Cyber-physical systems (CPS) are complex systems that integrate computation algorithms, communication networks and physical systems. Thus intrinsically making many of these systems heterogeneous. Using formal models of computation to model these systems is an increasing practice that enables early verification at the design phase.  However, a current challenging problem is to analyze and verify timing at the initial stages, the design is still conceptual and agreements have to be made for system functionality requirements at the expense of a practical architecture that would be able to implement the functionality. To solve this problem, we present a framework that separates the architecture modeling and functions. The framework also addresses the heterogeneity aspects of CPS design. The Functional Mock-up Interface (FMI) standard is rapidly becoming an industry standard for model exchange and co-simulation. The proposed framework utilizes the FMI standard to handle the interactions between the functional model and architectural model and utilizes the SysML language to represent the systems interactions and to auto generate Functional Mock-up Units (FMUs). Three case studies on automotive systems are used to demonstrate the effectiveness of our approach.