From concept to the manufacture and testing of a novel heat exchanger technology, Reaction Engines’ hybrid Sabre engine is set for great things. RICHARD WARRILOW gives an in-depth update to the on-going development story that has the potential of making the UK a key player in the commercialisation of space.
This is a full article published in Aerospace International: June 2011
In the January 2010 issue of Aerospace International Richard Gardner took a detailed look at Reaction Engines’ plans for Skylon, the reusable spaceplane concept, and its potential role in the commercialisation of space. The article touched briefly on the Sabre engine that will propel Skylon into low Earth orbit (LEO). A great deal of engine development and testing has taken place during the intervening 18 months, so now seems the ideal time to consider Sabre in more detail; and to reveal a few facts and figures that will provide testimony to the remarkable work being done in Oxfordshire.
As the earlier article noted, it is the weight penalty associated with carrying (in liquid form or as a solid) the oxidiser needed for combustion that defeats most single-stage-to-orbit (SSTO) rocket concepts. Sabre’s trick is to have two modes of operation: air-breathing and rocket.
In air-breathing mode Sabre will use oxygen from the atmosphere up to an altitude of 26km, which will be attained at Mach 5.14. It will then switch over to rocket mode, burning on-board liquid oxygen for the remaining distance to low Earth orbit (roughly 300km — which will be attained at Mach 25). As Alan Bond, managing director of Reaction Engines, commented: “If Skylon’s engines were not able to make use of the oxygen present in air for those first 26km of ascent, we estimate the spaceplane would need to carry an additional 250 tonnes of liquid oxygen — a weight penalty far greater than that of the components needed for Sabre’s air-breathing mode.”
To minimise engine mass and base drag the Sabre engine design uses a common combustion system and expansion nozzle for both modes of operation. It will use liquid hydrogen as a fuel because of its higher energy density compared to hydrocarbon fuels and because of its thermodynamic properties. Along with a closed loop helium system (detailed later), the liquid hydrogen will be used in a thermodynamic cycle in which work transfer will play as significant a role as heat transfer.
Why the need for heat transfer? In air-breathing mode Sabre will face the same problem as any gas turbine aero-engine operating at high speed, namely the considerable heat generated by the compression of air. Concorde’s engines, for example, dealt with a temperature of about 160ºC at Mach 2 and SR-71 Blackbird’s engines coped with 400ºC at Mach 3.
However, the relationship between temperature and Mach speed is not linear and at Mach 5.14 the temperature of the air going into the intake will be an estimated 1,000ºC. Richard Varvill, Reaction Engines’ technical director and chief designer, noted: “Unless this heat can be dissipated it will severely restrict the materials that can be used within the engine.”
Sabre therefore sees the addition of a pre-cooler in front of its compressor to reduce the air temperature down to about –120ºC; which is roughly the vapour boundary of air and about as low as you can go without liquefying it. This approach also avoids the need for an air condenser but still provides enough margin to then pressurise the air to around 150bar; the pressure needed to inject it into the rocket combustion chamber.”
With reference to the above simplified Sabre cycle diagram, in air-breathing mode the air flow will pass from the intake to the pre-cooler (HX1-2), which will be cooled by cold, high pressure (200bar) helium. The cold air will then be compressed and delivered to the pre-burner and the main combustion chamber (‘rocket’ in the diagram).
After leaving HX1-2 the temperature of the high pressure helium varies with the air temperature and so it is further heated to a constant delivery temperature of around 900ºC in HX3, before passing to the turbine to drive the air compressor. It then passes to HX4 where it is cooled back to cryogenic temperatures by the liquid hydrogen. The helium circulator then drives it back to HX1 to complete the cycle.
All of the hydrogen will pass to the pre-burner where, in combustion with some of the air flow, it will produce a hot (circa 1,500ºC) hydrogen-rich pre-burner exhaust product. This will be used to heat the helium in HX3 before passing to the main combustion chamber; where it will meet the remaining air flow.
The pre-burner temperature will be controlled at different flight conditions by adjusting the air flow split between the main combustion chamber and the pre-burner. And when Sabre switches over to rocket mode, the turbo-compressor will be removed from the power loop, and liquid oxygen will be pumped in, vapourised and will substitute for air.
Progress to date
Technically speaking Sabre, and a spin-off engine concept called Scimitar are both pre-cooled turbine-based combined cycle (TBCC) engines. Engines of this basic type have been on the drawing boards of several organisations such as NASA for over 50 years. The Sabre is a new, very novel variant of this basic engine class and its realisation will hinge on the practical feasibility of manufacturing low-mass, high-surface area heat exchangers; and that is what Reaction Engines claims to have solved.
Bond says that pre-cooler technology has been under continuous development in his company for about ten years and that work is most advanced with a super alloy shell and tubular matrix technology that will be used to make HX1-2 and HX4. Indeed, a Technology Demonstration Programme (TDP), initiated in February 2009 and which has the objective of validating key technologies like the matrix, is scheduled to produce results and draw conclusions in the summer of this year.
A significant part of the TDP has been the construction and test of a scale version of the SABRE pre-cooler; using a Viper jet engine as the test bed. The scaled pre-cooler is being built from 21 (full-size) modules, each containing several hundred matrix tubes, arranged in a drum. Note: production drums will contain nearer 80 modules.
The tubes are manufactured by tube-drawing, a process that dates back to the early days of the Industrial Revolution but Reaction Engines claims current manufacturing capabilities are being pushed to the very limits; due to the combination of small diameter (outer = 1mm), small wall thickness (40 microns) and material used (Inconel 718).
