A variety of engineering problems involve dynamic simulation and optimization, but exhibit a mixture of continuous and discrete behavior. Such hybrid continuous/discrete behavior can cause failure in traditional methods; theoretical and numerical treatments designed for smooth models may break down. Recently it has been observed that, for a number of operational problems, such hybrid continuous/discrete behavior can be accurately modeled using a nonsmooth differential-algebraic equations (DAEs) framework, now possessing a foundational well-posedness theory and a computationally relevant sensitivity theory. Numerical implementations that scale efficiently for large-scale problems are possible for nonsmooth DAEs. Moreover, this modeling approach avoids undesirable properties typical in other frameworks (e.g., hybrid automata); in this modeling paradigm, extraneous (unphysical) variables are often avoided, unphysical behaviors (e.g., Zeno phenomena) from modeling abstractions are not prevalent, and a priori knowledge of the evolution of the physical system (e.g., phase changes experienced in a flash process execution) is not needed. To illustrate this nonsmooth modeling paradigm, thermodynamic phase changes in a simple, but widely applicable flash process are modeled using nonsmooth DAEs.