Russian route

The successful launch of the developmental flight D5 of the Geosynchronous Satellite Launch Vehicle (GSLV) on January 5 marks the end of a 20-year-long and difficult journey to the indigenisation of the Russian cryogenic technology acquired in 1995. The upper stage of this three-stage GSLV was powered by the indigenously built cryogenic engine and stage, which was modelled entirely on the Russian engine KV1 designed to deliver 7.5-tonne thrust (75 kN) and the Russian stage 12KRB designed to carry 12.5 tonnes of liquid propellants, liquid oxygen (LOX) and liquid hydrogen (LH2).

When Glavkosmos, the commercial arm of the Russian space agency, reneged on the agreement on transfer of cryogenic engine technology to the Indian Space Research Organisation (ISRO) in 1993, U.R. Rao, the then ISRO Chairman, famously claimed to the media at various fora that since some data, drawings and information had already been received from the Russians, indigenous development based on that technology would be achieved by 1997. In off-the-record conversations, however, ISRO scientists admitted that it would take a minimum of 10 years. Retrospectively, even that turned out to be highly optimistic. It has taken a full two decades.

While indigenisation of an acquired technology has taken two decades, ISRO’s efforts at an entirely indigenous development of the technology are four decades old. Once import of Russian technology became possible, indigenous development, which, since the 1970s, has been progressing in fits and starts (see separate article on page 22) right up to the new millennium, seems to have given way to an indigenous development route that has been strongly influenced by the acquired technology. (The indigenised GSLV is called GSLV-MkII as against GSLV-MkI, which uses the Russian engine and stage.) It is moot whether this was the right approach to develop a complex technology given the extraordinarily long time that the indigenisation process has taken. The truly indigenous GSLV-MkIII, powered by an engine and stage that are totally indigenous, awaits its test next year.

As recounted by S. Ramakrishnan, former Director of ISRO’s Liquid Propulsion Systems Centre (LPSC) and the current Director of the Vikram Sarabhai Space Centre (VSSC), in his Aeronautical Society of India presentation in June 2012, as part of the fully indigenous developmental efforts that began in the 1980s, many ground tests were carried out on a one-tonne thrust subscale engine in order to generate data for designing a large-scale 12-tonne thrust engine (C12).

Experimental tests

The first experimental hot test (real life test burning with the actual propellant) on the subscale engine was conducted in 1987 using gaseous hydrogen (GH2) and gaseous oxygen (GO2). This was followed by a test with the propellant combination of LOX and GH2 that used a heat-sink thrust chamber. The first fully cryogenic regeneratively cooled subscale engine, which used LOX and LH2, was tested after a five-year gap, in June 1993, at a newly built subscale test facility. In a regeneratively cooled engine, one of the propellants from the engine outlet passes around the chamber to prevent it from overheating.

The test had to wait for the setting up of the first ever liquid hydrogen plant in the country, which ISRO did. Also the revival of the indigenous development that was hitherto on a slow mode was spurred following the uncertainty over the Russian deal. The test was a failure, which was caused by the failure of an isolation valve (also called the “latch”), an important component of the test-firing equipment, resulting in a hydrogen explosion. The test had apparently been attempted using a valve unsuitable for cryogenic liquid firings. Identification of the problem led to the use of an Oxygen Free High Conductivity (OFHC) copper, instead of stainless steel (SS), thrust chamber. A subsequent test in November 1995 after modifying the test procedure was, however, successful, which was followed by more tests.

According to Ramakrishnan, a sea-level thrust chamber for C12 was designed, realised and two hot tests in de-rated conditions, using LOX and LH2, were carried out in 1998. Experimental studies were also conducted on turbo pump systems. These were tested with gaseous nitrogen as turbine drive fluid and pump handling water in May 2000. “These studies,” says Ramakrishnan in his paper, “gave the ISRO team the experience on design, analysis and realisation of cryo thrust chambers and turbo pumps, as well as in production, storage and handling of cryogenic propellants.” The regeneratively cooled subscale engine incorporated some important technology developments, like electroforming where thin metallic layers are deposited on a base surface by electrodeposition. Electroforming has been used in the French HM7, the main space shuttle engine, and the Japanese LE7 to form the coolant channels through which one of the propellants would pass around the engine. The electroforming technique for the one-tonne cryogenic subscale engine (and later for the 12-tonne (C12) engine as well) had been developed by the CSIR’s Central Electrochemical Research Institute (CECRI) at Karaikudi, Tamil Nadu.

The Russian engine, on the other hand, uses the vacuum brazing technique to fabricate the coolant channels and, for indigenising this aspect of the engine, a large, power-guzzling, brazing apparatus has been imported and the facility established at Godrej, Mumbai, because of which the indigenous development of electroforming technology seems to have been given the go by. According to K.N. Srinivasan of CECRI, a 12-tonne electroformed thrust chamber had been supplied to ISRO for testing at the Mahendragiri (southern Tamil Nadu) liquid engine testing complex. However, he was not aware of what the status of the test was and whether the technology will be used by ISRO at all.

