Zoom in the Lab

Micro-turbines, maxi-difficulties

This combustion chamber only measures 20 mm in diameter and is 2.7 mm thick. Supplied by a hydrogen-air mixture, it produces a power of up to 1200 W.

Future miniature drones must be supplied by miniature energy sources. The best adapted seem to be micro-turbines. However, designing of turbines measuring a few cubic centimetres is a bit chancy. Researchers are looking into flow modification, the design of new geometries, micro-manufacturing and thermal studies.

 
Imagine drones the size of a bird, capable of flying for hours, of filming scenes and transmitting information. Many engineers are bent on designing such miniature drones, for civilian and military applications. One of the many difficulties is supplying energy to the engines, which must be both powerful and also extremely light. Batteries are too heavy and have too little independence for these micro-drones with wingspans of 15 centimetres and similar length, weighing around a hundred grams. As far as fuel cells are concerned, these do no yet exist in the range of power under research. That leaves gas turbines remain, which could provide ten times more energy than a battery with the same mass.

A turbine transforms energy from fuel into rotational motion, either to directly supply a propeller, or to produce electricity. "Our goal is to make a micro-gas-turbine, for future micro-drones", says Joël Guidez. These micro-turbines will supply the electric motor driving the drone wings, as well as the electrical equipment, such as transducers, and even a small camera.

However, to miniaturize a turbine, it is not enough to reduce the dimensions of each component. The flows do not occur in the same way at very small scales, for example in the combustion chambers of these micro-turbines, which only measure a few hundred cubic millimetres. The flows are much less turbulent, and gases thus mix with greater difficulty. This is not desirable for a combustion chamber, where the fuel must mix with air! It is thus necessary to create structures that favour the mixture of the gases within the combustion chamber. "We design circulation areas leading hot gases toward cool gases", explains Joël Guidez. This allows the combustion to be maintained, otherwise it would be extinguished.

  
The elsA aerodynamic simulation code allows the characteristics to be optimized by revealing details of the flows between the vanes.

Meanwhile, these microscopic structures can not be too complex, or it may not be possible to manufacture them. The first chamber constructed had a quite simple geometry: It was a cylinder with a 20 millimetre diameter, 2.7 mm tall, having a tube in its centre off which the gas bounces. It is thus sent around the periphery of the chamber, where it mixes with the gases present. This chamber serves above all to test the manufacturing and measuring methods of ONERA's laboratory. The manufacture of a second, more complex and better performing chamber is underway.

To speak of small volume combustion chambers is to speak of great thermal losses. Indeed, small objects have a greater surface in relation to their volume compared to large objects, which generates more thermal losses. This is both an advantage and a disadvantage. On one hand, it prevents the walls from overheating and melting, but on the other, it may extinguish the combustion if too significant. It is thus necessary to design chambers for which the losses are just at the right level.

 

	The micro-combustion-chamber and its experimental enclosure. The aerothermal simulation code Cedre allows the 3D combustion to be reproduced and to better understand the phenomena in play.

Current experiments are being carried out using hydrogen as fuel for several reasons. Since it is very light, it diffuses easily. Additionally, its chemical reaction time (the necessary time for the combustion chemical reaction to take place) is 50 microseconds, ten times shorter than that of the hydrocarbons that are usually used. Also, in a micro-combustion-chamber, the time it takes the gases to cross the combustion chamber is very brief. The more the dimensions of the chamber are reduced, the more this time is reduced. However, it must remain as five times greater than the chemical reaction time, without which combustion would be poor. On the other hand, a hydrocarbon is easier to store than hydrogen. Near future studies shall therefore be directed towards the hydrocarbon combustion stability within these very small chambers.

Which materials will be used to make these micro-turbines? The Massachusetts Institute of Technology (MIT) is endeavouring to manufacture a turbine entirely from silicon, in order to benefit from the silicon etching technologies of micro-electronics to create channels which will allow the gases to mix. ONERA prefers to manufacture micro-turbines made up of several materials.

Micro-turbine rotating parts etched on silicon (SilMach).

Finally, let us not forget the rest of the gas turbine, specifically the rotating parts. The turbine vanes can be made to spin at a very great speed, up to a million revolutions per minute. The manufacture of 8mm turbine vanes can prove to be complex. Silicon etching may be a solution, unless micro-machining techniques are favoured. "This manufacture will require all sorts of specific technologies", says Joël Guidez. But the crucial question is that of the bearings and stops, which hold the turbine shaft. In the usual turbines, ball bearings serve this purpose. Here, hydrodynamic bearings would be used: the gas flows sustain the rotating parts without these touching the fixed parts.

Micro-drones are not the only potential applications of micro-turbines, which combine a high rated power and a very small size. The power supply for portable devices could also benefit from this, for example, to equip the future infantryman, transporting increasingly more electronic equipment. Nevertheless, the extremely hot gases must be evacuated. ONERA is allowing itself one year to examine the feasibility of gas micro-turbines, and in the event of reaching favourable conclusions, it wishes to manufacture a prototype between now and 2010.

Last Update: October 29, 2007 - © ONERA 2009 - Terms of use