Among the key takeaways:
- Engineers are investigating several possible causes for why ISRO’s first cryogenic upper stage failed on a GSLV launch in April 2010;
- The complexity of the investigation has pushed the next cryogenic engine test back to March 2012, about a year later than officials originally hoped;
- A GSLV rocket failed on Christmas Day because the Russian supplied cryogenic engine was not built to dimensions specified in documentation, causing a protective shroud to give way and snap vital connectors;
- The Chandrayaan-II lunar probe, set for launch around 2013, is being delayed not only because of problems with cryogenic upper stage technology but also due to Russia’s decision to reduce its contribution to the joint program;
- ISRO is aiming for greater transparency and a more arms-length relationship with Antrix, its commercial arm.
Mastering cyrogenic engine technology and making GSLV reliable are two key priorities. Although ISRO has had great success with its smaller PSLV rocket, the GSLV is central to India’s plans to launch commercial communications satellites and astronauts into orbit.
The GSLV has a very poor record since it was introduced a decade ago. It has been launched only seven times in 10 years, with two successes, four failures, and one partial success. Two of the four failures occurred last year.
The small number of flights over such a long period of time creates challenges. Rockets typically need to be launched dozens of times before their reliability is understood. A low flight rate is not very good for keeping the teams that build and launch a rocket sharp.
The other problem is that each rocket is a brand new, unique vehicle, one that incorporates upgrades that ISRO has been introducing over the past decade. Introducing changes to a complex system that is still being matured can solve some problems but create others.
Radhakrishnan says that practice is continuing:
“There are also changes in the whole GSLV vehicle per se. From 1.5-tonne payloads, we’ve now reached about 2.2-tonne payloads. The biggest change effected is the size of the heat shield [a protective sheath that also determines the size of communication satellites to be put into orbit] and the materials we use for it. We’ve gone from 3.5m heat shields to 4m, and use composites [such as plastic fibre] instead of metal. For GSLV Mark-3, we may have to use 5m shields.”
In short, it doesn’t seem that ISRO hasn’t been able to settle on a single GSLV design, launch the vehicle enough to make it reliable, and then introduce changes only after the rocket’s flight performance is fully understood.
Radhakrishnan was asked about ISRO’s recent failures, and his answer is interesting:
Relative to the number of failed satellite launches, does Isro have greater success compared with international agencies?
There are two things: on-ground tests and eventual flight tests. Each country has its own philosophy. Russia is very conservative; America less so. I can’t really say where India is on that scale, but we have our own approach, too, that we have debated since our space programme began. Our main endeavour is to get as many parameters tested in static (ground) tests, because each of these tests involves a lot of time, money and effort. Russia also does a lot of ground tests. That said, however, between 1957 (when the so-called space age began) and 2010, there have totally been 320 failures across nations, and there’s no constant success-failure ratio (across nations).
ISRO clearly needs to do more ground testing before it loses any more satellites. But, it also needs to fly more often. And all of that is costly.
Read the full interview.