It was not the cryogenic stage involving the imported Russian engines everyone was watching with bated breath that failed. The launch vehicle countdown did not even progress to the third stage of the launch, where the famous cryogenic engines would have been fired. Barely four seconds before ISRO's pride and joy, its last-generation launch vehicle, was ignited for take-off on March 28, the flight was stayed by a computer that was doing the final checks.

The reason? According to ISRO Chairman Dr K Kasturirangan, one of the four strap-on engines, which are the very first to be ignited in the launch, simply failed to generate enough thrust.

Stripped of the jargon, it means this version of the GSLV is a three-stage vehicle. The launch vehicle is 49 metres tall and weighs about 401 tonnes at lift-off. Its first stage consists of a solid propellant motor, known as the S125, and four liquid-propellant strap-on motors (the L40s). This stage -- 20.3 metres long and 2.8 metres in diameter -- develops a 4,700 kiloNewton thrust and burns for 100 seconds. The entire vehicle is launched at an azimuth of 104 degrees. It takes about 1,040 seconds from lift-off for the spacecraft to be injected into its ultimate destination, a Geo-Synchronous Transfer Orbit.

The L40s -- 19.7 metres long and 2.1 metres in diameter each -- are fabricated using aluminium alloy. Each of them is loaded with 40 tonnes of hypergolic propellants, namely Unsymmetrical Dimethyl Hydrazine (UDMH) as fuel and Nitrogen Tetroxide (N2O4) as oxidiser, stored in two tanks mounted in tandem. Together, they form the liquid propellant.

On the launch pad, as soon as the final countdown is complete, the L40s are ignited. The solid propellant core stage ignites exactly 4.6 seconds later, after the normal operation of each L40 stage has been confirmed. This is an automated, computer-controlled process.

The solid propellant core burns for 100 seconds after ignition. The four liquid strap-ons continue to burn for 160 seconds. By this time, in a successful launch, the space vehicle would have reached an altitude of about 73km. Then, the second stage of the GSLV, which uses liquid propellants, namely UDMH as fuel and N2O4 as oxidisers in two compartments of an aluminium alloy tank separated by a thin metal sheet, ignites. This stage uses engines similar to those used in the liquid strap-ons.

Exactly 1.6 seconds after the liquid propulsion second stage, the GS-2, ignites, the first stage separates from the launch vehicle and falls off as space debris. So, as you can see, the liquid strap-ons, which are the very first to ignite, and the solid propellant core, are crucial for transporting the launch vehicle for the first 73km.

The entire launch sequence is so programmed that if any part of any stage fails, the entire launch is put immediately on hold by the computers. And the vast banks of computers at the Sriharikota launch control room are manned by dedicated scientists, who examine every parameter displayed and recorded minutely once the countdown begins.

The tension in this huge control room is, quite understandably, electric whenever there is a launch, especially one as major and significant as the first developmental GSLV flight. The media is barred from this sanctum sanctorum at such times. On other occasions, though, ISRO has given many journalists, including this reporter, a grand tour of these facilities.

During the launch, journalists are usually asked to view the proceedings from the terrace of the Brahma Prakash Hall (named after one of ISRO's former directors, now dead), located almost seven kilometres away from the launch pad. This time was no exception. Journalists waiting for the launch heard a loud roar, which the experienced know usually goes with the ignition of the strap-ons of the first stage. Television viewers too saw the strap-ons ignite with leaping flames.

Then, it was all over. According to Dr Kasturirangan, one of the four strap-ons failed to generate enough thrust. The computers detected this immediately and put the launch on hold. If this problem had not been detected, the entire mission would have failed because the launch vehicle would not have reached the second stage. Then, the satellite, as well as the invaluable Russian cryogenic engine and cryogenic stages -- bought, according to ISRO sources, at close to Rs 450 crore [Rs 4.5 billion] each, from Glavkosmos, the Russian space agency -- might all have raced towards earth and crashed into the sea. Launch pads are always located at coastal areas so that the debris from bad launches does not cause damage to life and property.

The satellite itself is valuable, but not as valuable as it could have been, given that this was an initial flight. ISRO scientists readily admit that they use only experimental satellites on first flights of new generation launchers. This one was no exception. Although the GSLV has been designed to ultimately help India launch her own INSAT communication satellites, this time it was to place only an experimental satellite called GSAT-1 in space. It was a deliberate decision to limit GSAT-1's weight to 1,540kg, as against the two or the 2.5 tonnes of the INSAT series.

