Acta Polytechnica CTU Proceedings doi:10.14311/APP.2020.26.0112 Acta Polytechnica CTU Proceedings 26:112–116, 2020 © Czech Technical University in Prague, 2020 available online at https://ojs.cvut.cz/ojs/index.php/app MECHANICAL AND IMAGE ANALYSIS OF ADHESION BETWEEN MINERAL AGGREGATE AND BITUMINOUS BINDER Jan Trejbal∗, Tereza Valentová, Václav Nežerka, Petr Semerák Czech Technical University in Prague, Faculty of Civil Engineering, Department of Mechanics, Thákurova 7, 166 29 Prague 6, Czech Republic ∗ corresponding author: jan.trejbal@fsv.cvut.cz Abstract. Asphalt mixtures often fail due to poor interaction between mineral aggregate and bituminous binder. Therefore, many efforts are being made on adhesion improvement between the two materials. In this work, paving grade bitumen 50/70 was doped with two types of adhesion promoters. Asphalt mixtures composed of crushed aggregate Brant coated by binder were made and exposed to stripping water. Then, they were subjected to visual and digital image analysis aiming to quantify aggregate residual bitumen-coated areas. Besides, two cylindrical samples of aggregate were bonded together by a thin film of bitumen doped with adhesion promoters. After solidification of binder, force needed for separation of the two parts was measured. It was shown that residual bitumen-coated areas were increased by 20-30% by adding adhesion promoters into binder. Mechanical adhesion of thus modified binder to aggregate was increased by up to twice. Keywords: Bituminous binder, adhesion, adhesion promoters, pull-off tests, image analysis. 1. Introduction Asphalt mixtures used in construction of road pave- ments are composite materials consisting of two main phases: mineral aggregate and bituminous binder. Adhesion between them is the key parameter which is responsible for road mechanical properties as well as lifetime. If adhesion is insufficient, the two mate- rials can be disintegrated. Consequently, pavement structure is thus damaged with potholes [1]. Physico- chemical interaction between binder and aggregate is complex process based on adhesive bonds. These are influenced by bitumen chemical composition (es- pecially presence of polar components), temperature of materials used, and surface properties of aggre- gate [2, 3]. It has to be also taken into account that adhesion between binder and aggregate, alongside the param- eters mentioned above, is strongly influenced by wa- ter. Most of aggregates used for production of road pavements are composed of hydrophilic acid-based minerals like granite and quartz. These exhibit strong affinity for water, higher than for bitumen [4, 5]. Given the fact that aggregate of the most frequented hot-mix asphalt mixtures is during production of as- phalt mixtures dried and preheated to high temper- atures often exceeding 150 °C [6, 7], the presence of internal water (water contained in row aggregate) can be practically excluded. However, the exposition to external water, ie., water entering pavements because of poor drainage or highly wet subgrade is unavoidable. Consequently, bitumen is thus replaced by water due to its low interaction with aggregate, followed by strip- ping effect – an adhesion failure between bituminous material and aggregate surface [4, 5]. The whole process of so-called asphalt moisture damaged is illustrated in Figure 1. The figure a) shows theoretical ideal state: aggregate is perfectly coated by thin film of bitumen and is thus protected against water penetration. However, the real state is depicted in figure b) where the continuous layer of bitumen is damaged due to either poor coating process or abrasion against other aggregate particles. Because of traffic loads and phenomenons described earlier, water gradually penetrates the aggregate through the bitumen coat defect and thus separates the bitumen thin film from aggregate surface [8, 9]. adhesion lost a) b) c) aggregatebitumen coat water Figure 1. Aggregate coated by bitumen thin film, a) ideal state, b) real state, and c) stripping due to water. To avoid stripping, different types of adhesion pro- moters at the form of liquid additives can be added into bituminous binder. The role of such chemical surfactants is to maximize the formation of inter- molecular interactions between bitumen and aggre- gate surface [4, 5] and to reduce aggregate hydrophilic- ity. These are usually based on alkylsilanes, phospho- 112 https://doi.org/10.14311/APP.2020.26.0112 https://ojs.cvut.cz/ojs/index.php/app vol. 26/2020 Adhesion between mineral aggregate and bituminous binder Mark Binder Adhesion promoter Admixture amount [wt.