‘Advanced Testing and Characterization of Bituminous Materials — ‘Loizos, Pat, Scarpas & ALQadi (ods) (© 2009 Taylor & Francis Group, London, ISBN 978-0-415-55854-9 Assessment of water sensitivity of asphalt rubber mixtures for wearing course FA. Batista & M.L. Antunes Laboratorio Nacional de Engentaria Civil, Lisboa, Portugal P. Fonseca RECIPAY, Cartaxo, Portugal ABSTRACT: To assess the water sensitivity of asphalt rubber mixtures an experimental study was carried out, in which different types of gap-graded mixtures applied in wearing ‘courses were used. The experimental study compares the results obtained by the test method traditionally used in Portugal - the American army standard MIL-STD-620A—method 104 - with those obtained through the EN 12697-12:2008-method A. The application of the European standard leads to lower values of retained resistance than those obtained with the American army standard and allows for a better distinction of the behavior of differ- cent mixtures. This paper presents the main results achieved so far, and contributes for the achievement of reference values for water sensitivity of gap-graded asphalt rubber mixtures determined according to the European standard EN 12697-12:2008—method A. 1 INTRODUCTION In Portugal, asphalt mixtures produced with bitumen modified with a high percentage of rubber (bitumen-rubber) are used in paving works since 1999. During these years, several studies have been developed by the Portuguese National Laboratory of Civil Engineering (LNEQ) in order to characterize the performance of this type of mixtures (Antunes ef al 2000, 2003, 2006, Batista er al. 2008, Miranda et al. 2008). The present paper refers in par- ticular to the evaluation of water sensitivity of asphalt rubber mixtures for application in wearing courses. In Portugal, before the introduction of European standards, the minimum values required for water sensitivity of asphalt mixtures were established based on the American army stand- ard MIL-STD-620A—method 104 (CRD-C 652-95). In the present context of the European standards, in particular the EN 12697-12:2008, it is necessary to establish reference values to be obtained in tests performed according to this, European standard. Indeed, there are significant differences between the test methods of the MIL-STD-620A and the EN 12697-12 standards, which may lead to very distinct results of | the water sensitivity for a given mixture. ‘This paper reports an experimental study carried out in order to assess the water sensitivity of asphalt rubber mixtures in which different types of gap-graded mixtures were used. ‘Theexperimental study compares the results obtained with the test method traditionally used in Portugal with those obtained through the European standard EN 12697-12—method A, in which the effect of saturation and accelerated water conditioning on the indirect tensile strength of cylindrical specimens of bituminous mixtures is determined. One conclusion is that the application of the European standard leads to lower values of water resistance than those obtained with the American standard and allows a better distinction of the behavior of different mixtures. ‘This paper presents the main results achieved, and gives a contribution for establishing ref- ‘erence values for water sensitivity of gap-graded asphalt rubber mixtures determined accord- ing to the European standard EN 12697-12—method A. 3232. EVALUATION OF WATER SENSITIVITY In general the assessment of water sensitivity of bituminous mixtures is performed by determining the mechanical strength of two groups of bituminous mixtures test specimens with identical characteristics, which are previously submitted to different conditioning con- cerning to the detrimental effect of water (Airey & Choi, 2002). The ratio between the aver~ ‘age resistance of the set of specimens submitted to more severe conditioning and the one in less severe conditioning provides an “index of retained strength”, which is used as an indica tor of water sensitivity ‘A brief description of the MIL - STD - 620A and EN 12697-12 test methods is presented in the following sections. 