An electrostatic fluid accelerator (EFA) is a device which pumps a fluid such as air without any moving parts. Instead of using rotating blades, as in a conventional propeller or in the turbine of an airbreathing jet engine, an EFA uses the Coulomb force from a high voltage electric field to accelerate electrically charged air molecules, a phenomenon studied in the academic discipline called electrohydrodynamics (EHD). Because air molecules are normally electrically neutral, not charged, the EFA has to create some charged molecules, or ions, first. Thus there are three basic steps in the fluid acceleration process: ionize air molecules, accelerate those charge carriers and, through ion-ion and ion-neutral collisions, push many more neutral molecules in a desired direction, and finally neutralize ions again to eliminate any net charge in the downstream flow. This principle is used for spacecraft propulsion in ion thrusters. The basic working principle has been understood for some time but only in recent years have seen developments in the design and manufacture of EFA devices that may allow them to find practical and economical applications, such as in micro-cooling of electronics components.
To understand how electrostatic fluid acceleration works it is necessary to review how air molecules are ionized and how those ions are then used to create thrust.
Giving a charge to air molecules is a process known as ionization. Ions are air molecules that have a net electrical charge. Air under normal circumstances has no net charge. Whenever a charge imbalance does occur, the natural attraction of positive and negatively charged ions tends to eventually cancel this charge out, as they attract and combine. Electric shock is an example of this, as is lightning.
It is possible to ionize air artificially, and there are many methods for doing so, as is done for example in arc welding and light bulbs. However many of the methods known to science do not operate in conditions which are conducive to everyday uses; for example, very high temperatures or very low pressures might be required. Or as in light bulbs, specialized materials and gases may be used and extraneous light and heat might be produced.
Because of these restrictions most applications of EFA have relied on a process known as the corona discharge, which has a number of attractive characteristics. It requires no exotic materials, temperatures, or pressures. It works using air at normal levels of humidity and at normal temperatures. It doesn't produce significant negative side effects such as heat or light. It also requires only fairly simple electrical principles in order to function, and uses only low electrical currents, making it relatively safe.
(For more information, see.)
Ionization is only the first step in electrostatic fluid acceleration. Once the ions are created they have to be used to generate a thrust. This process relies on the same electrodes and electric field as the corona process.
The ions that have been created have a net charge. Because of this charge they will be repelled from the electrode having the same charge and simultaneously attracted to the other electrode with the opposite charge. However, in between these electrodes are enormous numbers of neutral air molecules that have been unaffected by the ionization process. To reach the attracting electrode the ions must literally push their way through the mass of neutral molecules. In doing so, they tend to push those neutrals along with them; this is the process which results in a thrust.
It is important to note that because the ions are continuously propelled by the electric field they can continue to bump into and accelerate neutral ions the entire distance between the electrodes. This is possible because they are given potential energy by the electric field. The electric potential energy of the ions is converted to kinetic energy of the neutrals in each collision. This is the mechanism whereby electrical energy is used to do work by accelerating the neutral air. Some energy is also wasted of course, by slightly raising the temperature of the air and the electrodes, and increasing motion of the molecules in unwanted directions.
Once the ions reach the attracting electrode most of them will lose their charge, i.e., by gaining an electron from the electrode. The fraction of the ions that do not collide with the attracting electrode will tend to be drawn back (up stream) to the attracting electrode. This causes the EFA device to be, more or less, being "driven with a foot on the gas and a foot on the brake." An alternating driving voltage of the right frequency can, in principle, minimize this effect. The neutralized molecules may bounce off the attracting electrode in any random direction. The neutral molecules are not influenced by either electrode and thus their net flow is unaffected as they exit the EFA device.
Physicists and engineers have developed models for some aspects of corona ionization and fluid acceleration; but in general, due to its complexity there is no general-purpose model which can predict exactly what will happen under any given set of circumstances. For example, air temperature, humidity, electrode shape, and airflow all can affect the exact amount of energy required, the number of ions generated, etc. Because of these difficulties, developments in EFA have relied on experimentation more than modeling to fine-tune and refine ionization designs.
From this basic principle engineers have made a number of specializations and refinements to apply EFA to cooling applications. For example, see work done by Thorn Micro Technologies.
The Thorn Micro design is intended to be mounted directly on top of a conventional microprocessor, where it would produce downward airflow onto the heated upper surface of the microprocessor package.
A potential implementation of micro-cooling is to achieve an even more fundamental integration of cooling components and microprocessor. This next step would be to fabricate electrodes and airflow surfaces on the micron scale using the same techniques currently used to manufacture microprocessors themselves. This level of integration could lower manufacturing costs and potentially increase thermal efficiency beyond what an externally mounted fan or EFA could achieve.
Microelectromechanical systems (MEMS) approaches have been suggested. Research into this particular approach is being conducted at the University of Washington, with support from Intel and a small company Kronos Air Technologies (www.kronosati.com).
