Learning Forum

[mehdi Abbas]

The above photo shows residual values for a transient simulation. I am confused as to why the continuity residuals are not falling . This is an enclosed box containing battery cells with time dependent heat generation and natural convection. There's an additional heater whose source values are modeled with a udf using heating table values. The step size for each Time step is 1 second. The mesh quality parameters are 0.55 max skewness and minimum orthogonal quality equal to 0.3. I am unsure why the continuity solution is not converging, I am using K omega sst viscous model. Theres no fluid entering or leaving the box and the gaps for the fluid are in millimeters.Can someone help me as to how to get this solution to converge?

[Murari Iyengar]

Can you provide more details about the Boundary Conditions as well? At the moment, it isn't very clear. 

[mehdi Abbas]

The base of the box has adiabatic boundary conditions whereas the remaining five walls have a convective boundary condition. The system is initially at room temperature.1

[mehdi Abbas]

 

 

[Murari Iyengar]

You mentioned there is no fluid leaving/entering the box, then what are the gaps for? Can you share screenshots of your geometry? 

[mehdi Abbas]

This is a battery pack with cells. The volume of the box not occupied by the cells is fluid . Since the pack is closed off from the surroundings, the fluid volume doesn't change.

[Murari Iyengar]

If the fluid inside the box is present without any change in fluid flow, you can just assign convection boundary conditions to the cells. Why solve flow equation?

[mehdi Abbas]

The cells lose heat to the surrounding fluid in the pack . I am trying to simulate thermal runaway propagation in a battery pack. Thus it is important to solve the flow equations

[Murari Iyengar]

If you don't have fluid leaving/entering the domain, then your net mass flux should be 0. Is that the case? 

[mehdi Abbas]

I haven't checked the flux . Yes it should be.

 

[Murari Iyengar]

Are you using Body Force Weighted or PRESTO! as the Spatial Discretization scheme for Pressure? If not, I recommend doing so. 

[mehdi Abbas]

Yes I am using PRESTO

[Murari Iyengar]

Can you share an image of the mesh cross section? Are the mesh cells refined near the battery cells? Apart from continuity, what do the other results look like?

[mehdi Abbas]

I did not draw boundary layers