‘Advanced Testing and Characterization of Bituminous Materials ~ Loizos, Partl, Scarpas & Al-adi (eds) © 2009 Taylor & Francis Group, London, ISBN 978-0-415-55854-9. Nanoclay for binder modification of asphalt mixtures M.EC. van de Ven & A.A.A. Molenaar Delft University of Technology, Delft, The Netherlands Jeroen Besamusca Bitumen Department, Kuvait Petroleum Research & Technology ABSTRACT: Based on the positive experience with polymers nanoclay modification of bitumen was introduced. Nanoclay modification can influence properties like stiffness, fatigue resistance, strength, resistance to ageing and thermal stability. In this paper the potential of modifying bitumen with nanoclay is investigated. In all research a reference was compared to the modified material. Test results are reported for binders and asphalt mixtures containing standard and nanoclay modified bitumen. Influence of the nanoclay on ageing and rheology of the binder and influence on the performance of asphalt mixtures is discussed. The influence of nanoclay modifiers on fresh (unaged) and aged bitumen was measured with dynamic shear rheometer (DSR) and empirical tests like penetration and softening point. The effect of a nanoclay modified binder on an asphalt mixture was researched with the indirect tensile strength-, fatigue-, resilient modulus test at lower temperatures and dynamic creep test at high temperatures. The Cloisite nanoclay modifications increased the stiffness, the indirect tensile strength and the rutting resistance of a mix with standard 40/60 Pen bitumen. The Nanofill modifica- tion improved the ageing resistance of a 70/100 Pen bitumen in the short term ageing test strongly and good in the long term ageing test. The first conclusion is that nanoclay modifications can improve characteristics of asphalt binders and asphalt mixtures like ageing resistance and mechanical properties. Improvements are not yet such to justify application at large scale. Further studies on the chemistry of the nanoclay and bitumen are recommended to utilize the full potential of the nanoclay modifications. 1 INTRODUCTION 11 General A number of physical properties of polymers can be successfully improved when a polymer is modified with small amounts of nanoclay under the condition that the clay is dispersed at nanoscopic level (Pinnavaia et al. 2000, wwwiprime.umn.edu, Krishnamoorti et al. 2001, Ke et al. 2005, Al-Malaika et al, 2001). Based on the same principle, nanoclay modification of bitumen was introduced in the Netherlands by research institute TNO (Bos et al. 2004) Two types of nanoclay modifications were suggested: Nanofill and Cloisite. The expectation with the nanoclay modification is that it can influence properties like stiffness, fatigue resist- ance, strength, resistance to ageing and thermal stability. The purpose of this research was to test the potential of nanoclay modification of bitumen in relation to two important actual problems. In the Netherlands raveling of Porous Asphalt (PA) is a major durability problem that is strongly associated to the aging of the binder. If nanoclay could strongly reduce age- ing during service life, this would be a very important contribution to tackle the raveling problem. The second special application includes the design of asphalt mixtures for heavily loaded airport pavement. In this paper the contribution to reduction of aging and mechani- cal properties are taken as leading subjects to research the potential of nanoclay modified 133bitumen. The influence of nanoclay modifiers on binder characteristics was determined with rheological studies of fresh and aged bitumen with dynamic shear rheometer (DSR) and empirical tests like penetration and softening points. The effect of a nanoclay modified binder on the mechanical properties of an asphalt mixture designed for heavy loading situ- ations (airport) was determined with the indirect tensile strength test, fatigue test, resilient modulus test and dynamic creep test. A short introduction will be given in nanoclay modifi cation. The testing program will be discussed in general terms as it is impossible to describe all details (Ghile 2006). Some important test results will be given and discussed. The paper closes with conchisions, 1.2 Nanoclay Hardly any published information concerning nanoclay modified bitumen is available, How- ever, a large number of studies were done and are still