Evaluation of air content on concrete compressive strength with classical approach and neural networks

Mnogi parametri utiču na karakteristike svežeg i očvrslog betona. Jedan od najvažnijih svakako jeste poroznost betona. Poroznost svežeg betona meri se količinom uvučenog vazduha. Uticaj uvučenog vazduha na beton i njegovu pritisnu čvrstoću istraživali su mnogi svetski istraživači. Na bazi tih istraživanja, izdvojene su dve, u praksi najviše korišćene, formule. Tehnike mekog programiranja, a posebno neuronske mreže, s formiranim bazama podataka laboratorijskih ispitivanja betona, otvaraju novi pristup u predviđanju uticaja količine uvučenog vazduha u svežem betonu na njegovu čvrstoću pri pritisku.Many parameters influence on the characteristics of fresh and hardened concrete. One of the most important characteristic of concrete is its porosity. Measure the porosity of fresh concrete is measured by the amount of entrained air. The effect of entrained air in concrete on compressive strength investigated by many authors. On the bases these works we have two formulas applicable in practice. Soft programming techniques especialy neural networks and the formation of databases related to the testing in laboratories for concrete opened up new approaches in predicting the impact of the quantity of entrained air in concrete on compressive strength