Experiment with mixtures is a special type of experiment in which the response depends only on the proportions of input variables x = (x₁, �, x_q)� for mixture experiment, not on the total amount of the ingredients. When a response variable may not be adequately represented by a single polynomial function of the input variable over the entire experimental region. A possible solution is to use a regression model which consists of grafted polynomial submodels, named as segmented polynomial regression. In this paper, we study a special segmented mixture model in which the contribution of an ingredient proportion to the response variable, for example, x₁, is different from the others and its D-optimal design on the whole experimental region.