Mixed factor/component problems
Design of experiments in formulation: optimization

We have seen in the previous paragraphs of this article and in article that the formulator is faced with both general planning problems (such as factor screening, multi-criteria optimization of a response surface, etc.) and specific ones (compositional variables inducing an experimental domain in the form of a triangular diagram, existence of various individual or relational constraints). The most complex case is when the problem involves both independent factors, such as temperature, processing time or any other manufacturing factor, and compositional variables of a formula. We will consider two different approaches.