Experimental investigation of mechanical properties of hybrid fiber reinforced concrete samples and prediction of energy absorption capacity of beams by fuzzy-genetic model

• Hybrid fibers effects were exhibited on prismatic and cylindrical samples.
• Hybrid fiber addition increased compressive strength of cylindrical samples.
• Hybrid fiber addition increased toughness and ductility of prismatic samples.
• S1.50/P0.15 and S2.00/P0.20 additions are suggested as the optimum fiber addition.
• Fuzzy-genetic model showed a satisfactory performance in predicting data.

There have been a great number of studies aiming to improve the mechanical qualities of concrete material which is a basic component of reinforced concrete bearing systems. Attempts to reinforce the concrete through fiber addition have an important place among these studies. This study is an experimental research on the mechanical features of hybrid fiber added concretes which are obtained using steel fibers and polypropylene (PP) fibers together in certain proportions. Test specimens consist of 180 cylindrical samples (150 × 300 mm) and 90 prismatic (150 × 150 × 750 mm) beams produced in C40 concrete class. The specimens were exposed to temperatures ranging from room temperature to 100, 200, 400, 600 and 800 °C in a certain order following their cure periods of 7, 28 and 90 days. It was observed that hybrid fiber addition has a significant contribution into the compressive strength, flexural strength and energy absorption capacity of the concrete. Significance of hybrid fiber addition was put forward for the increase in ductility of the concrete which displays a brittle behavior when forced to bending.In the experimental studies with a numerous test specimens, losing of data due to some reasons and being unable to receive data was considered as a problem to be solved. Therefore, a fuzzy-genetic model was suggested to predict not only the experimental gaps, but also the untested value ranges of experimental parameters. Model was taken consideration in three different applications and it showed a satisfactory performance in predicting data.