When asked about this, Ramakrishnan said in an email response: “We had realised a 12-tonne thrust chamber by electroforming route and tested it at about 60 per cent thrust level after which we stopped the programme. However, for GSLV-Mk III, the 20 t thrust engine (CE20) development is progressing well. Of course, C12 experience has helped. However, the chamber for CE20 is a brazed version based on Russian engine technology. We are not pursuing electroforming route anymore.” Having invested in a brazing apparatus, ISRO has perhaps decided in favour of the brazing route.

The ISRO-Glavkosmos contract (see article on page 22) provided for seven cryogenic stages. The stage had a propellant (LOX+LH2) loading of 12.5 tonnes and was powered by an engine designed to deliver a thrust of 7.5 tonnes. This engine, originally designated as RD-56, had been developed for the Soviet moon mission in the 1960s. This was modified for use in the GSLV, and the stage too was specially designed for the GSLV. Besides the main engine, the propulsion unit had two pressure-fed vernier (steering) engines as well, which could provide attitude control by swivelling or gimballing. The engine used the complex staged combustion cycle (SCC) as against the simpler and more flexible gas generator cycle (GGC).

Significantly, both the subscale and the C12 engine were based on the GGC. “The C12 cryo engine development programme,” says Ramakrishnan, “was abandoned once the Russian contract was signed to get cryo technology and right now no work on C12 is going on. The focus changed to indigenous version of Russian cryo engine which incidentally operated on SCC unlike GGC proposed for C12.” Interestingly, however, for the GSLV-MkIII, powered by the indigenously designed 20-tonne thrust CE20 engine (and C25 stage) with a much greater target payload capacity of four tonnes-plus, ISRO scientists have returned to the GGC-based thrust generation. (Also see interview with ISRO Chairman K. Radhakrishnan.)

According to Ramakrishnan, the contract provided for only the propulsion hardware to be supplied by Glavkosmos. The control systems required for the stage and mission management—including sequencing, tank pressure control, thrust and propellant mixture ratio control, gimbal control and post-flight passivation, the facilities for stage preparation and propellant servicing—which have provided the much-needed experience in the indigenisation process were to be developed by ISRO. Further, during the development and qualification of the engine and stage for carrying out engine hot tests in Russia, ISRO’s avionics were used for thrust and mixture control and steering engine gimbal control. Stage-level cold flow and hot tests were also conducted in Russia using ISRO’s GSLV equipment bay.

Stage test in Russia

According to the authors of the book A Brief History of Rocketry in ISRO , a stage test in Russia was not part of the original contract. It had been hoped that necessary facilities would be ready at Mahendragiri within the projected time frame for these tests. But there was a delay in getting these facilities ready. Therefore, an additional contract was concluded for stage tests in Russia, presumably at a higher price than it would have cost otherwise.

Following the cancellation of the technology transfer part of the deal, ISRO proposed a Cryogenic Upper Stage (CUS) project, which was approved in April 1994. The objective of the project was to realise an indigenous Cryogenic Stage with the same specifications, configuration and interfaces as the Russian CS for the continuance of operational GSLV flights after all the stages supplied by Glavkosmos were flown.

“The 2.8 m diameter Indian CUS, with 73.5 kN [~7.5 t] nominal thrust SCC engine incorporated several changes from the Russian CS at component and sub-system level, such as pressure regulation and pyrogen igniters, based on the expertise and experience at ISRO,” points out Ramakrishnan in his paper. “However, the main stage architecture of fixed main engine with two steering engines was adopted from CS as such.”

For cryogenic operations, new facilities had to be created. ISRO established LH2 and LOX production plants at Mahendragiri. An integrated LH2 plant at 500 kg/d capacity was set up under a contract with Linde AG, Germany, in 1995. Since cryogenic filling operations before the launch are a great deal more complex than filling operations for earth-storable liquid propellants such as UDMH and NO, a separate contract with Glavkosmos was arrived at for establishing the cryogenic propellant filling plant at the SHAR launch complex. For fabrication of engine and stage hardware of the main engine and stage systems, infrastructure was established at Godrej, and MTAR, Hyderabad, and Hindustan Aeronautics Limited, Bangalore.

The first indigenous engine, A0 was tested on February 16, 2000. The test failed after firing for only 13.7 seconds, and when it was aborted by manual intervention as a fire was detected around the engine ion test stand. The engine and hardware were totally damaged. The first test with A1 was terminated prematurely too as the ignition did not take place in the gas generator. Following a failure analysis, the pyrotechnic igniter was replaced by a pyrogen igniter. Following a series of nine tests on A1, accompanied by several problems, changes and modifications of the configuration, the first successful test of A1 happened on February 9, 2002.

