GUY NORRIS / LOS ANGELES
Pulse detonation engines may be a long way off, but hold the promise of greater efficiency as well as power
During a brief period just over a decade ago, sleeping citizens of Los Angeles were woken on more than one early morning by mysterious "soundquakes", or sonic booms. No official explanation was ever given for the booms, which were monitored by hundreds of seismographs across southern California.
By tracing the supersonic passage of the sound waves as they swept across the coast and cities, seismologists agreed that something overhead seemed to be travelling inland at high velocity. When the US Defense Department and NASA reported they had nothing in the air on those days to cause such a phenomenon it fuelled growing speculation that some sort of classified new supersonic, or even hypersonic, aircraft was indeed either in development or service.
The 1990s also saw, coincidentally or not, several reports throughout the southern half of the USA of unusually shaped contrails. These apparently consisted of sequential circles of condensation like smoke rings, arranged axially around a central plume. The unusual "doughnuts-on-a-rope" shape gave rise to speculation that the new classified aircraft, whatever it was, could be using a hitherto little known propulsion concept known as a pulse detonation engine (PDE).
Whether true or false, the widespread and often wild speculation over the existence of classified aircraft and PDE powerplants probably helped promote a broader interest in pulse detonation concepts that today is manifested in research projects around the globe. There are PDE studies under way in the USA at NASA, the US Air Force Research Laboratories (AFRL), Office of Naval Research, Defense Advanced Research Projects Agency as well as within industry and academia. Parallel studies are also under way in several countries, including Belarus, Canada, France, India, Japan, Russia and Sweden. All were spurred by the simple and rugged PDE concept and the promise of harnessing the chemical energy content of a detonated fuel-oxidizer mixture. All are interested in the vast potential of developing an engine that, on the surface, appears to consist of little more than an intermittent explosion in a tube.
Although the PDE cycle is theoretically inherently simple (see box), the current state of research indicates that practical application could be a lot more challenging. Researchers in Japan, for example, have questioned the practicality of PDEs and whether they will pay off.
Is it viable?
Joe Doychak, project manager of PDE development technology at NASA Glenn Research Center (GRC) echoes these thoughts. "The question still remains about if it is a viable propulsion system. There are various aspects that have been demonstrated, but an entire propulsion system itself has never been developed. The fact is, that it is an unsteady device with gas-dynamic interactions through the inlet to the nozzle - or in other words the entire length of the engine. This means you don't know what you've got until you put the whole engine together."
NASA's GRC is in the forefront of US research into PDE concepts and was set to run the world's first known complete flying PDE testbed beneath a NASA Boeing F-15 when the project was recently cancelled. "That's caused us to focus on what we'd like to do next," says Doychak, who believes the most immediate step is to assess critical technologies across the whole engine. The testbed engine was a small-scale PDE built by Pratt & Whitney for the US Navy's now cancelled HyStrike hypersonic missile project. Although smaller than most potential applications of a PDE, it gave NASA a ready-made powerplant with which to start tests. Now, "even if we were given all the money we would need, it would still be slow starting out, and it could be around eight years or so before we get to a sensible configuration", he adds.
A big hurdle for PDE developers is a lack of specific applications. Since 1999 alone, PDEs have been proposed for supersonic reconnaissance vehicles, mini cruise missiles, long-range sensor platforms, low-cost unmanned air vehicles (UAV) and unmanned combat air vehicles (UCAV), single-stage-to-orbit launchers (SSTO), mini-satellite launch vehicles, combined cycle engines and even as afterburners in the bypass ducts of turbine engines. Yet, despite this myriad of ideas, no single application has emerged. "There is not really a customer out there today, and therein lies the question," says Doychak. Configurations differ depending on the application, yet no specific customer can adopt a PDE until the concept is somehow proven. "Whatever is successful as the first application will be responsible for the development of much of the science," says Doychak. "It's a real chicken and egg situation."
One specific opportunity that appears to hold promise is the SSTO and launch vehicle arena. Although key competing technologies such as rocket-based combined cycle (RBCC) and particularly turbine-based combined cycle (TBCC) have already achieved higher technology readiness levels, he believes that "some of these technologies are showing signs of plateauing. So it may be time to look at extremely new and potentially breakthrough, revolutionary propulsion concepts like PDE." Rocket engine applications have proved attractive to PDE studies mainly because hydrogen is an acceptable fuel and inlet operation, which troubles the air-breathing concept studies so much, is no longer an issue.
The rocket mode of operation is similar to the air breathing engine with ignition at the closed end. In rocket mode, however, the oxidizer also needs to be injected into the system periodically.
Early focus
As so much of PDE technology is in its infancy and the uncertainty levels correspondingly higher for each component, an early focus for NASA is expected to be the development of an integrated approach "to help get a good characterisation of what's going on", says Doychak. This would then be spun off into investigations of specific parts of the engine, starting at the inlet which is widely considered one of the most vital areas. Two main inlet questions appear to be uppermost: can a stable shock system be established in the inlet of a PDE and will the sudden closure of valves cause an inlet unstart - the surge-like problem that occurred on high Mach aircraft like the Lockheed SR-71. Preliminary tests at the University of Texas of a single inlet duct which supplies air to multiple PDEs have shown that a "shock trap" established with boundary layer bleed air could stabilise the shock in the inlet and give insufficient time for the formation of "hammershocks" - the destabilising effect which can trigger an unstart.
