OPTIMIZING MIX PROPORTIONS OF REACTIVE POWDER
CONCRETE USING GROUP METHOD OF DATA HANDLING
AND GENETIC PROGRAMMING
32nd Our World in Concrete and Structures (OWICs) - 2007
Ali Haghighi*, University of Tehran, Tehran, Iran
Mohammad Reza Koohkan, University of Tehran, Tehran, Iran
Mohammad Shekarchizadeh University of Tehran, Tehran, Iran
Genetic Programming, Group Method of Data Handling, super plasticizer, silica fume, steel micro
fiber, Reactive Powder Concrete
In this paper, the effect of super plasticizer amount, water/cement ratio, the cement,
silica fume and steel micro fiber content is studied on the ultimate compressive strength of new generation of Ultra High Strength Concrete (UHSC).Known as Reactive Powder Concrete (RPC) and also have used two non-linear functions; Genetic Programming (GP) and Group Method of Data Handling (GMDH), which are two methods in soft computing family. In this kind of concrete, the strength can reach to 200-800 (MPa) depending on the curing type and the manufacturing process. The emphasis is on the RPC which the strength ranges of 200-350 (MPa).
Testing more than 28 different mix proportions and the result shows that GMDH is
much exacter than the other methods. Finally sensitivity analysis has been showed the shares of each used parameters in the best model.