2.1 Test method 104 of the MIL - STD - 620A standard In order to assess water susceptibility of bituminous mixtures it has been of common prac- tice in Portugal to use a test method based on the American standard MIL - STD - 620A, According to this test method, the reduction of Marshall stability of bituminous mixtures specimens caused by immersion in water for a prescribed period is measured, With this purpose, the bituminous mix with the given composition is manufactured and ‘a minimum of $ Marshall test specimens of 101.6 mm in diameter and about 63.5 mm in hheight are prepared. These specimens are divided into two groups with approximately the same average bulk density ‘The frst group is immersed in water at 60°C for 30 minutes (for 20 t0 40 minutes according to the method 100 of the military standard (CRD-C 649.95), and then the Marshall stability for each specimen is determined and the average value (F,)is calculated. ‘The second group is conditioned for 24 h ina water bath at 60°C, after which the Marshall stability for each specimen is determined and the average value (F,) is calculated. Based on the results for the average Marshall stabilities, the Index of Retained Stability (IRS) is estimated as follows (Equation 1) a 2.2. Test method A of the EN12697-12:2008 standard ‘The evaluation of sensitivity to water of bituminous mixtures recommended in the method A of the EN 12697-12: 2008 is based on the comparison between the average values of indirect tensile strength (determined in accordance with the EN 12697-23) of two groups of eylindri- cal samples previously conditioned indifferent conditions. In this ease, 3 cylindrical test specimens are prepared for each group, whose dimensions are defined taking into account the dimension of the maximum aggregate. For mixtures with ‘maximum aggregate size not exceeding 22 mm, specimens with a (100 # 3) mm in diam- eter can be used, which means that test specimens are similar to those used in the previous ‘method, compacted by the impact compactor, according to the EN 12697-30 “The test specimens are also be divided into two subsets having approximately the same average bulk density, as well a a similar average length. In this case, the dilerence between the average bulk densities of the two subsets is limited to a maximum of 30 kg/m’ for the and the difference between the average lengths is limited to 5 mm. Iti also established that the test specimens must be of the same age (prepared within one week), and have a curing time of IGh to 24h, ‘The first subset of specimens (dry specimens) is stored in air at (20 + 5)°C. The second group (wet specimens) i Firstly submitted to a vacuum (absolute pressure (6.7 + 0.3) kPa) In ater at (20-4 5) in order to ensure that water penetrates adequately in the pores of the test specimens, after which they are stored in a water bath at (40+ 1)°, fora period of 68 to 72h. The volume of specimens from the wet subset must be checked after the application 324of vacuum. Any specimen that has suffered an inerease of more than 2% in volume must be jected. ‘ter conditioning, the indirect tensile strength of each specimen is determined, at atest temperature in the range of $*C to 25°C. The EN 12697-12 standard includes a note which recommends that a test temperature of 25°C should be selected to obtain the maximum influ- ‘ence from binder adhesion and to minimize influence from broken aggregates in the break line surface. Nevertheless, the EN 13108-20 standard establishes a temperature of 15°C for the initial type esting of CE marked bituminous mixtures. Based on the ratio between the average indirect tensile strength of water conditioned speci sens (ITS, and dry specimens (ITS,), the ITSR value, in percent, i calculated according the formula (Equation 2 ms, Ts, 1% @ ITSR= 2.3. Features related tothe implementation of the method A of the EN12697-12 standard Earlier experiences developed at LNEC with the EN 12697-12 test method A have shown that its application presents some difficulties, mainly concerning the application of vacuum. to the wet subset of test specimens. The first results obtained for ITSR using EN 12697-12 test method A, showed a relatively high dispersion, with values which were much lower than those expected considering the type of mixtures tested. This type of problems were attributed. to the application of vacuum to the wet specimens. ‘Taking into consideration the experience gained so far, itis recommended that the follow- ing steps are strictly followed in the application of vacuum (Fig. 1) ~ Place the wet subset of specimens in the vacuum container (pycnometer filled with dis- tilled water at 20 + 5)°C up to level atleast 20 mm above the top of the test specimen; ~ Apply a vacuum to obtain an absolute pressure of (6.70.3) kPa within (10-1) min. Itis ‘very important to reduce the pressure gradually to avoid any damage. For thi, the