Learning Forum

[greg.harrington]

I have a heated plate inside a box. The box enclosing the plate has a thickness of 0.055 inch and is cooled by natural convection on all 6 sides. The box is defined as the air volume around the plate offset by 50mm
If I set the HTC on all 6 sides of the box created by the air volume to a constant value like 5 W/m-K (a standard value for these problems) the model runs just fine.
If I set the HTC to "Use correlation" to and use 4 vertical sides, 1 top and 1 bottom then the model doesnt run and gives the error: "Solver initilization failed"

For the solver settings:
Flow Regime: Laminar
Include Gravity: ON
Radiation Model: OFF
X Velocity Initial condition: 0.01 m/s (Direction opposite gravity)
Use Model Based Flow Initialization: ON

Pressure Under-Relaxation: 0.70, Body Force Discretization scheme
Momentum Under-Relaxation: 0.30

All other solver settings are default.

 

 

[escalindamy]

The "Solver initialization failed" error you are encountering when using the "Use correlation" option for the heat transfer coefficient (HTC) may be due to the complexity of the flow regime or the specific boundary conditions set for the model. Here are a few suggestions to troubleshoot this issue:

1. Check Boundary Conditions: Ensure that the boundary conditions for the bitlife correlation method are correctly defined and compatible with the laminar flow regime you are using.

2. Adjust Initial Conditions: The initial velocity condition of 0.01 m/s might be too high for a laminar flow scenario. Consider reducing this value or setting it to zero to see if it helps with initialization.

3. Relaxation Factors: The momentum under-relaxation factor of 0.30 may be too low. Try increasing it to 0.5 or higher to facilitate convergence.

4. Flow Regime: If the flow is expected to be turbulent in certain areas, consider switching to a turbulent flow model or adjusting the parameters to reflect the actual flow characteristics more accurately.

5. Mesh Quality: Ensure that the mesh quality is sufficient and that there are no skewed elements that could affect the solver's ability to initialize.

6. Solver Settings: Experiment with different solver settings, particularly those related to initialization and convergence criteria.

By carefully reviewing these aspects, you may be able to resolve the initialization failure and successfully run the model with the correlation-based HTC.