underway on nanoclay modified poly- mers (Pinnavaia et al, 2000, www.prime.umn.edu, Krishnamoorti et al. 2001, Ke et al. 2005, Al-Malaika et al. 2001 ). This study focuses on nanoclay modified bitumen, but understanding of nanoclay modified polymers is of significance since both modifications can be based on similar principles. The term nanocomposite is in the broadest sense referring to every type of material with fillers in the nanometer size range at least in one dimension. A number of physi- cal properties are successfully enhanced when a polymer is modified with a relatively small amount of nanoclay on the condition that the clay is dispersed at nanoscopic level. Material variables which can be controlled and which can have a profound influence on the nature and properties of the final nanocomposite include the type of clay, the choice of clay pre-treatment, the selection of polymer component and the way in which the nanocomposite is incorporated into the polymer. Separation of the clay dises from each other will result in a nanoclay with an enormous large active surface area (it can be as high as 700 to 800 m* per gram). This makes an intensive interaction between the nanoclay and its environment possible (bitumen in our case). The process to realize the separation (surface treatment) is dependent on the type of material to be mixed, which can be explained as follows. Clay discs are negatively charged, but stay together in a clay particle because of the positive ions between the clay discs. In this way the clay particle as a whole is neutral, In addition, unmodified montmorillonite clays are generally highly hydrophilic species and therefore naturally incompatible with a wide range of polymer types. A necessary prerequisite for successful formation of polymer-clay nano- composites is therefore alteration of the clay polarity to make the clay ‘organophilic’. Dutch national research institute TNO has developed some nanoclay modifications that can be well mixed with polymers or bitumen (6). Bentonite and montmorillonite are the clays selected for this purpose. The developed nanoclay which is called organoclay can be processed to fit the binder. If the modifier is fully compatible with the binder, itis possible to get a homogeneous distribution of the clay at nano scale within the binder by intensive mixing However, there is a difference in the structure of bitumen and most polymers. Bitumen has avery complex molecular composition. The structure of polymers is mostly well defined and it can be altered and tailored to meet the demands much more easily. The complex and highly variable nature of bitumen could probably be a drawback to the success of the interaction between bitumen and the clay. It could demand a special approach to get interaction between clay and bitumen. TNO has selected two different nanoclays specially for bitumen. The treat- ment of the clay modifications was done by TNO (Bos et al. 2004). Based on the TNO research a Nanofill and a Cloisite were selected for use in bitumen, No detailed information on the nanoclays was made available from the supplier. 2 EXPERIMENTAL PROGRAM. 2.1 Overview testing program The testing program was designed first to give an indication if modification with nano- clay considerably reduce the aging of the binder: this question is of special importance for 134mixes with high void content (15%) like porous asphalt. Secondly it was questioned if modification with nanoclay can improve mechanical properties like stiffness, strength and fatigue and permanent deformation: this question is of special importance for heavy loaded airport pavements ‘As itis clear that these two different requirements could ask for different nanoclays, some pre-investigation was done on the two nanoclays selected for the research. The outcome of the binder testing of these two modifications resulted in a decision to select one modifica- tion for improved aging resistance (nanofill) and one for improved mechanical properties (cloisite). The testing program is summarized in table 1 The rheological properties of the modified binder were determined and compared to the standard binder. The binders considered for the test were a Penetration 40/60 bitumen for dense asphalt concrete and Penetration 70/100 bitumen for porous asphalt (10). The quanti- ties of modifications used were 3% and 6% cloisite and 6% nanofill by weight of bitumen, The 