pressure should be reduced in steps between the atmospheric pressure (101.3 kPa at sea level) and the target value in steps, which can be achieved monitoring the duration of each step with a stopwatch, ~ Keep the vacuum during (30 + 5) minutes ~ Allow pressure to grow slowly till the atmospheric pressure. ~ Leave the specimens immersed for a further period of (30-£ 5) minutes. Figure 1. Vacuum being applied to the wet subset of test specimens 3253. WATER SENSITIVITY OF ASPHALT RUBBER MIXTURES In order to assess water sensitivity of gap-graded asphalt rubber mixtures, two mixtures were produced using the same bitumen rubber binder produced with a 35/50 grade virgin bitumen and two different gap-graded mixtures: one produced with granite and another with crushed gravel. Table | identifies the tested mixtures, The characteristics of the materials used in the production of the AR mixtures are described in the following paragraphs. 3.1 Bitumen rubber binder The binder used for the production of the asphalt rubber mixtures, usually known as itumen-rubber” (BR) was obtained through the in-situ modification of a virgin bitumen with crumb rubber though the so-called “wet process”. Presently, there isa wide range of bitumen rubber binders in the Portuguese market, such as in situ manufactured binders with high content of crumb rubber and ex situ manufactured binders with lower contents of erumb rubber. This paper refers specifically to AR mixtures produced with bitumen rubber containing more than 18% of crumb rubber, known as “modified bitumen with high content of crumb rubber”, and produced with a 35/50 penetration grade bitumen. Table 2 presents the main properties of the BR used in this study. 3.2. Aggregates The mixtures tested in this study were gap-graded mixtures produced with 0/12.5 mm aggre- gates of two different natures: granite and crushed gravel. Table 3 presents the composition of the aggregates used in the production of the referred AR mixtures, and Figure I represents the corresponding grading. 3.3. Experimental programme This study comprised AR water sensitivity assessment through the MIL - STD - 620A and the EN 12697-12 standards. In the last case, the indirect tensile tests were conducted at Table L__ Tested asphalt rubber mixtures identification. AAR mixtures identification GG-pravel-BR GG-granite-BR Aggregate nature Crushed gravel Granite Binder characteristics 82% B3SIS0 + 18% rubber 82% BSSIS0+ 18% rubber Binder content s5% 8 Table 2. Rubber modified bitumen (I 10 22% of rubber) properties Viscosityat 17S°C* Penetration a 25°C Softening point (20pm) (100 2,55) (ing and ball method) EN 133022003, EN 1426:1999 EN 1427:1999 Bitumen eubber = Ur firs Pas Om © Rubber modified 35/50 grade bitumen 2500 404500, Isto30 268 "Determined using a Brookfield apparatus with thermosel and a SC4 -27 spindle 326Table 3. Aggregates compositions ‘AR mitures identification Fraction dimension Fraction percentage GG-gravelBR 10/16 fraction 4110 fraction 6 fraction Commercial filler GG-granite- BR 10/15 fraction | 6110 fraction 5 fraction Commercial filler 100 ‘Gradation composition %” specification limits ,™ 66-Grante = 66.Gravel Fo 2 00 5 50 E40 zw “20 0 0 oot a1 1 0 100 Sieve size (mm) Figure 2. Aggregates gradation Figure 3. Snapshots ofthe tet method A of the EN 12697 -12:2008 327Table 4. Results of the water sensitivity tests (MIL -STD - 620A—method 108), ‘Water sensitivity test Specimens 30 min, Specimens 24 AR mixtures Average air voids immersed in water immersed in water identification Property a °C aac CG-gnvelBR 28 ‘Average bulk 278 B3 12 9 GG-sranite BR 46 Average bulk 2202 ‘density (kn) Average Marshall 11.8 3 96 able S._ Results of the water sensitivity tests (EN 12697-12:2008-method A), Water sensitivity test Specimens submited Specimens tovacuum and AR mixtures air voids conditioned immersed in water at identification __ content Property Bdaysat 27°C 40°C for = 70h GG-grvelBR 25 ‘Average bulk 286 228s density (kg/m) Average indirect 1157 m tensile strength (kPa) TTSR 7 GG-zraniteBR 48 Average bulk 221 density (kg/m) Average indirect 990 si tensile strength (ea) TTSR 86 temperature of 25°C taking into account the benefits outlined in the EN 12697-12 for the selection of this test temperature. In both cases, Marshall specimens were prepared using @ impact compaction (EN 12697-30) with an energy level of 2x75 blows. Figure 3 illustrates two aspects of the application of method A recommended by EN 12697-12. Tables 4 and 5 present respectively, the results