Purpose The purpose of this paper is to model the corrosion rate behavior for two ferrous materials, carbon steel AISI 1020 and stainless steel AISI 304, immersed in ferric sulfate and ferric chloride solutions using D-optimal design with response surface methodology. Design/methodology/approach Experimental design addresses two factors (concentration and contact time) with multilevel categories, in order to predict and compare the corrosion rates of the studied materials immersed in flocculants solutions. A corrosion rate of specimens was calculated from mass loss determinations. Findings The authors used a polynomial model to fit the experimental values, thereby predicting significantly higher corrosion rates in ferric chloride solutions, as compared to ferric sulfate. Originality/value The authors propose a high fidelity model of the corrosion rate of each carbon steel and stainless steel material using D-optimal design with a response surface method (RSM).