Evaluation of the effects of hydrated lime addition to bituminous hot mixtures by fuzzy logic method

Deterioration of highway pavements is one of the primary problems faced in highway construction and operation. Principal leading factor to this problem is increasing traffic volumes and axle loads. Because of this problem, which is generally not taken into account in the highway design phase, shortening in the service lives of the roads is experienced. This is undesirable for both users of and authorities responsible from the highways. Nowadays, many different methods are being attempted to overcome this problem. One of the methods proposed is modification of the bituminous mixtures. A variety of additives can be used for the modification process. Usage of hydrated lime as an additive in hot bituminous mixtures is one of the foregoing hot mixture modification methods. Up to now, numerous studies have been conducted to determine the positive and negative effects of hydrated lime on the mixture and the subject is still widely being studied. Effects of any additive on the mixture depend on not only the type but also the amount of the additive. In this study, usage of hydrated lime as an additive to bituminous hot mixtures used in highway pavements is investigated. The effects of the amount of hydrated lime additive are examined in a wide range with high sensitivity intervals. In the first part, a comprehensive literature review about the usage of hydrated lime to solve different problems with highway pavements, as well as its benefits, methods of addition and amount of usage of it is presented. In the second part, experimental materials are introduced and results of the material property determination tests are presented. Three primary materials used in this study are aggregates, binder and hydrated lime (as additive). Aggregate comes from the limestone quarries in Ömerli, İstanbul. The aggregate, which is obtained from İsfalt Inc. Ümraniye asphalt production plant, is mostly grey – dark grey colored. Particle size distribution analysis of the aggregate is done according to the Highway Technical Specifications Wear – Type – 1 mixture percentages and it is found to be 51,3% coarse aggregate, 41,9% fine aggregate and 6,8% filler. Bitumen from Tüpraş refinery with production code AC 50-70 is used as asphalt cement. The additive (hydrated lime) is a commercial product of Öztüre holding Inc. Bartın lime factory, with production code S-KK 80-T. In the next stage, to assess the effect of the amount of hydrate lime addition on the mixture, the amount of filler is reduced 0,5% by 0,5% by weight, coupled with the addition of exactly the same amount of hydrated lime. Hence, 15 different Marshall Design sets, each having a different hydrated lime content, are obtained. There are 18 specimens per Marshall Design set, adding up to 15x18 = 270 specimens. Among the various mixture parameters of those specimens, bulk specific gravity (Dp), Marshall Stability (MS), flow (F), air voids (Vh), voids filled with asphalt (VFA) and voids in mineral aggregates (VMA) are determined to calculate the optimal bitumen amount for every different Marshall Design. Since hydrated lime is varied with high sensitivity intervals, this study enables the effect intervals of hydrated lime amount and transitions between them to be accurately observed. In the following phase of the study, to confirm the bulk specific gravity, Marshall Stability, flow, air voids, voids filled with asphalt and voids in mineral aggregates values corresponding to optimal bitumen amounts, 45 specimens, 3 for every different Marshall Designs, are produced, using the calculated optimal bitumen amount. In the third part, the results are evaluated. For this, design results from the 270 specimens of Marshall Design and computational results from the 45 specimens, produced with the calculated respective optimal bitumen amounts the original designs, are compared. To model the relation between the design results and the computational results, a fuzzy logic model is developed, using the design results of 270 specimens for calibration. Model outcomes are then compared with both design results and computational results, each of them being obtained by using optimal bitumen content with its respective hydrated lime amount. It is seen that, model outcomes are within a tolerable range of design results. However, the error with respect to computational results is gets larger. It is thought that, this study will be beneficial, on the basis that it would shed light on future studies, which are likely to include further performance tests on the mixtures with optimal bitumen contents. Keywords: Bituminous hot mixture, Marshall test method, modifying additives, hydrated lime, Fuzzy logic