The GSLV F10 lifts off from Sriharikota, at 5:43 am on August 12, 2021. Photo: Doordarshan/YouTube
- ISRO launched its GSLV F10 mission on August 12 with the GISAT-1 Earth-observation satellite. The mission failed 10 minutes after liftoff.
- ISRO chief K. Sivan had suggested in August that he thought the failure was the result of a “process issue” and within an expected failure rate.
- But a Reddit user found a senior ISRO official saying in a talk that the F10 mission failed because of a technical glitch.
- On March 25, the Failure Analysis Committee submitted its report. Its findings were in line with the ISRO official’s account.
This article was first published on December 2, 2021, and was republished on March 27, 2022. This is because on March 25 the Failure Analysis Committee of the GSLV F10 mission submitted its report on the reasons the mission failed and its subsequent recommendations. The report confirmed what the ISRO official quoted below said in December, that:
“… the lower liquid hydrogen tank pressure at the time of cryogenic upper stage engine ignition, caused by the leakage of vent & relief valve, resulted in the malfunctioning of the fuel booster turbo pump, leading to mission abort command and subsequent failure of the mission.”
GSLV-F10/EOS-03 Mission Failure: Insufficient pressure conditions in the propellant tank of the Cryogenic Upper Stage at the time of its ignition led to the failure, Failure Analysis Committee finds in its report submitted yesterday. https://t.co/HOLwv4QRz3
— ISRO (@isro) March 25, 2022
The GSLV rocket on the F10 mission flew in its Mk II configuration, whose third and final stage uses a cryogenic engine, which combusts liquid hydrogen with liquid oxygen. If the liquid hydrogen tank had a leak, the liquid wouldn’t have been pressurised, which means not enough of it would have reached the engine for combustion. This would have caused the engine to fail, like a heart attack.
According to the report, the vent & relief valve is likely to have developed a leak after a soft seal in the valve got damaged – due to regular wear/tear, contamination or anomalous stresses. The committee has also recommended that ISRO incorporate additional steps before each of the involved systems are activated during the rocket’s launch.
Bengaluru: The Indian Space Research Organisation (ISRO) launched its second mission of 2021 on August 12 – a Geosynchronous Satellite Launch Vehicle (GSLV) carrying the GISAT-1 Earth-observation satellite. However, the mission, designated F10, failed 10 minutes after liftoff as one of the rocket’s engines failed to perform as expected. The whole ensemble crashed into the Indian Ocean.
While ISRO has since launched a probe into the incident, it has divulged no other details – not even the progress of the inquiry. K. Sivan had reportedly told Times of India on August 19 that he had asked the Failure Analysis Committee to submit its final report by the end of the month. If it has been submitted, ISRO hasn’t placed neither the text nor any details from it in the public domain so far.
But thanks to a tradition of ISRO employees sharing important details in obscure speeches and communiqués, it has now come to light that the F10 mission failed specifically because of an anomalous drop in pressure in a tank containing liquefied hydrogen.
Reddit user Ohsin spotted the detail in a talk by V. Narayanan, the director of ISRO’s Liquid Propulsion Systems Centre, delivered in mid-November on the occasion of the Vikram Sarabhai Space Centre Quality Day 2021. Ohsin subsequently shared their findings on Reddit’s ISRO forum. According to Ohsin, Narayanan said that “50 millibar reduction in LH2 tank pressure led to failure of GSLV F10 mission.”
For the F10 mission, ISRO launched the GSLV rocket in its Mk 2 configuration. One of its most significant parts is the cryogenic engine powering the third and uppermost stage. The Mk 2 rocket uses the CE7.5 cryogenic engine, which ISRO developed from Russian technology. In the Mk 3 configuration, the rocket uses ISRO’s indigenous CE20 cryogenic engine. (The now-retired Mk 1 configuration used the Russian RD-56 engine.)
Rocket scientists prize cryogenic engines because they prize the fuel used in these engines: hydrogen. Of all the known rocket fuels, hydrogen has the highest specific impulse – which is “how much more push accumulates as you use that fuel” (source). The CE7.5 engine has a specific impulse of 4.452 km/s.
But these engines are very difficult to operate because rockets need to carry both the hydrogen and the oxygen needed to combust it in liquid form.
As gases, they occupy too much space and are harder to pump. However, both elements become liquid only at extremely low temperatures: oxygen at -183º C and hydrogen at -253º C. This means the engine and its various attendant parts need to be able to operate at low temperature, with tightly controlled pressure, flow rates and flow profiles.
According to Narayanan, the CE7.5 engine failed to perform as expected because the tank holding liquefied hydrogen had 2.5% lower pressure than it needed to have. This may seem like a small margin of error, but as various commentators on the ISRO Reddit forum have also pointed out, the CE7.5 engine is sensitive to such small changes.
Pumps work by creating a pressure differential between two tanks. If the hydrogen tank didn’t have a sufficient amount of pressure, there may not have been enough impetus to cause the liquid hydrogen to flow into the combustion chamber. And without liquid hydrogen, the cryogenic engine wouldn’t have provided thrust, causing the rocket to stop climbing.
It’s also notable that this engine uses the staged combustion cycle. As this author wrote in a previous article:
Two booster pumps supply cryogenic hydrogen and oxygen to a turbo-pump, which then feeds the engine’s combustion chamber. The turbo-pump supplies fuel and oxidiser. It is powered by a turbine, in turn driven by combusting a small amount of the fuel in a pre-burner. The name of the cycle is derived from the fuel being combusted in two steps; this leads to increased fuel efficiency. The ratio of the mixture is controlled by a regulator. While the main engine provides the thrust, two separate vernier engines ensure the rocket follows its trajectory by firing smaller pulses.
The indigenous CE20 engine operates using the gas-generator cycle, in which the fuel is combusted in a single step. The engine as a result has lower fuel efficiency but is also less complex and, thus, easier to operate.
Comparing Narayanan’s revelation with something ISRO chief Sivan said shortly after the F10 mission failed may be instructive:
While the effect of such malfunction may be severe [failure of mission], the cause may actually turn out to be insigificant. Sometimes, the systems that appear to have a good margin during pre-launch tests and previous missions don’t perform as expected when launched. The GSLV Mk 2, for example, has had six consecutive successful flights before last week, which is why we think it may be a process issue.
That is, Sivan suggested that ISRO didn’t have reason to suspect a technical issue and that the mission could have failed for reasons that ISRO had anticipated. But what Narayanan said in his November talk contradicts this view by indicating that the mission failed for a reason that could not have been expected (e.g. a sensor failure).
Error margin is there more than this to accommodate this drop .
This is happening for the first time.
The associated Auto Launch Sequence Software may not be synchronised to accommodate this .This is what I think
— Dr. P V Venkitakrishnan (@DrPVVenkitakri1) November 20, 2021
An unexpected mode of failure is considered more dangerous. That this happened in a way related to a cryogenic engine may be additional cause for concern since ISRO is working towards human spaceflight through its Gaganyaan mission, which will use the CE20-powered GSLV Mk 3 rockets.
In any case, ISRO may need to check for these modes of failure in additional tests – or it may already have, if the Failure Analysis Committee submitted its report by end-August as Sivan had asked.