Closed Loop Water Cooling System Model

[ta1234]

Hi Everyone,

I would like to set up a thermodynamic/heat transfer model of the following system. The system is a closed system which cools an electronic device which is attached to a heat sink, the water circulates through tubes between the heat sink and a cold plate, rejecting heat to the cold plate. The outer surface of the cold plate will be kept at a constant temperature. I know the heat load output, cold plate outer wall temperature, device temperature limit and the wetted surface areas of the heat sink and cold plate.

I would like to generate a model which will allow me to adjust parameters such as heat sink and cold plate wetted area and surface temperature, giving me a required mass flowrate of water. I have come up with a couple of models but I still have not enough equations for the number of unknown variables.

The main thing that I'm finding quite difficult is that using the Dittus Boelter equation, the convection coefficient for a pipe wall is dependent on fluid flowrate (Link). Does anyone know a simplified way of calculating pipe convection coefficients which avoids dependency on flowrate?

Can anyone provide me with some pointers for a good way to solve this problem? I'm not totally sure where I should be drawing my system boundaries? I'm not looking for anyone to solve it for me, just a method and perhaps which heat balances to use for my system.

Many thanks in advance.

 

[MortenA]

The point do Dittus Boleter is that there IS a dependecy of heat transfer on flow velocity! If you want to be conservative use the lowest flowrate that you may operate at and solve for this once - and then use the transfer coefficient you get from this calc.

Best regards, Morten

 

[ta1234]

Hi Morten,

Thanks for your quick reply.

Ok in that case I'd be simplifying things too far if I didn't take into account the dependence on mass flowrate in the equations.

I'm designing the system from scratch so I can accommodate whatever flowrate I want, within practical limits obviously.

I'd like to end up with a system of simultaneous equations that I can use as a model for the cooling system. Has anyone ever set up this type of thing before? The page here (Link) is very similar to my system, except in place of the air/water heat exchanger I have a heat sink, with one side kept cold at a known temperature.

Many thanks.