obtained in the water sensitivity tests per- formed according to MIL - STD - 620A and European standard, Figure 4 summarizes the retained strength obtained for the two different mixtures tested according to the referred test methods From the results presented it can be noticed that: ‘* The application of US. military standard (MIL-STD-620A) leads to very high values of | retained strength, providing virtually the same results for both mixtures; 328IRS, % (MIL. STD - 620A) IITSR, % (EN 12697-2008—method A) 100 z & eo 2 a 3 3m 0 GG-gravel- BR GG-geanite-BR AR mixture ID FFigure 4. Variation of retained strength obtained using different test methods (MIL - STD - 620A standard versus EN 12697-12:2008—method A). + ‘These values were reduced when the water sensitivity tests were performed according to the European standard (EN 12697-12:2008—method A); + The European standard provides a better distinction between the behaviour of the referred mixtures in the presence of water: as expected, the gravel AR mixture presented lower val ues for the retained strength (ITSR values). 4. FINAL REMARKS This paper presented the first experiments undertook at LNEC in order to evaluate the water sensitivity of bituminous mixtures with a high percentage of rubber, according to the Euro pean standard EN 12697-12. The experimental study compares the results obtained through the test method tradition: ally used in Portugal with those obtained through the European standard EN 12697-12. The application of the European standard leads, in general, to lower values of retained resistance when compared with those obtained with the US military standard and allows for a better distinction between mixtures produced with different aggregates. There have been, however, some difficulties with the implementation of the European test method, especially during application of the vacuum to the immersed specimens. This paper presents some recommendations to overcome these difficulties. REFERENCES, Ai s. GD. & Choi, ¥-K. 2002. State of the art report on moisture sensitivity test methods for bituminous pavement materials International Journal of Road Materials and Pavement Design 3 (4):355-372, Antunes, M.L; Batista, FA; Eustbio, 1; Costa, MS. & Miranda, C.V. 2000, Characterization of asphalt rubber mixtures for pavement rehabilitation projects in Portugal, Proc of the “Asphalt Rub ber 2000”, Vilamoura, Portugal, 14-17 November 2000:285-307 Antunes, M.L.; Domingos, P; Euscbio, I. & Costa, M.S. 2003. Studies concerning the use of Asphalt Rub- ber in Portugal. Proc. ofthe “Asphalt Rubber 2008", Brasilia, Brasil, 1-4 December 2003:195-210, 329Antunes, M.L; Batista, FA. & Fonseca, P. 2006. Asphalt rubber mixtures in Portugal: practical application and performance. Proc. of the “Asphalt Rubber 2006”, Palm Springs, California, USA, 25-27 Octoher 2006, pd, Batista, FA; Antunes, M.L; Miranda, H. & Fonseca, P 2008. Fatigue resistance of asphalt rubber mix- tures Proc of the TRA2008 - Transport Research Arena Europe 2008 conference, Ljubljana, Slovenia, 21-24 April 2008. CRD-C 649-95 Standard Test Method for Unit Weight, Marshall Stability, and Flow of Bituminous Mixtures (Formerly MIL-STD-620A, Method 100, 13 January 1996), Construction Criteria Base ~ Handbook for Concrete and Cement, Whole Bulding Design Guide, bup:}www-whds org, 2008. CRD-C 652.95 “Standard Test Method for Measurement of Reduction in Marshall Stability of Bitu- ‘minous Mixtures Caused by Immersion in Water” (Formerly MIL-STD-620A, Method 104, 13 Janu ary 1966), Construction Criteria Base - Handbook for Concrete and Cement, Whole Building Design Guide, bateliwww.vtodg.org), 2008, EN 12697-12:2008—European standard EN 12697 - Bituminous mixtures—Test methods for hot mix ‘asphalt, Part 12 - Determination of the water sensitivity of bituminous specimens. EN 12697-23:2003European standard EN 12697 - Bituminous mixtures Test methods for hot mix ‘asphalt, Part 23 Determination of the indirect tensile strength of bituminous specimens. EN 12697-30004 + A1:2007—European standard EN 12697 - Bituminous mixtures—Test methods for hot mix asphalt. Part 30 - Specimen preparation by impact compactor. EN 13108.20:2006European standard EN 13108 - Bituminous mixtures~Material specifiations. art 20: Type testing Miranda, H.; Batista, FA.; Neves, J; Antunes, M.L. & Fonseca, P2008. Asphalt Rubber Mixtures in Portugal: Fatigue Resistance. Proc of the “Ind European Conference on Pavement and Asset Manage ‘ment, Coimbra, Portugal, 7-9 July 2008, CD-ROM, 330.