Computers Cooled in Mineral Oil
A recent study by Intel using Green Revolution Cooling's CarnotJet System described immersing computer servers in mineral oil for electronics cooling. Of course the question that springs immediately to mind is why would you want to submerge sensitive electronics in a liquid?
It all boils down to a fluid's specific heat capacity (Cp). Cp is a measure of how much energy is required to raise the temperature of a fluid. The higher the Cp the better for cooling applications. You want a cooling fluid that can absorb energy and transport it away (through convection) from temperature sensitive electronics.
The specific heat capacity of typical cooling fluids at room temperature is:
- 4183 J/kg.K for water
- 1670 J/kg.K for mineral oil
- 1005 J/kg.K for air
Clearly water appears to be an excellent candidate as a cooling fluid, with a much higher Cp than air and mineral oil. It is for this reason that water serves as the basis for many mechanical cooling applications. However, for electrical components, water has a problem - it is an electrical conductor. So when air cooling is no longer adequate or cost effective for electronics, then mineral oil, being an electrical insulator, is a good alternative.
Another advantage of using a liquid (mineral oil) coolant over a gas (air) coolant is that the liquid will typically be denser (800x for mineral oil relative to air) and therefore able to transport energy away at a much reduced flow rate.
Intel's study confirms what custom computer system builders have known for a while, i.e., that mineral oil can serve as an excellent substitute for air cooling without adversely affecting the immersed electronic components.
Check out a Caedium CFD simulation example where we compare air, mineral oil, and water (virtually of course!) as fluids for an electronics cooling application.
Recent blog posts
- External Aerodynamics with CFD
- A Case for Renaming the Navier-Stokes Equations
- Ludwig Prandtl: Real Fluids Explained
- Osborne Reynolds: A Giant in Fluid Dynamics
- Sliver Treatment Strategies for CFD
- How to Fix Small Acute Angles for CFD
- Small Feature Removal for CFD
- Fluid Visualization in Nature
- Design is Compromise
- CFD Is Not Enough