%] R 50/70 - 0 M1 50/70 unsaturated fatty acids with diethanolamine 0.2 M2 50/70 phosphorus-based 0.3 Table 1. Summarization of three variants of bituminous binder 50/70 used. rus, chemical surfactants comprising amines, or un- saturated fatty acids combined with diethanolamine. Their amount differed from 0.2 to 0.4% of binder weight. The choice of a suitable additive depends on the origin of used aggregate, the type of binder, and temperature of asphalt mixture preparation [10, 11]. With using of adhesion promoters, a new question about their efficiency arises: How do they improve adhesion between the two materials? Generally speak- ing, two evaluating basic methods are known [12]: (i) loose asphalt mixture tests and (ii) mechanical anal- ysis. The first of them is based on subjective visual or digital image analysis [13, 14]. Asphalt mixture is made according to relevant technical standard [6, 15] and then exposed to hot stripping water or abrasion for certain period. After that, the stripped aggregate areas are quantified and assessed. These methods are indirect and it is therefore clear that adhesion between the two materials can not be quantified from the me- chanical point of view. The second mentioned method – mechanical analysis – is then divided into two sub- categories as indirect and direct. Indirect testing is based on mechanical behavior evaluation of asphalt mixture specimens. In recent technical practice, we are talking primarily about tensile strength tests of prismatic or cylindrical asphalt mixture specimens. The durability and resistance of compacted specimens is checked before and after water exposition [16]. It is obvious that the indirect testing can be influenced by many variable factors, for example by mixture inhomogeneity, aggregate shape, etc. The all shortcomings listed above can be over- come using the direct mechanical adhesion analysis. Aggregate-bitumen bonds can be determined by means of peel or pull-off tests. In case of the first mentioned, polymeric or metal belt is bonded using bitumen as the adhesive thickness-controlled layer on preheated pris- matic sample of aggregate. After cooling of the speci- men and solidification of bitumen, the belt is peeled off under angle of 90° (to aggregate surface plane), while the fracture development is observed [17, 18]. Pull-of tests are based on the same general princi- ples. Thin layer of bitumen is sandwiched between two cylindrical aggregate samples. One part of aggre- gate is after technological pause pulled off, aiming to evaluate interfacial bond strength between the two materials [18]. Such a method seems to be the most effective of all mentioned, because it combines only two main materials and provides the most valuable data from the perspective of interface interaction. Although the pull-off method is not widespread in the Czech Republic, we decided to use it in order to determine bond strength between mineral aggregate and differently modified bituminous binder. To extend such obtained results, we made asphalt mixtures and subjected them to standard loose asphalt mixture tests. 2. Materials 2.1. Aggregate and bituminous binder For purposes of this study, aggregate from Brant quarry was chosen. It is composed primarily of granite porphyry, or porphyry microgranite (quartz, potas- sium feldspar, and mica rocks), therefore it is hy- drophilic and thus more susceptible to water dam- age, with poor adhesiveness to binder and elevated potential of stripping failures. Its surface can be char- acterized as porous and its composition as relatively homogeneous. Paving grade bitumen 50/70 was used, standardized by [19]. 2.2. Adhesion promoters To ensure stronger adhesive bonds between both mate- rials, two types of liquid-form chemical-based additives were used. These were applied at the trace amount of 0.2 and 0.3wt.