Tests are also being conducted on a rotary valve inlet for single and multitube combustors. This meters the flow to the combustor as well as isolating the inlet from the high pressures produced during detonation.
Investigators believe that by using multiple combustors that fill and detonate out of phase, all the inflowing air can be used. Firing rates of up to 12Hz per combustor have been shown in a hydrogen-fuelled system. Tests are proposed of a fluidic valve which has no moving parts. This work will involve Texas A&M University and others.
One of the potential problems of air-breathing PDEs is expected to be operations at high altitude, where low static pressure will influence the total pressure recovery of inlets and, therefore, the size of the bang from the propellents. As pressure drops, the mixture becomes less easy to detonate. To avoid making PDEs altitude limited, researchers are also examining ways of increasing back pressure. Work is expected to build in part on tests at Ben Gurion University of a supersonic inlet of PDE with a mechanism to excite back pressure. The university showed that, even by blocking up more than 80% of the exit and varying the amplitude of the excitation in the engine, pressure oscillations in the inlet were confined to an area downstream of the throat and no problems were found upstream.
After the inlet come the problems of mixing and igniting the fuel and oxidizer to make the biggest or most efficient detonation possible. Several investigations are under way at locations as diverse as universities in Moscow and California into the effectiveness of different shaped cavities, blockages and jet impingement to enhance fuel-air mixing. Not unexpectedly, perhaps, the best results during multi-cycle operations have been found when the fuel-air is pre-mixed before injection into the detonation chamber. This area is of critical importance to the entire PDE concept and the issue of mixing under flight conditions along with any experiments that may have been done in the US or elsewhere remain highly classified.
Initial tests
The use of liquid fuels such as JP-10, which is generally preferred over gaseous fuels for practical purposes, makes mixing even more difficult. Some initial tests have been conducted at University of California San Diego on a system that operates with low air pressure drop and yet still makes small droplets below 10 microns. Almost the only good news for the PDE community in this area are results from previous research showing that atomisation under pulsing or transient conditions is virtually identical to steady-state conditions. Smaller droplets are expected to be generated by the addition of a swirl, which will also help tailor radial and axial fuel distribution.
Generating reliable and repeatable ignition is perhaps the main challenge faced by the PDE hopefuls. Early research quickly established that the amount of energy required to spark detonations in hydrocarbon-air mixtures (other than acetylene) is impractical. The only alternative to making a PDE concept possible is a process dubbed DDT, or deflagration to detonation transition, in which high-speed flame transitions to detonation down a tube. Most of the detonation research is now, therefore, focused on DDT tube lengths and on means of reducing the time taken for the process to occur. Various shape blockages, orifice plates and even spirals have been tested in DDT tubes, as have chemical and fuel additives such as nitrate sensitisers and hydrogen peroxide.
Alternative research has focused on initiating detonation in a smaller chamber or tube containing an easily detonatable mixture such as ethylene-oxygen and then transitioning it to the main chamber with a larger mixture of ethylene-air.
However, proponents of PDEs for small-scale practical uses of the engine on hypersonic targets or cruise missiles (such as those being studied in Japan by Mitsubishi Heavy Industries), point out that this would require the need for on-board oxygen generation and add weight and complexity. Detonation transmission from a small tube to a larger chamber has been shown to be easier when combined with devices that focus the shock waves, or have other design adaptions to smooth transition.
Recent Russian PDE research, for example, shows that a more gradual area change instead of an abrupt change works well. The Russian experiments include a convergent section at the end of the main chamber causing a multistep detonation and producing flow regimes higher than those in a stationary detonation wave.
While many of the detonation studies have concentrated on gaseous fuels, the focus for volume-limited applications means close attention is also being paid to liquid fuels such as JP-10. This is also a good endothermic fuel and therefore an effective heat sink and is expected to help improve heat management in PDEs.
Fuel tests
Researchers also believe it may also be sufficient to help pre-vapourise some of the fuel and aid in the start of the detonation in fuel-air mixtures. Tests with this fuel indicate that JP-10/air detonations occur with droplet sizes of around three microns, and an estimated fuel-vapour content of about 70%.
The unsteady nature of the PDE process means that nozzle design is a key design challenge for researchers, even if made to appear more steady by using a high-frequency, multi-tube system. Tests on nozzles of various shapes and sizes to-date indicate that divergent designs give higher impulse while, within this group, the bell-shaped nozzle have shown "significant improvements in performance in at least three independent studies", says PDE studies watchdog, the Washington DC-based Office of Naval Research. It adds that studies of convergent nozzles generally "show shock reflections that propagate upstream and interfere with the refilling process".
Basic research is also under way at AFRL into methods of self-aspirating, or essentially self-starting PDEs using ejector pumps and a turbo-charger. Experiments using valves adapted from a General Motors "Quad 4" car engine cylinder head were used to stop and start the fuel-air mixture and purge air flow into the detonation tube. Results showed the ejector induced three times the flow, showing self-aspiration was possible, while a turbine and compressor survived a 25min self-aspirating run.
Further tests are planned with axial flow turbines and compressors, as well as experiments that use the turbine to increase pressure in the detonation tube.
While the current state of research indicates severe challenges ahead for successful PDE applications, it appears to do little to suggest that such a powerplant yet exists in the classified world. "Based on what I've seen so far, I'd say we still have a long way to go," concludes Doychak.
Source: Flight International
