A digital twin is a combination of different representations of a technological product and its development is a complex endeavor that requires expertise from many different domains and employs several different strategies. It can include everything from system-level models to full 3D models. Each of these models fulfills a specific need and answers a specific set of questions. During the design cycle, each team needs to communicate its results with other teams. Typically, the model-based systems engineers and 3D analysts work in a disjointed manner where data sharing occurs through file exchange. This approach to data sharing is frequently manual (i.e., via USB, email, etc.), which is difficult to scale, prone to error, and less conducive to seamless teamwork. In the increasingly interconnected systems developed today, it is more important for experts to collaborate more easily, particularly within multidisciplinary teams. This presentation will show how co-simulation between system and 3D simulation software represents an efficient solution to enable direct communication between the system and 3D models, thereby allowing seamless data transfer. The presentation deals with the case of a mechanically driven loop heat pipe modeled both in Simcenter 3D Thermal Multiphysics and Simcenter Amesim. Modeling heat pipes can be very complex due to the inherent two-phase nature of the heat transfer. While systems-level models and correlations can be used for modeling heat pipes, the true three-dimensional nature of the heat transfer is not captured. Conversely, modeling the process fully in 3D can be both cumbersome and time consuming. In this presentation, we demonstrate a way to connect the system and 3D models of the loop heat pipe for leveraging the strengths of both solvers. Comparison with experimental data and different design configuration will be discussed. We will also describe the challenges inherent in modeling physics with very different time constants and the strategies we employed to mitigate them.