FLUG REVUE July 2006: Compressor technology advances

July 2006

HIGH-PRESSURE COMPRESSORS – THE HEART OF THE ENGINE

By Patrick Hoeveler

Art or science? The development of a high-pressure compressor (HPC) is both. The HPC has for all times required the greatest development effort of any engine component. As Dr. Erich Steinhardt, Senior Vice President Technology at MTU Aero Engines, explains, “The compressor is the heart of the engine. The high-pressure compressor has the job of achieving the highest possible pressure ratio. The greater the exit pressure from the compressor, the more efficient the circular process in the engine.” It is very important to have a good efficiency rate, as this has a huge impact on fuel consumption. “A one percent increase in the efficiency rate alone cuts fuel consumption by 0.6 percent.”

A modern compression system must cover every possible permutation of engine use from idle to full throttle and at the same time it must be as small as possible with as few stages as possible. “We have to achieve the same building height but with fewer steps,” is how Dr. Steinhardt explains the challenge that designers face. A new compressor always carries a risk. Failure to achieve the specification values can have unfortunate consequences for the entire programme, as occurred, for example with the initially five-stage HPC in the Pratt & Whitney PW6000, such as extended delays and customers abandoning ship. Moreover, compressor development is not cheap: it costs an average of $100 million and accounts for up to 18 percent of the respective costs of a powerplant.

Despite this, following the success of the HDV12 developed jointly with Pratt & Whitney, MTU has embarked on the development of a new compressor for the next Airbus and Boeing narrow-body families. The German company is responsible for the first four stages and the Americans for the rear four. The compressor, which will have a pressure ratio of 17:1, is expected to be running in Munich by the end of the year (see FLUG REVUE 4/2006). According to Dr. Steinhardt, the engineers have had to accept a compromise between efficiency rate and number of stages. “Up to now reducing the maintenance costs to a minimum has been the priority. But the recent hike in oil prices brought with it a change of emphasis. Fuel consumption and hence efficiency rate are becoming more important again.” At the same time, there is a requirement for a low number of stages and hence higher stage pressure ratios. These in turn require higher rotational speed and hence higher flow speeds, causing increased aerodynamic losses. “With three or four stages more, we would have a better efficiency rate, says Dr. Steinhardt. “But weight continues to play an important role.”

The use of blisk (Bladed Disk) technology, in which the discs and blades are integrated into a single piece of metal, has become established as a virtual standard. Normally they have fewer, but larger blades and are lighter than multi-part components. In the past it was very difficult to manufacture them at a reasonable cost. “But today blisks are cheaper to manufacture than multi-part blade and disc assemblies,” says Steinhardt. On the MTU high-pressure compressor for the Advanced Technology Fan Integrator (ATFI) demonstrator, all seven stages consist of blisks.

Materials offer the prospect of further improvements. Here, for example, the Munich engineers are working on new alloys using titanium and nickel bases. Blades constructed out of carbon fibre composites have already been tested, but due to their inadequate temperature resistance they are only suitable for low-pressure compressors. Metal-matrix composites (MMC), such as titanium strengthened with silicon carbide fibres, have the advantage of low weight combined with high strength. This technology can be used to manufacture bladed rings, known as “blings”. But a new problem then arises, namely the need to reconcile disparate thermal expansion and rigidity, which in turn means higher development and production costs. This technology is therefore only worthwhile for military products which have to be very compact. In the medium term, Dr. Steinhardt does not see any demand for it, though it might conceivably be employed on a possible successor to the Eurofighter's present EJ200 engine.

Computer-aided design, which permits ever more accurate simulations, has become an indispensable part of the development process. “We only properly understood the PW6000 compressor after we had included the return flow effects.” Here he is referring to air which flows out of the cavities and labyrinths under the stators back into the main flow, and influences the compressor behaviour. “Good tools are necessary in this area. In the last few years computer technology has made a huge leap forward possible.” But there is still some way to go. The designers currently calculate the transient dynamics of each stage on its own. “In the future we will be able to calculate the transient compressor dynamics in their entirety all at once. At present, this would cause the computing time to explode.”

Despite this, the predictions obtained on the computer are already proving very accurate, as Dr. Steinhardt explains: “On the ATFI compressor, the discrepancies in the efficiency rate were of the order of 0.5 percent and in the surge margin one to two percent.” The new methods also allow the casing to be modified to extend the surge margin. This is known as “casing treatment”. Vortices which lead to flow separation and hence aerodynamic losses develop on the blade edges. A casing treated in this way has specially formed openings which direct the air back in the form of a jet. The jet acts on the blade edge and reduces the vortices by influencing the boundary layer. In this way modification of the casing can extend the surge margin by up to 15 percent, while also offering the possibility of dispensing partly or wholly with variable guide vanes. This approach has already been tried on two stages of the ATFI compressor. The next step, according to Dr. Steinhardt, will be to design future compressors for casing treatment, rather than vice versa as has been done up to now.

He expects the use of intelligent systems such as active surge control to produce an even greater leap forward. Here, an injection of air could prevent compressor surge. Tests in a low-pressure compressor on a Larzac-04 turbofan have already delivered promising results. Clearance control would also be actively adjustable. To avoid aerodynamic losses, the gap between blades and casing must be kept as small and even as possible. Normally this clearance lies between 0.1 and 0.5 mm. In view of the enormous speed at which the blade tips rotate, at over 1600km/h, under no circumstances must the rotor strike or touch the casing.

“Optimisation of the clearance is an art, as there are so many factors that have to be taken into account. The gap changes with rotational speed and temperature.” The casing is not as inert as the rotor and accordingly expands at a different rate. This behaviour can be influenced by air flowing out of the rear stages or by leading the air for the air-conditioning system along the compressor casing. Research is also under way into a mechanical means of control, such as using bimetal effects. In any case such active systems require small but robust sensors. In this area, further development work lies ahead of the engineers.

All in all, the effort required to produce ever smaller improvements, which have their limits due to the laws of physics, is growing, as Dr. Steinhardt explains. “On large civil engines the efficiency rate cannot be raised significantly above 90% due to the many losses. The upper limit on the pressure ratio is 23:1 and with a single-stage high-pressure turbine it is not much more than 13:1. There is not much point in trying to get any more because of the circular process.” The trend towards engines operating without bleed air is not making the development work any easier, as Dr. Steinhardt points out. “The challenge in this case is proving greater, as one variable is missing, namely the possibility of bleeding air to improve compressor stability. The efficiency rate does not necessarily have to be better. The effectiveness depends on the aircraft application.” With or without bleed air, a new compressor can today be designed in almost record time thanks to advances in computer technology. The development time up to the first tests is usually only a year. In the old days it often took twice as long and required a lot more build versions.

From page 80 of FLUG REVUE 7/2006

Last updated 8 June 2006

FLUG REVUE, Ubierstr. 83, 53173 Bonn, Germany