Development and optimization of roller compacted coloured concrete with high albedo to reduce urban heat island impact

Urban heat island effect is the difference between air temperature of urban area and the nearest rural area. Industrialization and urbanisation cause urban heat island effect. Within the scope of this doctoral thesis, pigment-added roller compacted concrete samples were produced to reduce the urban heat island effect caused by asphalt pavements. Within the scope of preliminary experiments, concrete samples with different color pigment additives were produced and Albedo determination test was carried out. The highest reflective property was determined in the samples containing yellow pigment up to 0.75% by weight of the cement. In the next stages, only yellow pigment was used. The samples produced are R0 without pigment, P0.25, P0.5, P0.75 and P1 with pigment. The consistency test, compressive strength test, splitting tensile strength test, reflectance determination test, sulphate resistance test, cantabro abrasion tests were performed on the produced yellow pigment added roller compacted concrete (RCC) samples. Primarily, it was aimed that the concrete samples produced with pigment additives would obtain results that comply with the relevant standards and the studies in the literature, and this target was achieved. 28 days of compressive strength of samples were deterimed to be 34.78, 34.95, 35.04, 35.06 and 35.07 Mpa in the samples respectively R0, P0.25, P0.5, P0.75 and P1.0.After the test results obtained, the albedo value of the sample with the highest reflectivity was added to the simulation program library. Two areas were selected from Konya , the use of asphalt as a road pavement material and the use of pigment-added RCC in this area were analyzed. In the scenario where pigment added roller compacted concrete was utilized as road pavement material revealed that utilization of this material can decrease the air temperature 1.18-1.8 oC , increase the relative humidity rate 5.88% and 10.08%, and decrease the coating surface temperature 8.06 oC – 13.11 oC. The experimental results obtained were utilized to do prediction via machine algorithms and the highest correlation number was determined to be 0.97 in the random forest algorithm, and the algorithm with the lowest performance was support vector machines. Then, the amount of pigment was kept constant in the applied design mixture and the surface response method was used to optimize the compressive strength. The optimized design mixture was tested in the laboratory environment.