Segmentos experimentais na BR 116/RS empregando TLAFlex, HiMA e 55/75-E (SBS): caracterização avançada e monitoramento

Abstract

Due to increase in heavy traffic, allied to department of transportation’s budgets restrictions, municipalities are seeking for long lasting pavements that perform better over time. Despite the increase in initial cost, modified binders proved to be an economically-viable alternative in long term evaluations. In this context, mechanistic-empirical methods emerge in several countries as a way to validate the high-quality materials use. Looking forward to evaluate the influence of modified binders on the asphalt materials behavior, three test sections were executed it BR 116/RS, using different asphalt mixtures in the surface layer: one containing Styrene-Butadiene-Styrene (SBS) polymer, AMP 55/75-E (PGH 64 V); one highly modified by SBS, Stylink HiMA (PGH 70 E); and another modified by Trinidad Lake Asphalt added to SBS, TLAFlex (PGH 70 H). The binder materials were characterized according to Superpave methodology, rheology, MSCR and LAS tests, suggesting a better rutting behavior for HiMA binder, and worst for AMP 55/75-E, and for fatigue, better for HiMA binder and worst for CAP 50-70 (used as asphalt base layer). The asphalt mixtures, collected during the pavement execution, were characterized with the tests of Resilient Modulus, Complex Modulus, Flow Number, Diametral Compression Fatigue and Uniaxial Cyclic Fatigue. Due to inconsistences in executed binder content, the correlation of the stiffness, rutting and fatigue between binder and mixture scales were not effective. In spite of that, all mixture had a great performance related to rutting. For fatigue, while the HIMA mixture (with Stylink HiMA) exhibited the best characteristics, the SBS (with AMP 55/75-E binder), with asphalt content 1,30% lower than dosed, had the worst. In parallel to this, since the opening to two years of opening to traffic, functional surveys of macro and microtexture, roughness, rutting and cracked area were conducted every semester in the highway. It was identified presence of water, mainly in SEG 01 (TLAF), which resulted in high levels of rutting. The evaluations using extracted reduced samples were promising, strengthening this methodology as a way to verify the field execution. Although the modified binders could not be directly compared due to divergences between laboratory controlled and field conditions, comparison between the segments was carried out separately. In this framework, the research conducted numerical simulations of the pavement structures using MeDiNa v.1.1.5.0 and LVECD Beta 1.1, with Nascimento (2015, 2021) transfer-functions, using retro-analysis before traffic opening, with wet subgrade material, and with the thickness obtained through field extractions. The rutting in the wet subgrade segment was assertively predicted by MeDiNa when using a material tested in the Wótima+4% condition. In relation to predicted cracked area for both programs simulations, it was concluded that among the retroanalyzed conditions, the LVECD, added with the updated transfer function, was the most assertive. Finally, when the triaxial conditions of the support layers were used in MeDiNa, it returned the prediction closest to the equality line.