Post-manufacture each tube is subjected to a series of dimensional and strength tests to ensure it is of adequate quality to use in a module. The testing is done by a purpose-built machine, capable of detecting flaws invisible to the naked eye, which automatically positions any detected flaws under a microscope to facilitate a more detailed inspection.
Defect-free tubes are processed to tailor their wall thickness, and then pressure tested and cut to length. The processed matrix tubes are then assembled in a braze fixture, in conjunction with headers and baffles, to form a complete pre-cooler module. Finally, the entire fixture is loaded into a furnace and subjected to an intensive time/temperature brazing cycle, then subjected to leak and pressure.”
Varvill: “In many respects, for the pre-cooler modules, we’ve already embraced the volume manufacturing and testing techniques — and quality procedures essential for release to service — that will be needed for Sabre production.”
As mentioned, via the helium loop, liquid hydrogen will be used to cool the incoming air while Sabre is in air-breathing mode. This means it is not available for other cooling duties, most notably for cooling the combustion chamber as is done for conventional rocket engines.
Aerospace International Contents – June 2011
News Roundup – p4
Paris in the springtime -p 13
Paris Air Show preview
Partners in partner - p 14
How Rolls-Royce’s academic network drives innovation
What next for Europe? – p18
The future of the European fighter industry
Cutting edge – p 22
Reaction Engines’ Sabre engine?
Northern exposure- p 26
A report on the Canadian aerospace sector
Steady as she goes- p 30
Latest progress on the A350XWB
Cabin fever – p32
Aircraft Interiors Hamburg show report
The last word – p34
Keith Hayward on the benefits of space exploration
Reaction Engines’ proposed solution to this is for, in air-breathing mode, SABRE’s combustion chamber to be cooled by a combination of air in a jacket of tubes that makes up the engine thrust chamber and LH2 film-cooling inside the chamber itself. Varvill adds: “Having established that the rocket chamber is cooled by the oxidiser this needs to continue during rocket mode, in which the on-board liquid oxygen replaces the air to cool the chamber.”
As part of the TDP, two contracts have been placed to explore the concepts. EADS Astrium in Ottobrunn, Germany, is currently investigating the heat transfer characteristics of liquid oxygen. One study of heat transfer correlations has already been completed and have enabled Astrium to build a sub-scale LOX-cooled combustion chamber; which was tested at the rocket test facilities at Lampoldshausen in 2010.
In the second contract — this one with the German Space Agency (Deutsches Zentrum für Luft- und Raumfahrt or DLR) — another test combustion chamber is being used to examine the effectiveness of compressed air for cooling. Varvill comments: “The tests also include hydrogen film-cooling within the combustion chamber. The technique has been used on previous rocket engine concepts to provide additional cooling but for Sabre the flow of hydrogen in the film will be somewhat greater.”
Another part of the TDP is focused on the propulsion system nozzle, and how it must cope with a wide range of atmospheric back-pressures as Skylon ascends. The nozzle must also have a large area ratio in order to be highly efficient at altitude when Sabre operates in rocket mode.
The University of Bristol is currently undertaking the development of an expansion-deflection nozzle concept for the Sabre engines. This work includes the testing of a variety of candidate nozzle contours in a newly refurbished test facility together with a programme of computational analysis. The university’s work will culminate in the design and test firing, this summer, of a hydrogen-air burning engine with expansion/deflection nozzle.
Designed to power a Mach 5 civilian airliner, Scimitar is a spin-off from the 2005 LAPCAT study, initiated to examine a number of propulsion concepts and technologies required to reduce long-distance flight times for new hypersonic commercial transport aircraft capable of between Mach 4 and 8.
However, there are also considerable differences between the Sabre and Scimitar concepts, because of their differing intended roles. For example, Sabre is optimised to be an accelerating engine — and to behave like a rocket even when in air-breathing mode. The Scimitar engine design is optimised for sustained Mach 5 flight. It must also fly with acceptable efficiency at Mach 0.9 during overland flight path segments to eliminate sonic boom, and be sufficiently quiet at take-off to satisfy international noise regulations.
Fundamental technical differences between the two engines include Scimitar’s six rather than four heat exchangers — although the main pre-cooler is simpler than Sabre’s — and and the inclusion of a fan in its bypass duct (shown in yellow in the above diagram).
Space without costing the Earth?
Reaction Engines is privately owned, has about 80 shareholders, and the majority of monies are coming in through private investment. For example, the European Space Agency (ESA) — through the UK Space Agency (formerly the British National Space Centre (BNSC) — part-funded the development of Skylon’s core heat exchanger technologies to the tune of €2m but private funding was almost quadruple that. “We’re not looking for funding from the UK Government,” says Bond. “It made its views on backing this kind of venture clear when it scrapped HOTOL. All we ask is that the Government be active in reforming the international legislation governing space, and change the way that projects like ours are insured so that they become increasingly attractive to further investment.”
As for Skylon’s overall development costs, the most recent formal costing exercise was done against the last but one build configuration and came to $12bn, a figure that is on par with Ariane 5. However, the mature operation price of Skylon is predicted to be about $1,000/kg, compared to that of a typical expendable launch vehicle which is approximately $10,000/kg (typically after a 50% government subsidy).
Looking back on the January 2010 issue of Aerospace International, and the assortment of artists’ impressions it contained, it is reassuring (and frankly very impressive) to see some of the technologies critical to Skylon’s success being realised.
Indeed, the Skylon project epitomises what is truly great about aerospace engineering; for example that tubes 1mm in diameter and with a wall thickness of only 40 micron can play such an instrumental part in providing better access to space. Bond concludes: “Pre-cooled engines, such as Skylon, herald a new age in transportation with easy access to space and nowhere on Earth more than four hours apart.”