And what does the GSLV have, by way of new launch technology, that its predecessor, the Polar Satellite Launch Vehicle did not? Apart from the cryogenic stage, about which enough has already been said, it was the liquid strap-ons. Until now, Indian strap-ons have always used solid fuel. It also has other new elements like a heat shield with a larger diameter than the PSLV did and a vented inter-stage between the first and second stages.

ISRO repeatedly tests every new innovation in simulated conditions that most closely resemble what the launch vehicle will face during its actual flight. Why, then, after such extensive and exhaustive testing, did a strap-on engine fail? This is the one crore question that nobody is willing or able to answer just yet.

On March 28, a visibly distraught Kasturirangan told journalists at Sriharikota that he and his colleagues would work day and night to find out what went wrong. Normally, in the case of such mishaps, a failure analysis committee comprising high-powered scientists from within ISRO is appointed within a day or two and they study in detail exactly what went wrong and why.

Actually, what happened on Wednesday was nothing new. The first developmental flight of every new generation of launch vehicles (this is the fourth) so far has failed. The Satellite Launch Vehicle (SLV-3 E1), the first one made in India, was launched on August 10, 1979. A jammed valve in its second stage control system caused the oxidiser to leak, leading to what was termed a partially successful launch.

Eight years and three successful SLV launches later, the Augmented Satellite Launch Vehicle failed the first two times it was fired. The first time, on March 24, 1987, the strap-on motors fired, but the first stage motor did not ignite, causing the launch to fail. A little over 15 months later, ISRO tried the ASLV again. This time, the flight failed 46 seconds after lift-off, causing the resultant debris to crash into the sea. It was only after another four years of design changes and improvements that the ASLV finally had two successful flights, in 1992 and 1994.

Likewise, the PSLV too failed on its first ever launch on September 20, 1993. The then ISRO chairman, Dr U R Rao, said this was because of a software error in the pitch control loop of the on-board guidance and control processor. But ISRO was fairly satisfied with this launch; its scientists said most of the launcher's systems had performed satisfactorily. Soon after Kasturirangan took over as ISRO chief, the PSLV was flown successfully for the first time on October 15, 1994. There have been three more successful PSLV flights (in 1996, 1997 and 1999). Since then, ISRO has been working steadily towards this GSLV flight.

The next PSLV launch is scheduled within the next couple of months; it has already been contracted to carry two commercial satellites, a German satellite called Bird and a Belgian satellite called Prova. ISRO scientists expect it to go on as planned as the launch vehicle a proven one and almost ready. Besides, a comfortable, tested launch is necessary for ISRO's own morale and its standing in the international space community.

The next INSAT satellite is also almost ready for launch. It is expected to be sent into space as usual by an Arianespace launcher from French Guyana, probably by the end of the year. ISRO sees no reason why the GSLV mission failure should cause any change in this schedule. As Dr Kasturirangan points out, the satellite programme and the launch vehicle programme were separated from each other long ago, since launchers are more difficult and time-consuming to make and operationalise.

The third project ISRO has on the anvil, scheduled either for the end of this year or early next year, is another PSLV launch, this time of a one-tonne meteorological satellite.

The GSLV failure, though, has dealt a blow to ISRO's plan of marketing it in the space bazaar in about two years from now. But since commerce has never been a priority with ISRO, it is unlikely to mourn this factor.

The other purpose of the GSLV was to make India fully independent as far satellite launches -- mainly the INSATs -- were concerned. But, as Dr Kasturirangan himself pointed out to rediff.com recently, our communication satellites might become heavier than the weight the GSLV -- eventually supposed to be able to launch two-tonne satellites -- could carry.

Research into the failure of the liquid strap-ons will now probably take place at ISRO's rocket manufacture and liquid propulsion centres in Kerala. The GSLV's next flight, which is now scheduled for next year, should also take place next year, reveal ISRO scientists. If enough time and energy is devoted to fault-finding, there is no reason why the flaw, whatever it was, should not be set right and the GSLV launched again next year.

RELATED REPORTS
GSLV launch aborted
ISRO chief plays down GSLV failure
Not a major setback, says Narasimha

ALSO SEE
The countdown: 10...9...8
Reaching for the Skies

Design: Dominic Xavier

The Rediff Specials