Reverse flow

[Sravan B]

Hello,

I am simulating a battery pack cooled by PCM (Phase Change Material) and liquid cooling. I aim to implement a delayed boundary condition where the velocity inlet boundary condition is activated only when the PCM liquid fraction reaches 0.6. For the liquid cooling setup, I have applied velocity inlet and pressure outlet boundary conditions. To achieve the delayed boundary condition, I used expressions to define the boundary condition.

The issue I am facing is that the velocity is initially set to zero, which leads to reverse flow occurring in the channel—a behavior that should not happen. I need to ensure that no flow occurs in the channel until the velocity inlet boundary condition is triggered. I have also tried using an outflow boundary condition instead of a pressure outlet, but the problem persists. Additionally, I experimented with mass flow inlet and mass flow outlet boundary conditions, but the issue remains unresolved.

Could someone please help me in solving this issue?

Thank you!

 

[Rob]

How much incomig mass are you seeing? 

[Sravan B]

Inlet mass flow rate is 0.008kg/s. 

If you are asking about how much reverse flow is occuring, in the console it is showing reverse flow in 65% area due to pressure outlet.

[Rob]

Does material density change with temperature? 

[Sravan B]

For PCM, density changes with temperature. For liquid(Water is used), everything is constant.

[Rob]

OK, so if the PCM changes volume (that's another consideration as it can't move) where does the other material (liquids) go? 

[Sravan B]

 

Liquid flow through channels of aluminium plate. 

Battery is surrounded by PCM, it takes out heat from the battery. Beside PCM, plate is placed. Liquid flowing channels of plate takes out heat from PCM.

Initially, inlet velocity for liquid is given zero. I wrote an expression when PCM liquid fraction reaches 0.6, inlet velocity boundary condition changes to 1m/s. 

There is reverse flow occuring when inlet velocity is zero.

 

[Rob]

Is the PCM in direct contact with the liquid circuit? 

[Sravan B]

No. There is plate which absorbs heat from PCM and heat from plate is taken by liquid flowing through channel.

[Rob]

Thanks. So, two separate flow channels. 

For the PCM, if density changes where does the extra volume go (or come from)? 

For the liquid flow boundaries how much volume is lost, or how fast is the flow? It could be convergence related. We'd also tend not to use zero flow velocity boundaries; equally swapping from wall to inlet isn't always trivial.