6% cloisite modified DAC was compared with the standard DAC. 2.2 Binder testing programme The theological tests were divided in empirical tests (penetration, softening point as used in Europe (EN-NEN 12591 2004) and fundamental tests with the DSR. For the ageing of the binder special attention was given to the possible influence on ageing of mixing in the nanoclay after I hour of high shear mixing. In principle the Rotating Cylindrical Ageing Tester RCAT (EN-NEN 15323 2005) was used for short term and long term aging. However, in the RCAT the binder is continuously moved during the ageing process. This is a reason- able assumption during production, but does not simulate the service life period very well. Because of the possible negative influence of the movements in the RCAT test on the effec- tiveness of the nanoclay against oxidation, it was decided to compare also with the pressure ageing vessel PAV (EN-NEN 14769 2005) 2.3. Dense mix testing program ‘The nanoclay is separately mixed in the bitumen and seen as part of the bitumen. ‘A dense mixture used at Amsterdam airport was taken as reference. Test specimens were compacted according to the Marshall compaction method with 2 * 75 blows. The average thickness of the test specimen was between 60 and 65 mm. The aggregate used in the prepa- ration of specimens is Norwegian granite and the composition of the dense mixture is given in table 2. The tensile strength, resilient modulus and fatigue resistance were determined with the Indirect Tensile test set up according to European specifications (EN-NEN 12697-23 2004, EN-NEN 12697-24 2004, EN-NEN 12697-26). The permanent deformation with dynamic creep (EN-NEN 12697-25 2004). For details of the test parameters see (Ghile 2006). No statistical analysis was done due to the low number of repetitions for most tests. For each Table 1. Overview of testing program. Group Materials Testing programme Rheology binders Bitumen 70/100, 70/100 Fresh, short term aged, long term (fresh and aged) nauofill Bitumen 40/60, aged 40/60 cloisite Empirical: penetration, softening point Fundamental: DSR master curves Mechanical properties dense Dense Asphalt Concrete Indirect tensile strength, mixture (DAC) DAC cloisite resilient modulus, modified binder Fatigue (all ITT) Perm, Deform. (uni-axial cyclic compression) 135Table 2, Composition of dense mixture. % [weight/(total weight weight of ‘Component type of aggregates)] % of components components 24-16 0.0% 0,0% 0.0 16-112 11,5% 108% 1323 1128 10.5% 1208 856 11.5% 1323 56-2 20.8% 239.2 Crushed sand 39.0% 48,5 Wigro filler 6.7% TL “Binder 6.0% 63.0 106,0° 1219.0 Binder = 40/60 for standard mixture and binder = 40/60 + 6% cloisite for modified dense mixture material one batch was produced, one batch was aged, etcetera. Penetration and softening point were just determined for the material used. One frequency sweep was done for each binder type. One Indirect tensile strength was done per temperature. Resilient modulus ITT was done 3 times, fatigue line determined on 8 samples per temperature, for the creep test three repetitions/temperature/loading condition were done. 3. TEST RESULTS The test results are discussed according to the general testing program given in table 1. A sum- mary of the rheological test results will be given first and after that the mechanical test results of the mix will be summarized. No statistical analysis is added, just an indication of what could be possible can be seen from the results. 3.1 Rheological test results on the (modified) bitumen Empirical rheological tests including the penetration and softening point test (because they are still leading tests in Europe) and fundamental rheological tests including master curves with the DSR were done. The penetration and softening point tests were performed on fresh (un-aged), short term and long term aged samples of standard and modified binders. Figure I shows that the modifi- cation improves the resistance to ageing of the 70/100 binder both for the retained penetration and the increment in softening point. Especially the PAV results look promising. From the PAV analysis it can be observed that there is some improvement in the resistance to ageing in the long term with the Nanofill modification. The results for Cloisite are given in Figure 2. From this figure it can be seen that the improvement is marginal, especially compared to the short term