall stability and tensile strength ratio were higher indicating higher resistance to moisture damage. Moreover, the film thickness for SMA was found to be higher in comparison to BC and DBM, attributed to higher void in mineral aggregates (VMA). This indicated better durability for these mixtures. The last part of the study dealt with the evaluation of rutting and fatigue performance of the asphalt mixtures. Rutting characteristics was evaluated using wheel rut testing at 60 ˚C. Fatigue performance of the mixes was evaluated using four point beam bending test (4PBBT). This test was performed at 20 ˚C at a strain level ranging from 200-1000 micro-strains. A new phenomenological model was proposed to characterize the variation of fatigue vii life at different strain levels. Correlation of rutting and fatigue test with binder performance was also attempted. It was found that SMA had the highest rutting and fatigue life in comparison to BC and DBM. Plastomeric modified mixes were found to be highly susceptible to change in strain levels. In rut depth was found to correlate fairly well with the unrecoverable creep compliance from the MSCR test. The fatigue life of asphalt binders from LAS test showed linear correlation with the fatigue life from 4PBBT. The new phenomenological model proposed in the study correlated appreciably with the measured response.