...
A little history about Vikram at this point can help one understand why a lunar lander is complicated business and why one out of two such missions ends in failure. India had not planned to make its maiden attempt at a soft landing on the lunar surface on its own. Even before the indigenously-built Chandrayaan 1 orbiter was launched, ISRO had decided that it could use the help and experience of Russia’s Federal Space Agency (Roscosmos) for the Chandrayaan 2 mission and signed an agreement with it in November 2007. For the joint Indo-Russian lunar mission, ISRO would have the prime responsibility for the orbiter; Russia for the lander and rover. The launch was planned for 2012. Though ISRO was ready with the orbiter on schedule, Roscosmos pulled out of the agreement in December 2011. This was after its Phobos-Grunt mission to put a lander and rover on a Martian moon in collaboration with the Chinese space agency failed. ISRO then decided it would build a lander and rover on its own and scheduled a launch for 2016. Meanwhile, the organisation repurposed its orbiter for the Mangalyaan, accomplishing it in record time.
....
Despite ISRO’s vast experience in building launchers and satellites, it soon found designing and developing a lander and rover a complex and uphill task. According to M. Annadurai, a former director of ISRO’s U.R. Rao Satellite Centre and till last year the in-charge of planetary missions, It is one thing to send an orbiter [to the Moon] as we did with Chandrayaan 1, fire an impactor probe to the Moon or send an orbiter to Mars. But to bring down an orbiting spacecraft and make it land softly on the lunar surface is vastly more complex and challenging. The key technologies ISRO needed to master were a flexible propulsion system that would regulate the lander’s descent, and a control system that would guide and navigate the spacecraft to a pre-designated spot on the lunar surface. Both these technologies have been developed in the past five years and are now the prime suspects in the premature termination of Vikram’s mission.
THE LANDING PLAN
After its separation from the Chandrayaan 2 orbiter on September 4, Vikram was orbiting the moon at a speed of 1,680 metres per second (or 6,048 km per hour, six times the speed of a commercial jet) and at a height of 30 km above the lunar surface. That velocity, along with the height, had to be brought down in a controlled manner to almost zero within 13 minutes of the descent phase. The lander would do so using the array of five rocket engines and eight tiny attitude control thrusters fitted on its base, which ISRO had developed for the mission. The engines were designed as throttle-able ones, their thrust varying with the regulating of the fuel flow, just like an accelerator in a car.
The control and guidance system was also developed to meet the complexities of a moon landing. With the distance between the earth and moon being 3.84 lakh km, there is a time lag of more than a second before commands sent from mission control reach the craft, and of another second when data about its implementation is relayed back. As decisions had to be taken in milliseconds during Vikram’s rapid descent to the lunar surface, ISRO developed a fully autonomous guidance and control system that would take care of all the exigencies and anomalies that may arise on the 15-minute flight. The craft was also equipped with highly precise measuring instruments to monitor its velocity, height, attitude, direction and position with relation to the moon’s surface, enabling Vikram’s onboard computer to take decisions in real time. The craft’s control and propulsion systems were also designed keeping in mind that the moon’s gravity is one-sixth of the earth’s. Both these systems were subjected to rigorous tests, simulating conditions corresponding to the moon’s erratic gravity profile.
THE BIG CHANGE
What was also under test was ISRO’s new plan for powered descent that was put in place just two years ago. When designing the lander, ISRO scientists had initially decided to work with only four engines instead of five. In this configuration, the engines and the guidance control system would gradually bring the speed and altitude down to around 10 metres above the moon’s surface. But then the concern arose that the engine thrusters, at this distance, would kick up a mini lunar dust storm that would envelop the craft and damage its vital equipment. ISRO then planned to shut all the four engines and instead strengthen the four legs of the craft to withstand the free fall from that height without damaging either the lander or the rover. A launch was scheduled for January 2018.
Meanwhile, a fierce debate had broken out among space scientists over the dangers of having a four engine-controlled descent for a moon lander. ISRO decided to circumvent the free fall by introducing a fifth engine at the centre of the lander’s base. This would have two advantages. The fifth engine would be fired only after all the other four engines were shut down at 10 metres and ensure a powered descent till touchdown. And since the engine was located at the centre of the craft, the plume of dust it would kick up would be pushed away from it. That decision would add more weight to the spacecraft: along with other changes in the configuration, the composite Chandrayaan 2 with the orbiter, lander and rover would now weigh 3.8 tonnes. This meant that Geosynchronous Satellite Launch Vehicle, or GSLV MarkII was no longer suitable as a launch vehicle, as it was capable of carrying a payload of only 2-3 tonnes. So the Chandrayaan 2 project team had to wait for GSLV MarkIII, ISRO’s heaviest rocket, then under development, to be validated. Rather than wait for the full range of trial flights, ISRO decided to take a risk by launching Chandrayaan 2 on GSLV MarkIII’s first operational flight. As it turned out, after an initial scare, the GSLV MarkIII fired beautifully on July 22, 2019, launching Chandrayaan 2 on its lunar journey.
