Design of Experiments (DOE) applied to high-alumina calcium aluminate cement-bonded castables

Abstract

Using statistical methods (such as Design of Experiments – DOE) is an important issue due to the increasing demand to reduce costs and optimize properties of different processes and products. In order to assess the feasibility of these techniques in the refractory field, the Response Surface Method (RSM) and Tukey test were applied to optimize the modulus of rupture of high-alumina calcium aluminate cement (CAC)-bonded castables after curing and drying, considering three continuous factors (the amount of water, cement and dispersant) and a categorical one (the drying process). In addition to the cold mechanical strength, the flowability and apparent porosity of all the castable samples were also measured. The Tukey test pointed out the composition with the highest strength after curing and after drying. Based on the attained RSM models, the first one (considering all four factors) indicated the composition with the highest modulus of rupture after curing, but it could not properly predict the castables' properties after drying. The other two RSM models analysed different processing conditions (post curing and post drying, separately) highlighting that the complexity of the drying stage related to the hydration of the cement phases invalidated the models used for the post drying condition. The major difficulty in using statistical techniques comes from the transient and non-equilibrium conditions experienced by refractory ceramic castables during their processing and working conditions.