Prediction and Optimization of Heavy Clay Products Quality - Advanced Materials for Agriculture, Food and Environmental Safety - page 87

4

Prediction and Optimization of Heavy
Clay Products Quality

Milica Arsenović 1, *, Lato Pezo 2 , Lidija Mančić 3 and Zagorka Radojević 1

1 Institute for testing of materials IMS, Belgrade, Serbia
2 University of Belgrade, Institute of general and physical chemistry, Belgrade, Serbia
3 Institute of Technical Sciences of SASA, Belgrade, Serbia

Abstract
h e ef ects of chemical composition, i ring temperature (800-1100 C), and sev-

eral shape formats of laboratory brick samples on the i nal product quality were
investigated. Prediction of the i nal laboratory products parameters was evaluated
by second order polynomial regression models (SOPs) and artii cial neural net-
works (ANNs), and at erwards compared to experimental results. SOPs showed
high r 2 values (0.897-0.913 for compressive strength models, 0.942-0.962 for
water absorption, 0.928 for i ring shrinkage, 0.988-0.991 for water loss during i r-
ing, and 0.941 for volume mass of cubes models). An ANN model, coupled with
sensitivity analysis, was obtained with high prediction accuracy: 0.866-0.939 for
compressive strength models, 0.954-0.974 for water absorption, 0.882 for i ring
shrinkage, 0.982-0.988 for water loss during i ring, and 0.920 for volume mass of
cubes models. h e optimal samples' chemical composition and i ring temperature
were chosen depending on a i nal usage of the raw material in heavy clay brick
industry.

°

Keywords : Heavy clay products, prediction, optimization

4.1 Introduction

Heavy clays are used as the main raw material in the brick-making indus-
try due to their specii c properties. Final properties of clay brick are