ageing results. In addition the modified 40/60 material has 2 higher viscosity than the regular 40/60 and therefore prob- ably sullers less contact with hot air or hot oxygen, Hence, the small improvement can also be due to the influence of the viscosity instead of the influence of the modification With the DSR test results it was possible to compare the master curves of the stiffness and the phase angle of the modified and unmodified binders and to analyze the change in ageing effect of the binder due to the modification and the rutting and fatigue resistance parameters for the modified and unmodified binders. Frequency sweep results can be used in the construction of master curves. All master curves in this project were composed at the reference temperature of 20°C. A typical example of master curves of the stiffness is given in Figure 3, All master curves for stiffness and phase angle are reported in (Ghile 2006). To get an indication of the effect of the nanoclay types is summarized with two parameters as developed in SUPERPAVE (NCHRP report 459 2001, NCHRP report 465 2002) for fatigue resistance and permanent deformation, with the assumption that RTFOT 136Ageing Bitumen 70/100 19 Bitumen 70/100 ‘0 ® 80) © 2 “ zr ga o si) 3 % “ ie ® » “Ni ” al [se fo B ° ° eRCAT ERAT Pav = ROAT |castendaa tumen 70/100 6 Navel (6%) + Blumen 7010 | Figure 1 bRcAT binder (s-RCAT = short term aged in RCAT, /RCAT = long term aged in RCAT), Ageing Bitumen 40/60 : & wo iii 2 80. aw Biss e-RCAT RAT oe ‘ate ()» Brame ao late (e+ Brame 460 (standard Bitomen 70/100 Nano (6) + Bitumen Toto Retained penetration and increment in softening point for standard and modified 70/100 "Gone (1) Sinan 400 Figure 2. Retained penetration and increment in softening point for standard (left column) and modi- fied 40/60 binder (s-RCAT = short term aged by RCA, LRA‘ T "Ofer reah and ing tr aged Cambined Master curve of 28°C for standard and Clos adie binder gree sen Figure 3. 40/60 binders (fresh and RCAT long term aged), 137 ng term aged by RCAT), Master curves at 20°C of stiffness of Cloisite 6% modified 40/60 binder and unmodifiedTable 3. Temperature values at minimum limits from RCAT of G*/sin8. ‘Temperature (°C) Binders Unaged-G*isind = 1 kPa Short aged- Gtisind = 2 kPa Standard 70-100 65.0 634 6% Nanofill, 65.2 628 Standard 40-60 69.0 68.4 3% Cloisite BS 69.5 6% Cloisite 168 44 jimum limits from RCAT of G#/sin6 Table 4, Temperature values at mi ‘Temperature (°C) Binders long aged-Gr*/sin8 = 5000 kPa Standard 70-100 (RCAT) 19,2 % Nanofill (PAV) 200 Nanofill (RC. 19.8 ‘Nanofill (PAV) 19.2 Standard 40-60 22.1 3% Cloisite 24.0 6% Cloisite 25.2 and PAV give similar results compared to s-RCAT and -RCAT. The rutting parameter (G*/sin8) is calculated at 10 rad/s (1.6 Hz) from the master curves. According to SUPERPAVE (NCHRP report 459 2001, NCHRP report 465 2002) the minimum limit of G*/sind is 1.0 kPa for unaged binder and 2.2 kPa for RTOFT aged binders and the limiting temperature corresponds to the highest temperature in which the binder is actually expected to serve. The maximum tempera- ture values at which the minimum limit of G*/sin8 is still satisfied both for unaged and short term aged binder is given in table 3. The results of Table 3 show that the 6% Cloisite modified binder shows significant better rutting resistance than the standard and the 3% Cloisite modi- fied 40-60 binders. This is not the case for the 6% Nanofil To improve the fatigue life the work dissipated during each loading cycle should be mini- mized (NCHRP report 459 2001, NCHRP report 465 2002, EN-NEN 12591 2004). Limit- ing the G*/sind parameter, decreasing G* and/or sin6 will result in better fatigue behavior. A summary for all binders of the minimum temperature values at which the maximum limit of G*/sind is satisfied is given in Table 4 The results of Table 4 indicate that the Cloisite modified 40/60 binder has a higher fatigue temperature than the standard 40/60 binder, whereas the Nanofill modified binder shows more or less the same fatigue temperature as the standard 70/100 binder in the long term aged samples 3.2. Mechanical properties of the dense asphalt concrete Indirect tensile strength, resilient modulus, dynamic creep and fatigue results are reported in this section, 3.2.1 Tensile strength Tensile strength was determined with the indirect tensile test (EN-NEN 12697-23 2004), A compar- ison of ITS values