% of bitumen according to recom- mendations of their manufacturers. Reference and two modified binder variants were used, as summa- rized in Table 1. In the case of mixture marked as M1, reactive products of unsaturated fatty acids with diethanolamine were used. The additive can be char- acterized as brown viscous fluid, which is recovered by condensation of unsaturated fatty acids at high temperature with disposal of water. The acids are dis- soluble in ethanol and acetone and can be emulsified in water. In the bituminous binders, such additives are soluble at standard blending/mixing temperatures. The additive according to the producer shall be re- sistant to long-term heating (150 hours) at 150 to 160 °C without any changes in its activity. The second mixture, M2, was doped with a new generation of phosphorus-based adhesion promoters. This type is specific due to versatile application and improved ther- mal stability allowing to resist higher temperatures. 2.3. Sample preparation Cylindrical specimens having 42±0.2mm in diameter were drilled from block of aggregate using table grill equipped with diamond tipped tile drill bit. Then, such obtained samples were cut off to 25mm length using saw Struers Secotom-15 with diamond cutting 113 J. Trejbal, T. Valentová, V. Nežerka, P. Semerák Acta Polytechnica CTU Proceedings blade to ensure perfectly planar cut, perpendicular to longitudinal axis of the specimens. Two thus obtained specimens were then preheated to 150±5 °C and sand- wiched with 1mm thick layer of the bitumen with the same temperature. The thickness of the bitumen layer was ensured by applying its properly weighed amount. After that, the sandwiched specimens were tempered to 15±1 °C. According to [6], crushed aggregate of fraction 8/16mm (in sum ca. 300 g, temperature 160±5 °C) was coated by the bituminous binder (12±0.3 g, tem- perature 170±5 °C). After 24 hours of samples stor- age, these were submerged into stripping water having 60 °C for following 60±5 minutes. Then, they were subjected to visual and digital image analysis as de- scribed in separate paragraph below. 3. Experimental methods 3.1. Pull-off tests Sandwiched cylindrical specimens, 5 pieces for each mixture, were fixed to clamping chuck composed from two mirror half circles. A space between the chucks and testing specimens was filled with rubber seal in order to ensure strong adhesion between them. Then, thus fixed specimen was placed to steel cage ensuring its direct guiding during loading. The cage is movable in the direction of specimen axis. It is equipped with tensile sticks from the both sides. These are used for fixing the whole construction in a loading frame, as captured in Figure 2. Sandwiched specimens were loaded through the construction by tension using a loading frame MTS Alliance RT/30. The experiment was displacement controlled at the constant rate of 0.5mm/min. Figure 2. Construction for pull-off tests fixed in loading frame. In the middle, there is a sandwiched specimen placed. 3.2. Visual and image analysis Following [6], the conventional visual analysis was based on subjective assessment of asphalt mixtures by two evaluators, who classified the rate of stripped areas of aggregate. In order to minimize "human factor" and inaccurate evaluators assessment, semi-automatic digital image analysis was employed. Asphalt mixtures were cap- tured using DSLR camera Canon EOS 6D (full-frame sensor) equipped with lens Canon EF 100mm f/2.8L Macro IS USM. Such device allowed to capture even insignificant details being necessary for proper assess- ment. The samples were lighted by daylight only. Consequently, thus obtained images were subjected to evaluating software enabling entropy-based image seg- mentation. Thanks to local entropy calculation, the software is capable to assess roughness of the texture. The areas with high entropy are considered to belong to aggregates while low entropy areas represent the bituminous matrix [14]. 4. Results and discussion 4.1. Pull-off tests Force as a function of displacement obtained during pull-off testing is for all three types of sandwiched spec- imens shown in Figure 3. All curves were calculated as arithmetic mean from 5 independent measurements. A significant difference was detected in comparison of specimens with modified and reference bitumen. Both modified specimens (M1 and M2) showed bond- ing strength approximately twice as higher as those containing reference binder (R). Moreover, reference samples showed perceptibly less gradual increase in recorded force in initial stages of loading. Comparing both samples sandwiched with doped bitumen to each other and considering measurement deviations, it can not be concluded which adhesion promoter is better. 