Is this the right way?

Using SimScale Fluid Flow / CFD
1

Hi everyone,

I am a student using SimScale for the first time to analyze fluid flow and fan performance inside a high-density electronics housing, which is structured similarly to a large computer chassis or a blade server rack.

[System Configuration]

  • External Inlet: 8 Fans (Max Static Pressure: 75 Pa each)

  • External Outlet: 4 Fans (Max Static Pressure: 180 Pa each)

  • Internal Recirculation Fans (as Momentum Sources):

    • 1st Floor (Bottom, cooling specific heat sources): Four 30 Pa fans + Two 40 Pa fans

    • 2nd Floor: Four 30 Pa fans

[My Simulation Setup]

  1. Created an Internal Flow Domain and extracted the Fluid Region.

    image
    image1274×1142 87.8 KB

  2. Assigned standard Inlet and Outlet boundary conditions to the external ventilation faces.

    image
    image1669×1179 75.2 KB

  3. Modeled the internal fans using CAD solid cylinders and assigned them as Momentum Sources (Fan Model with P-Q Curves).

  4. image
    image2560×1440 314 KB

image
image933×1145 325 KB

[The Problem] The simulation results do not match my theoretical expectations. It seems like the internal fans are either not being recognized by the solver or underperforming drastically. The volumetric flow rate through the internal fans drops to just 15% - 40% of their rated specifications. Surprisingly, when I ran a baseline test with the internal fans completely disabled, the overall velocity profiles and pressure distributions looked almost identical to the run where the fans were enabled.

[My Hypotheses]

  • Geometric/Meshing Issue: The CAD cylinders representing the internal fans have slight geometric overlaps. Could this cause the mesher to merge the volumes into the fluid domain, thereby neglecting the Momentum Source inputs? (Should I use Face-based sources instead?)

  • Aerodynamic Phenomenon (Stall): Since the external pull (180 Pa outlets) is significantly stronger than the internal fans, a strong negative pressure fields develops inside the enclosure. Are the internal fans simply experiencing a severe aerodynamic stall (choking) due to excessive system resistance?

Could anyone point out if there are flaws in my setup, or share similar validation models/public projects (e.g., server rack cooling, chassis flow balancing) that I can refer to?

Thank you so much for your help!