1. Mixture optimization. Response surfaces

In the article , the first part of this presentation, we have demonstrated the use of more qualitative than quantitative techniques, whose conclusions are generally the starting point for a more detailed study. We are now going to develop quantitative techniques for building predictive models of the behavior of [23] mixtures, or for finding a compromise between several more or less contradictory characteristics (responses). It is very often the case that the formulator has to take account of different experimental responses, optimize some of them, keep others within the limits imposed by a standard, and so on.

Response surface methodology (RSM) applied to independent factors has been developed in numerous works [3]