4.2. Visual and image analysis Visual analysis of two independent operators revealed that bitumen surface coated area reached on 60% only in the case of the reference mixture (R), with belonging verbal classification as "unsatisfactory" ac- cording to [6]. Unlike both modified mixtures (M1 and M2) exhibited the area about 80% – "satisfactory" classification (see Table 2). Entropy-based image segmentation, however, pro- vided quite different results than visual analysis. Con- cretely, reference mixture (R) exhibited 67.1% of coated areas, while M1 and M2 94.7 and 96.0%, re- spectively (Table 2). Given the fact that the semi- automatic analysis is considered to be more accurate than the subjective one, such obtained results are as- sumed to be more valuable and conclusive. Moreover, as it is shown in Figure 4, the green colored regions, marking the coated areas, prove that the amount of stripped areas is negligible. Regardless to differences between results from both methods, it can be concluded that adhesion promoters significantly contributed for improving in adhesion between mineral aggregate and bituminous binder. 114 vol. 26/2020 Adhesion between mineral aggregate and bituminous binder Fo rc e [N ] Displacement [mm] 0 500 100 200 300 400 0 2.01.51.00.5 R M1 M2 Figure 3. Force as a function of displacement revealed during pull-off experiment. Mark Coated area by Classification by Coated area by Classification by visual analysis (%) visual analysis DIC analysis (%) DIC analysis R 60 D-E: unsatisfactory 67.1 D: unsatisfactory M1 80 C: satisfactory 94.7 A: good M2 80 C: satisfactory 96.0 A: good Table 2. Summarization of binder coated area results obtained by both visual and digital image (DIC) analysis. Figure 4. Images of aggregate coated by binder (top) and highlighted coated areas as a result form DIC analysis (bottom). 115 J. Trejbal, T. Valentová, V. Nežerka, P. Semerák Acta Polytechnica CTU Proceedings 5. Conclusions Adhesion between paving grade bitumen 50/70 and mineral aggregate from Brant quarry was analyzed by means of visual, digital image, and mechanical analysis. In order to increase an interaction between the two materials, bitumen was doped with adhesion promoters based on unsaturated fatty acids with di- ethanolamine and phosphorus-based at the amount of 0.2 and 0.3wt.% of binder. Asphalt mixtures composed from crushed aggregate (8-16mm) coated by binder were exposed to strip- ping water. Consequently, areas coated by binder were quantified. Then, the adhesion was analyzed mechanically using pull-off tests, when two cylindrical parts of aggregate were sandwiched with 1mm layer of bitumen binder and consequently pulled-off. It was shown that adhesion between the two mate- rials was increased using adhesion promoters by ca. 20 and 30% according to visual and digital image analysis, respectively. Pull-off tests revealed that me- chanical adhesion between aggregate and modified bitumen was increased by up to twice. Acknowledgements This work was supported by the Czech Science Founda- tion (research project No. 18-13830S) and by the Czech Technical University (No. SGS19/148/OHK1/3T/11) and the support is gratefully acknowledged. Special thanks to Tereza Horová for editing assistance and to Dr. Petr Bittnar for helping with experimental testing. References [1] N. Ahmad. Asphalt mixture moisture sensitivity evaluation using surface energy parameters – PhD thesis. The University of Nottingham, 2011. [2] D. Lesuer. The colloidal structure of bitumen: Consequences on the rheology and on the mechanisms of bitumen modification. Advances in Colloid and Interface Science 145(1-2):42–82, 2009. doi:10.1016/j.cis.2008.08.011. [3] C. Rossi, B. Teltayev, R. Angelico. Adhesion promoters in bituminous road materials: A review. Applied science 7(524), 2017. doi:10.3390/app7050524. [4] Y. Gao, Y. Zhang, F. Gu, et al. Impact of minerals and water on bitumen-mineral adhesion and debonding behaviours using molecular dynamics simulations. Construction and Building Materials 171:214–222, 2018. doi:10.1016/j.conbuildmat.2018.03.136. [5] A. Sevilla, M. Montero, J. Aguiar-Moya, et al. Influence of