of the modified and unmodified dense specimens is summarized in Figure 4. ‘A comparison of the ITS results of the modified and unmodified dense mixtures show that the ITS results of the modified specimens is higher than the ITS of standard mixtures at all test temperatures, a percentage increment in ITS values due to the modification varies from 8% to 40% and the percentage of increment is higher for the higher testing temperature, 138Comparison of indirect tensile strength ct tensile strongth(MP) temperature) Figure 4, Comparison of ITS (in MPa) of modified and unmodified dense mixes, Reallant modulus (MPa) we Loading time (ms) wee SCONE ETRE COX w20C6KC .ILECELC uIsCENe $5C8 xiases x20es MSCS owes Figure 5. Comparison of resilient modulus (MPa) of standard and 6% Cloisite modified dense mixes. at 35°C the specimen of the modified mixture fails due to pure crack where as the specimen of the standard mixture fails due to crack and deformation. Also fracture energy was determined from the load-displacement curves in the ITS, In all, cases the Cloisite modified mixture showed increased fracture energy compared to the refer- ence mixture (Ghile 2006). 3.2.2 Resilient modulus results comparison of the average resilient modulus between the mixture modified with 6% cloisite and the standard mixture is given in Figure 5. It shows that the dense mixture modified with 6% clois- ite has a higher stiffness than the standard dense mixture at all test temperatures; tthe 6% cloisite modification increases the stiffness between 3% to 17%, An average 12% increase in stiffness is obtained with 6% cloisite modification for the frequencies and testing temperatures considered 3.2.3 Dynamic uni-axial creep test results. ‘The dynamic creep test was carried out at three temperatures and at different stress lev- els, according to (EN-NEN 12697-25 2004). An example of typical creep test results at a 139temperature of 40°C is given in Figure 6. From the cteep curves the mean rate of permanent deformation in the secondary stage (K) and the flow number (the number of cycles where the tertiary creep stage starts), are the two parameters recommended as rutting resistance indica- tors Ghile 2006, NCHRP report 465 2002). In this paper K is summarized in figure 7 for all mixtures and test conditions. All rutting resistance indicators show that the dense mixture modified with 6% Cloisite results in better rutting resistance compared to the standard dense mixture. This is in line with the indications from the rutting analysis of the binders. 3.24 Fatigue results ‘The indirect tensile test (EN-NEN 12697-24 2004) was used to compare the fatigue resistance of the standard and modified dense mixtures. In this test, a constant repetitive haversine compressive load is applied and the resulting vertical deflection is measured in tow (ACS standard mi [ACC - Mocs ubie.c) ‘Accumulated strain(umm) Teper ACE pulse counts Figure 6. Dynamic creep test results at 50°C. ° 100 200 300 400 500 600 stress (kPa) x 40°Cstandard = 40°C-6% coiste "50°C-standard” + 50°6-6% cloiste —x- 60°C-standard —* —60°0-6% cloisite Figure 7. Rate of permanent deformation (K) in relation to the vertical stress level (in kPa). 140Table 5. _k and R-square values at 5°C and 20°C test temperature for N = kyo". Temp Mixture type k k, R xc Standard dense mix 1,549 x 10" 41077 0,964 6% cloisite modified 7,568 x 10 4,362 0.9995 dense mix 20°C Standard dense mix 2.716 x 108 3.3249 0,964 Gp cloisite modified 4.467 x 108 3.748 0,995 dense mix relation to time (pulse counts). The fatigue life of a specimen was defined as the number of load repetitions at which specimen failure takes place. Here it is reported as stress versus fatigue life. Fatigue tests were done at 5°C and 20°C with the ITT test set up. The relationship between ¢ and Nf at 5°C was determined by means of linear regression analysis at log scale of the formula: N = ko, . The k values and R-square values are summarized in Table 5. ‘As can be seen from the results at 5°C the standard dense mixture performs better in fatigue than the 6% cloisite modified dense mixture, whereas at 20°C the dense mixture modi- fied with 6% cloisite performs better than the standard dense mixture. The comparison of the fatigue parameter, G*/sin8, of the standard 40/60 binder and 6% cloisite modified 40/60 binders at 20°C shows that the fatigue life of the standard binder is better than that of the 6% cloisite modified 40/60 binder. At 20°C, G/sin8 of the 6% cloisite modified 40/60 binder is about 1.4 times that of the standard 40/60 binder and this could result in a reduction in fatigue life due to the modification. This is not in line with the conclusion drawn from the fatigue performance of the modified and unmodified dense mixtures in table 5. A physical explanation could be that at 20°C the test set up is more vulnerable to creep and the modified mix performs better in creep. 