"Dipole moment" is a measure of the electrical polarity of a system of charges. Consider two molecules, one of ethyl alcohol and the other of di-methyl ether:
Ethyl alcohol has an obviously asymmetric distribution of charge, but the charge distribution of di-methyl ether is more symmetric. Therefore the dipole moment of di-methyl ether is very low while the dipole moment of ethyl alcohol is larger.
Most molecules have some non-zero dipole moment and that is very much true for petro-chemical oils that are suitable for dielectric service in high voltage applications. There is a consequence of this fact.
Consider a vessel filled with an oil and into that vessel, we place two electrodes between which we may apply a high voltage.
When the high voltage is not applied, the oil is at rest and exhibits some particular fluid density with the molecules of that fluid arranging themselves in the usual close packed fashion.
However, when we apply a sufficiently high voltage, the electrostatic field of that high voltage causes the oil molecules that are positioned between the two electrodes to rotate into alignment with that field so that those molecules are no longer in the physically close packed arrangement that they had been in when we started. The density of the altered fluid between the electrodes is reduced.
However, fluid outside of the electrodes is still as dense as it was before and thus pressure from that outside fluid causes an upward displacement of the lower density fluid which then spills over the top and regains its original density now that it is no longer subject to the electrostatic field.
The fluid at the bottom that is newly introduced to the electrostatic field undergoes its own density reduction and a cycle of flow and density modulation is set up. The fluid circulates.
Do bear in mind that none of this would be happening if the oil molecules had zero dipole moment; the molecules would not line up in response to the electrostatic field.
The vigor of this circulation can be difficult to predict. In the case of one high voltage assembly delivering 150 kV, there were two oils on hand, one from Texaco and the other from Shell.
Memory fails me as to which did which, but one of them circulated at some moderate rate while the other one circulated so energetically that there was danger of oil being thrown right out of the vessel.
A patent, number 3,267,859, "Liquid Dielectric Pump", was granted to Sakari T. Julia for the United States Air Force in 1966 to apply this process as a fluid pump with no moving parts for use in heat removal.
Additional investigations of this idea were later done at Oklahoma State University. The material is covered in the textbook: Dielectrophoresis, The Behavior Of Neutral Matter In Nonuniform Electric Fields, by Herbert A. Pohl, Professor of Physics, Oklahoma State University, Cambridge University Press, Copyright 1978.
The fluid pump idea never "took off" though. In my view, the development of thermoelectric coolers based on the Peltier effect offered more easily constructed cooling systems with no moving parts.
This is very interesting. Thank you.
Do you have any information on the efficiency of such pumps? How large are the power losses, in comparison with conventional mechanisms for similar heat transfer rates?
Posted by: Ophir | December 24, 2010 at 11:39 AM
Just curious, but does this phenomena require DC or will the molecules flip back and forth fast enough for it to work with AC too? I'm thinking oil-submerged 60Hz power transformers...
Posted by: Brad Peeters | December 28, 2010 at 02:16 PM
Hi, Brad.
I've only looked at this effect for DC. However, having the molecules pivot around in a few mSec doesn't sound all that far fetched.
Maybe a Google search would yield some applicable literature.
Posted by: John D. | December 28, 2010 at 03:39 PM
Thanks to share this interesting phenomenon. Does anybody have information about how the energy conservation law can applies to it? I mean, considering the oil viscosity, some energy is lost, and then, some current should be drawn from the voltage source... But how this current can move through the isolating oil and how it perform its work? (Considering the amount of polarized molecule remain constant...)
Posted by: Remi Demers | December 30, 2010 at 12:03 PM