The basic physical principles involved in electrostatic fluid acceleration have been understood at least since the industrial revolution. Experimenters such as Lord Kelvin (see also Kelvin's thunderstorm) applied similar principles in their studies of electromagnetism. Even in the 21st century researchers and experimenters are trying to explore the practical applications of EFA, using more detailed analysis, as for the Plasma actuator.
An early idea which garnered popular attention but didn't reach the prototype development was the use of EFA to produce thrust for aircraft. The same basic principles used in a cooling application, at a larger scale, have been shown to provide sufficient thrust to provide some lift, and early experiments were encouraging.
“Ionocraft” devices, as they are popularly known, never achieved any practical use because the thrust generated was insufficient to lift much more than the (very lightweight) electrodes themselves. Including the power supply or any other significant cargo vastly exceeded the maximum achievable thrust. Today these devices can be built as science experiments. While impractical, lifters demonstrate in a dramatic fashion the simple physical principles involved in EFA.
Cooling applications, however, do not suffer from the same weight restrictions. Cooling for electronic components has had a series of developments as the need for thermal management increased along with the capability of modern CPUs. The earliest microprocessors in personal computers required no cooling apparatus at all, consuming very low power. Gradually as clock speeds and component density increased, heat sinks were added to the surfaces of CPUs, but the cooling was still only passive, relying on naturally occurring air convection.
However since the early 1990s high-performance CPUs such as found in typical desktop computers have required active cooling. This also includes secondary processors, such as graphics processors which also consume a large amount of power. The most common and inexpensive method of active cooling is to mount one or more conventional fans directly on the processors in conjunction with a large heat sink, and possibly one or more others elsewhere in the case of the computer to increase the overall airflow. Much larger computers have sometimes relied on more sophisticated active cooling techniques such as water or refrigerant -based methods.
Other than cooling, EFA has been considered for use in other applications. These mainly have included particulate removal (“air cleaning”) and dehumidification. These applications rely on electrostatic effects to assist in the collection and removal of particles in the air. Prior to recent developments, the airflow velocities and efficiencies of EFA pumps were too poor for consideration in cooling products. One reason for this was that the relatively simple geometries which enabled analytical studies were not good performers, and more sophisticated mathematical and experimental models were needed to improve their performance. The designs discussed above may be the first examples of this new wave of more sophisticated designs.
The typical cooling fan is limited by noise and wear due to their high-speed moving parts. They also may suffer from inefficiency resulting in higher power requirements and a given fan may have a limited range of uses because they operate at a fixed speed and have a fixed blade geometry – i.e., the amount of airflow cannot be varied. Turbulence introduced by fan blades is one of the leading causes of inefficiency and vibration.
Potential advantages of EFA cooling devices include improvement on all these issues; EFA fans produce no vibration and have no parts to wear out. Their power consumption and airflow are controllable electronically, allowing them to be run in an optimal fashion to maximize cooling and efficiency. In particular, EFA fans can produce flow which is fairly laminar, and the velocity profile of the flow can be controlled to a much greater extent than conventional fans. Control of the airflow may have the most important impact on cooling effectiveness. Electrostatically accelerated flows have been shown experimentally to alter the boundary layer along a fixed surface so that the heat transfer rate is increased, in some cases by more than twofold.
According to recent research papers, there are several factors which affect the outlet velocity of an EFA fan, all of which may be improved by future research and development.
The size of the EFA device limits the amount of airflow that can be generated. A possible improvement involves layering or stacking a series of EFA devices to increase the airflow. Existing limitations and focus areas in this approach are that larger device sizes are inconvenient, especially for smaller applications such as netbooks. There can be interference between the stages, where the attractor electrode of the first stage and the ionizing electrode of the next stage produce an unwanted corona effect, which can actually result in reverse airflow. Alternatively, sparking can also occur if the layers are placed too close together. It is believed that careful design of the geometry of the devices can improve on all these issues.
Higher velocity airflow can be generated if more ions are generated by the ionization process; a greater density of ions would mean that a greater number of neutral molecules are pushed along. However the maximum voltage which can be applied to the electrodes is limited by the breakdown strength of the air – too high a voltage would result in a spark, which is a sudden discharge of electrical energy (much like lightning), instead of a corona. To avoid this a larger number of electrodes are required. Again, finding the optimal geometry will be the key determinant of success, as too many electrodes too close together tend to diminish the corona effect.
Optimization of the energy efficiency requires optimization of the overall fluid flow, so that a maximal amount of electrical energy is used to produce kinetic energy of the flow in the right direction. One approach to “tune” the fluid flow is to use additional accelerating electrodes, strategically placed to draw the charged fluid in the desired direction. This concept also requires substantial development and testing.
Future research is likely to consider improvements in some or all of these areas as compared to the simple models. Finally, although as-yet unproven, the potential for MEMS-level integration may further reduce costs, increase electrical and thermal efficiency, and also increase reliability.
Other work remains in testing prototype devices to ascertain if they are ready for widespread commercial use. For example, a determination of the operating life of an EFA device would need to be made before they could be widely adopted. Long-term running effects might include collection of particles on the electrodes, with as-yet-unknown effects. Likewise, air humidity effects need to be fully considered.