4 CONCLUSIONS The conclusions should be seen as indications from a preliminary test programme in which no statistical analysis could be done for most of the parameters due to low number of test repetitions. It was not possible to design this test programme statistical, due to the limited possibilities. Also the decision to use the Superpave parameter values for RCAT aged mate- rial is questionable and is only done for comparison (ranking) indication. 4.1 Rheological test results on binders The Cloisite (3 and 6%) modification increases the stiffness of the bitumen considerably ‘Nanofill hardly changes the stiffness. The nanofill (6%) modification improves the ageing resistance of the 70/100 binder in the short term and long term too. Both the empirical tests (Pen SP) and fundamental tests (DSR) show the same trend for Cloisite and Nanofill, Permanent deformation prediction based on stiffness and phase angle (G*/sin6) show strong improvement with cloisite and no improvement with Nanofill. Fatigue prediction (Gt/sin8) shows higher temperature for Cloisite and same temperature for Nanofill 4.2. Mechanical properties of dense asphalt mixture Cloisite modification (6%) increased the strength, stiffness, and strongly improved the rut- ting resistance of mixtures with the standard 40/60 binder. Fatigue resistance performance of mixtures with 6% Cloisite modification is not improved at low test temperatures. M143° General ‘The improvements observed with the nanoclay modification are not at a level yet to justify applica- tion at large scale, Assuming that the improvement is related to the chemical additive that opens up the clay structure and is responsible for the interaction of the clay mineral with the bitumen, further studies on the chemistry of the nanoclay and bitumen and further development of the nanoclay technology is necessary to utilize the full potential of the nanoclay modification. Also storage sta- bility in bulk applications should be investigated especially with nanoclay percentages above 3%. REFERENCE: Almalaika, Golovoy A. & Wilkie C.A. 2001, Specialty polymer additives principles and applications, Blackwell science Ltd, London. Bos B., Fischer H.R. & Timmer K. 2004, Possibilities to improve the wear properties of bituminous anticskid with nano composite technology (ia Dutch). Report DMP-RRI-04-071, TNO-TPD, The Netherlands Ghile D-B, 2006, Effects of nanoclay modification on the rheology of bitumen and performance of asphalt mixtures. MSc thesis Delft University of Technology. Ke ¥.C. and Stroeve P. 2005, Polymer layered silicate and silica nanocomposites, Elsevier science pub: lishers B.V., Amsterdam, Krishnamoorti R. and Vaia R.A. 2001, Polymer nanocomposites, Oxford University Press, Washington Pinnavaia TJ. and Beall G.W. 2000, Polymer-clay nanocomposites, John Wiley and Sons Ltd, England, EN-NEN 15323, 2005, Accelerated long-term ageing conditioning by rotating cylinder method (RCAT), European standard. EN-NEN 14769, 2005, Accelerated long-term ageing conditioning by pressure ageing vesel (PAV), European standard EN-NEN 12591, 2004, Bitumen and bituminous binders-Specification for Paving Grade bitumen NCHRP report 459, 2001, Characterization of modified asphalt binders in superpave mix design, Washington D.C. NCHRP report 465, 2002, Simple performance test for superpave mix design, Washington D.C. EN-NEN 12697-23, 2004, Determination of the indirect tensile strength of bituminous specimens, European standard. EN-NEN 12697-24200, Resistance to fatigue of bituminous mixtures, European standard EN-NEN 12697-26, 2004, Bituminous mixtures — test methods for hot mix asphalt -part 26 — stiffness, European standard. EN-NEN 12697-25, 2004, Bituminous mixtures ~ test methods for hot mix asphalt -past 25 ~eyeic compression test, European standard |http:/Avwv:iprime.umn.edu/pdfs/NanoWorkshop/Hunter1.I’MPnano050310.pdf, Nanocomposite Organoclays. 142