A2, whose configuration is similar to A1 but for a few additional subsystems, underwent 10 tests, including an endurance test that lasted for 1,000 s. The first vernier engine was successfully tested at the subscale test facility using GH2 and LOX on June 25, 2001. According to Ramakrishnan, after extensive development and qualification tests at engine level, the indigenous stage was realised and subjected to successful hot test for the flight duration of 720 s on November 15, 2007. The tests confirmed that the indigenised engine was capable of delivering 27 per cent more thrust than the Russian version.

Indigenous flight stage

The first indigenous flight stage flew in the developmental flight GSLV-D3 on April 15, 2010. The engine fired successfully but could not develop the requisite thrust owing to a malfunction of the LH2 turbo pump. Before the first successful launch of D5 took place, nearly 40 hot tests on seven main engines for a cumulative duration of nearly 8,000 s and about 80 firings on 18 steering engines for a cumulative duration of about 10,500 s had been carried out.

The idea to replicate the Russian design for GSLV-MkII was done with a view to minimise the development time and lessen the risks for early operationalisation of the GSLV. From that perspective, even the special materials required for the engine and stage were procured from Russia by 1996 and successfully indigenised. Also, the six launches with the Russian stages, of which four were failures, however, have given ISRO scientists the necessary lessons for successful indigenisation (see interview with K. Radhakrishnan).

These launches—successful, sub-optimal and failed— also provided the necessary experience and data for precise calculation of aerodynamic loads on the upper cryogenic stage, optimisation of the launch azimuth and “range safety” considerations, increasing the thrust rating, higher propellant loading in the solid booster and the liquid stages, increasing the pressure of the liquid stages L40 and L37.5, reduction in the mass of the equipment bay and use of additional composite materials where possible, and obtaining optimal performance of the vehicle in terms of its payload capacity against the targeted delivery of a 2.5-tonne satellite in the Geosynchronous Transfer Orbit (GTO).

During the initial interactions with the Russian engineers, with the best values for the aerodynamic loads and the optimal launch azimuth given the “range safety” constraints (which meant unspent fuel in the second stage liquid engine L37.5), the mass of the cryogenic stage had increased considerably from the original indications and it seemed that the satellite mass could not exceed 1.5 tonnes. Indeed, the satellite GSAT-1 aboard the first developmental launch GSLV-D1 weighed only 1,540 kg.

But later exercises of optimisation resulted in the weight of EDUSAT, launched by the first operational flight GSLV-F01, being increased to 1,950 kg. In the failed launch GSLV-D3, which used the indigenous cryo engine and stage for the first time, the weight of the satellite GSAT-4 could, however, be increased to 2.2 tonnes. In the recent successful launch with the Indian cryo, the weight of GSAT-14 was kept somewhat less at 1,980 kg.

“The first Indian cryo stage which flew in GSLV-D3,” says Ramakrishnan, “was indeed optimised and was lighter than the Russian cryo stage. Also, in D3 we operated the engine at an uprated thrust level. Another factor was more favourable launch azimuth. All these pushed up the payload to 2.2 tonnes. Of course in D5, we played it more conservative. On the basis of shroud failure experience from F06 flight, we made the structure more robust. Also in general, everywhere we kept more margins since our main objective was to prove the working of the Indian cryo stage. We can progressively enhance the payload to 2,250 kg and maybe ultimately 2,500 kg in the coming flights” (see interview with Ramakrishnan).

The next phase of cryogenic technology development, following the expertise gained in the CUS project as well as the successful launch of GSLV-MkII, is the development of GSLV-MkIII, to be powered by a 20-tonne thrust cryo engine CE20 based on GGC. It is expected to have a specific impulse of 443 s. “GGC was deliberately chosen,” writes Ramakrishnan in his paper, “to facilitate faster development by way of individual subsystems testing and qualification before integrated engine test.”

According to him, the various subsystems have been completed and the gas generator, injector, LH2 and LOX turbo pumps, start-up systems and so on have undergone development tests. The integrated power head of the engine developing a turbine power of 2 MW was successfully tested in bootstrap mode on July 30, 2010. “Indigenously designed turbo pump unit of this power level to be successfully realised within the country marks a major milestone in the area of aerospace technology in India,” he says.

The CE20 and C25 are entering the crucial stage for validation tests, which are aimed at the launch of GSLV-MkIII next year. The immediate target for MkIII, however, is a successful launch of the experimental LVM3-X mission in the next few months. LVM3-X will be a GSLV-MkIII rocket except for a passive cryo stage with non-functional cryo engine to test all other elements of this new indigenous ISRO beast.

Despite the recent success of the GSLV and an upbeat ISRO looking into the future with an entirely indigenously developed cryogenic engine, which no doubt will have a mix of the legacy of the Russian engine and domestic efforts of the 1980s and the 1990s, it is still worth asking if the long-drawn indigenous developmental route via Moscow was the best way of achieving it.