Closed Loop Conjugate Heat Transfer Simulation — How to Set Up Forced Flow Without Inlet/Outlet

Using SimScale Fluid Flow / CFD
1

Hi SimScale community,

I’m a student working on a project called Project ARCTIS — a simplified magnetocaloric refrigeration system using a Halbach array and a Gadolinium (Gd) magnetocaloric bed. I’m trying to run a Conjugate Heat Transfer IBM simulation to generate a temperature distribution heat map across the assembly for a student research submission.

System description:

  • Closed loop fluid system — water circulates continuously through a serpentine channel inside the Gd bed, through connecting stub pipes, through heat exchanger channels, and back through a curved return pipe

  • No true inlet or outlet — the loop is continuous

  • Solid bodies: Gadolinium bed, Copper HX plates, Aluminium housing, Steel pipes

  • Fluid: Water

What I’m trying to achieve: Show heat transfer from the Gd bed surface (fixed at 295.93 K) into the circulating water and through the surrounding solid assembly — producing a colour heat map showing temperature distribution across all components.

What I have successfully done:

  • Uploaded STEP file from Fusion 360

  • Selected CHT IBM as simulation type

  • Successfully extracted Internal Flow Volume — SimScale correctly detected the entire connected fluid path as one body

  • Assigned materials to all solid bodies

  • Set Fixed Temperature boundary condition on Gd bed surface (295.93 K)

  • Set gravity, initial conditions, numerics and simulation control

  • Generated mesh and obtained a result using stagnant water

The core problem: Because my system is a closed loop, I cannot use velocity inlet and pressure outlet boundary conditions — SimScale flags the faces as internal and ignores them. I want to simulate forced flow at approximately 0.2 m/s to show realistic heat transport from the Gd bed into the water and through the pipes, rather than just conductive diffusion from stagnant water.

What I have tried:

  • Pipe split to create open ends for inlet/outlet — SimScale still flags faces as internal

  • Average Velocity momentum source across entire flow volume — uncertain about direction for complex serpentine geometry with multiple flow directions

  • Fan momentum source — same problem with direction for complex geometry

My specific questions:

  1. What is the correct SimScale approach for forced flow in a closed loop CHT simulation?

  2. Is momentum source the right tool, and if so how should direction be configured for a multi-directional serpentine geometry?

  3. Is there a better simulation type for this specific case that still captures fluid-solid heat transfer across the full assembly?

Thank you in advance.