bitumen and aggregate polarity on interfacial adhesion. Road Materials and Pavement Design 18:304–317, 2017. doi:10.1080/14680629.2017.1304265. [6] čsn en 76 6161: Determination for adhesion of asphaltic binders to aggregate, Praha: Czech Office for Standards, Metrology and Testing, 2000. [7] H. Sivilevičius, S. Šukevičius. Manufacturing technologies and dynamics of hot-mix asphalt mixture production. Journal of Civil Engineering and Management 15(2):169–179, 2009. doi:10.3846/1392-3730.2009.15.169-179. [8] W. Wang, L. Wang, H. Xiong, R. Luo. A review and perspective for research on moisture damage in asphalt pavement induced by dynamic pore water pressure. Construction and Building Materials 204:631–642, 2019. doi:10.1016/j.conbuildmat.2019.01.167. [9] A. Mehrara, A. Khodaii. A review of state of the art on stripping phenomenon in asphalt concrete. Construction and Building Materials 38:423–442, 2013. doi:10.1016/j.conbuildmat.2012.08.033. [10] J. Trejbal, P. Demo, V. Nežerka, et al. Adhesion improvement between bitumen and mineral aggregate: theoretical principles and their impact on pavement structure lifetime. In Conference Proceeding of 10th aniversary International conference on Nanomaterials - Research and Applications. Tanger, Ltd., Ostrava, 2018. [11] J. Trejbal, Z. Prošek, T. Valentová. Assessment of adhesion between mineral aggregate and bituminous binder using digital image analysis. In Acta Polytechnica CTU proceedings, Nano and Micro mechanics conference. Czech Technical University in Prague, 2018. doi:doi:10.14311/APP.2018.15.0126. [12] S. Amelian, S. Abtahi, S. Hejazi. Assessment of adhesion between mineral aggregate and bituminous binder using digital image analysis. Construction and Building Materials 63(30):294–302, 2014. doi:10.1016/j.conbuildmat.2014.04.012. [13] J. Trejbal, T. Valentová, T. Horová. Assessment of adhesion between bituminous binder and mineral aggregate using digital image analysis. In Acta Polytechnica CTU proceedings, Contemporary materials and technologies in civil engineering 2018. Czech Technical University in Prague, 2018. doi:doi:10.14311/APP.2019.21.0038. [14] V. Nežerka, J. Trejbal. Assessment of aggregate- bitumen coverage using entropy-based image segmentation. Road Materials and Pavement Design pp. 1–12, 2019. doi:10.1080/14680629.2019.1605304. [15] čsn en 12697-11: Bituminous mixtures – Test methods for hot mix asphalt Part 11: Determination of the affinity between aggregate and bitumen, Praha: Czech Office for Standards, Metrology and Testing, 2006. [16] J. Grönninger, M. Wistuba, P. Renken. Adhesion in bitumen-aggregate-systems. Road materials and pavement design 11(4):881–898, 2010. doi:10.1080/14680629.2010.9690311. [17] S. Cui, B. Blackman, A. Kinloch, A. Taylor. Durability of asphalt mixtures: Effect of aggregate type and adhesion promoters. International Journal of Adhesion and Adhesives 54:100–111, 2014. doi:10.1016/j.ijadhadh.2014.05.009. [18] J. Zhang, A. Apeagyei, G. Airey, J. Grenfell. Influence of aggregate mineralogical composition on water resistance of aggregate–bitumen adhesion. International Journal of Adhesion and Adhesives 62:45–54, 2015. doi:10.1016/j.ijadhadh.2015.06.012. [19] čsn en 12591: Bitumen and bituminous binders – Specifications for paving grade bitumens, Praha: Czech Office for Standards, Metrology and Testing, 2009. 116 https://doi.org/10.1016/j.cis.2008.08.011 https://doi.org/10.3390/app7050524 https://doi.org/10.1016/j.conbuildmat.2018.03.136 https://doi.org/10.1080/14680629.2017.1304265 https://doi.org/10.3846/1392-3730.2009.15.169-179 https://doi.org/10.1016/j.conbuildmat.2019.01.167 https://doi.org/10.1016/j.conbuildmat.2012.08.033 https://doi.org/doi:10.14311/APP.2018.15.0126 https://doi.org/10.1016/j.conbuildmat.2014.04.012 https://doi.org/doi:10.14311/APP.2019.21.0038 https://doi.org/10.1080/14680629.2019.1605304 https://doi.org/10.1080/14680629.2010.9690311 https://doi.org/10.1016/j.ijadhadh.2014.05.009 https://doi.org/10.1016/j.ijadhadh.2015.06.012 Acta Polytechnica CTU Proceedings 26:112–116, 2020 1 Introduction 2 Materials 2.1 Aggregate and bituminous binder 2.2 Adhesion promoters 2.3 Sample preparation 3 Experimental methods 3.1 Pull-off tests 3.2 Visual and image analysis 4 Results and discussion 4.1 Pull-off tests 4.2 Visual and image analysis 5 Conclusions Acknowledgements References