Level: Intermediate Job Function: Analyst / Scientist / Engineer Bernard Ennis, Principal Researcher, Tata Steel David Neal Hanlon, Scientific Fellow, Tata Steel Patricia Gobernado Hernandez, Research Scientist, Tata Steel
Best E-Poster Winner
Optimization of the development cycle for new product development is a challenge for R&D in the steel industry. Steel is produced using a long chain of interdependent process steps. Consequently, large volumes of experiments are required to fully explore process property relations.This is more critical for new, highly differentiated products for which understanding of the underlying physical metallurgy is incomplete and thus a-priori exclusion of factors not possible. In this context, the use of statistical design of experiments to shorten development timescale, maximize the output of experimental capacity and minimize material consumption is essential. In this paper, an assessment of the response of a complex steel product to annealing process is presented. Comparison is made between a classical full factorial (FF), six-factor design with two levels, and a novel definitive screening design (DSD). Results are evaluated and discussed based on the capability of each design to screen for significant variables. While both designs are able to identify the most relevant factors, DSD results in a significant reduction of experiments. For the DSD the experimental burden is four times smaller than that of the FF design.