Research into product modularity has created methods to partition a product system in to modules, including DSM algorithms to minimize various within- or between-module connectivity assumptions, as well as heuristic methods for combining functions into common modules, including dominant flow, convert-transmit, repeated elements, and branching flows. We here re-examine these methods in view of flows between field potentials. Fields are spatially defined scalar functions representing temperature, pressure, voltage, etc., each with associated flows such as heat, fluid, current, etc. It is hypothesized that isolation of elements to desired field values can form a physical basis for module definition. Product functional descriptions were examined from the literature. Those found sufficiently detailed with function structures, module definitions, part lists and subassembly definitions were studied here. Within these examples, there were 183 functions grouped into 51 modules. Of these, a statistically significant 67% of the modules had boundaries which isolated a field. For example, all elements within the module were at a high temperature and all elements outside the module were at a low temperature. Such agreement between actual modularity and field isolation provides evidence that an effective module definition strategy is to use field boundaries to separate into modules the necessarily high and low field values in the product structure. A second analysis considered how desired flows are designed to cross field boundaries. In 84% of the cases of flows crossing field boundaries, specific field separation functions were defined. Care was taken through specific functionality provision to ensure field boundary isolation. In summary, we find containing fields within a product can form a physics based guideline for defining product modularity.