THE WOBBLE
For Vikram’s descent to the moon, ISRO homed in on a parabolic powered descent trajectory divided into four distinct phases. The process would begin when the spacecraft was at a height of around 30 km above the lunar surface and 650 km away from the landing site. In the first phase, known as the Rough Braking Phase and lasting for 10 minutes 20 seconds, Vikram would use the brute force of its engines to brake its horizontal speed of 1,648 metres per second down to around 150 metres per second. In this phase, it would come down from 30 km to 7.4 km. While detaching from the orbit and independently revolving around the moon, Vikram was ejected with the exhaust funnels of its five engines facing the direction of its revolution instead of on the opposite side. At the beginning of the descent phase, its onboard computers ignited four of the five engines to steadily kill its velocity. To ensure that both the craft’s horizontal and vertical velocities were within parameters, all four engines had to fire with perfect synchronicity. If one of the engines deviated, the computer was pre-programmed to use the other engines to provide differential thrusts to correct the anomaly.
The live telecast by Doordarshan showed scientists clapping at the completion of the Rough Braking Phase, indicating it was successful. But some experts believe that there are indications that errors may have been building up in this phase. For while the horizontal velocity (the speed the craft was moving at) was to be around 150 metres per second at the end of the phase, the readings on the large console in the mission operations complex showed that it was around 200 metres per second, faster than what it should have been. On the other hand, the vertical velocity or the speed with which the lander was descending, hovered between 70 metres and 68 metres per second for several seconds.
It was at this point that the second phase, termed the Absolute Navigation Phase (ANP) and lasting around 40 seconds, kicked in. In this phase, Vikram should have corrected any errors in calculations of the key navigation parameters such as its height and velocity during the Rough Braking Phase. It did this by double-checking the readings of its on-board measuring instruments, including cameras photographing the lunar terrain, to measure Vikram’s velocity and height. Variations in the velocity, altitude or inclination of the spacecraft were to be corrected by the autonomous control systems, which arrive at their own logical decisions on the adjustments that need to be made. As a senior scientists put it, The number of exigencies and errors you can calculate and feed into the computer is only limited by your imagination. The best control systems are the ones where scientists let their imaginations run free and plan for as many contingencies as possible.
{Unknown unknowns or planning for uncertainities}
It was at the brief ANP phase that the anomalies in Vikram’s powered descent began to mount. In the control room, the large console simulating Vikram’s descent showed the lander deviating from its 45-degree inclination. It inexplicably executed a somersault, making the engines face upwards instead of downwards (see graphic: 15 Minutes to Despair). One explanation is that the onboard computer was correcting the spacecraft’s attitude to enable the cameras to position it properly for taking the pictures it needed to calibrate vital parameters. But that manoeuvre went haywire and resulted in increasing the vertical descent velocity rather than decreasing it. The other explanation is that the control system noticed a drop in the velocity and corrected it even though it was still within the threshold. In doing so, it first erroneously rotated the craft by 140 degrees to boost the velocity, then reversed it to the original position. By then, the spacecraft had lost its orientation and control.
LOSS OF CONTROL
It was at this point that the third phase, the Fine Braking Speed Phase lasting 90 seconds, began. To bring down Vikram’s horizontal and vertical speeds to near-zero and the craft to an altitude of 400 metres, two of the four engines were to shut down. There is evidence to show that the spacecraft was desperately trying to regain its orientation and was pitching from side to side. The console showed that the vertical speed had increased; it was also at this juncture that all communication with the control room snapped. There was no evidence to show that the two engines had shut down as per plan. All the console showed was that the horizontal velocity was still a high 48 metres per second and the vertical velocity 59 metres per second. Both these key parameters should have been considerably lower for the lander to go into its terminal descent phase. Its speed at this point should have been near-zero and it should have been hovering over the lunar surface at a height of 400 metres. Its onboard cameras were then to take pictures for its control system to check whether the landing site was suitable.
ISRO had decided that Vikram would land near the colder South Pole where water molecules were expected to be found in greater abundance. This was the first time a lander was doing so for good reason, as there are more craters on the lunar poles than its equatorial belt. Vikram’s control system, using its instruments including the cameras, was to ensure that the craft would land on a flat surface. If it landed on any surface that had an incline beyond 12 degrees, it would topple over. Vikram was to then descend to 10 metres before its on-board control system would switch off two engines. The fifth engine located at its base would then be switched on for a controlled descent. All this was to happen if everything had gone well in the earlier phases. But, since communication snapped at a height of 2.1 km, there is no evidence so far to show whether the terminal descent phase was activated or not.
According to experts, Vikram’s abrupt end can be attributed to three major reasons, but they do not quite agree which one was the main culprit. 1) Some believe that the propulsion systems malfunctioned during the transition from the Rough Brake Phase to the Absolute Navigation Phase, when the engines were to fire synchronously to reduce the lander’s speed. Since the throttle-able engines were based on a new technology, there is suspicion that one of them could have misfired, causing unstable conditions beyond the system’s tolerance, and confused the command and control system. Others 2) believe the error lay in the control system itself, with an improper logic built in, that made the lander do a complete turn during the transition between the absolute navigation and fine braking phases. Yet 3) another section of opinion argues that it was a combination of errors in both propulsion and control systems that led to the setback. Meanwhile, ISRO scientists are gathering every bit of data the lander transmitted before its signal was lost. They are using such data to simulate all possible scenarios and explain Vikram’s aborted landing.
........
.
.