The landing engineers worked on fiddly details of the order of events during landing, finding ways to shave off risk here and there. For example, after spending time with models and empirical tests of mobility system deployment, they shifted the event until quite late in the landing sequence, during the sky crane phase. Figure 1.16 depicts the entry, descent, and landing sequence in its final form.
Figure 1.16. Schematic diagram of approach, entry, descent, and landing. From the Mars Science Laboratory launch press kit, November 2011. Minor changes from previous diagrams include a very late deployment of the wheels and an emphasis on the lengthy period of near-horizontal flight during the “hypersonic aero-maneuvering” phase.
Engineers on the landing team performed simulation after simulation, testing their landing system against a huge variety of landing-day scenarios, increasing their confidence in their system.
A fourth landing site selection workshop took place on September 27, 2010, to discuss the merits of the final four sites in the context of the new scientific results from Mars Reconnaissance Orbiter. All four sites remained possibilities.
Not all the development news was good. The Department of Energy measured the power output of the stored MMRTG on a quarterly basis. As of late summer 2009, the Department of Energy predicted that the MMRTG would be producing 110 watts of power on landing day, and planning proceeded according to that assumption. But the next measurement, taken in late 2009, revealed that the MMRTG was under-performing. Now they had two problems: the lower-than-predicted power levels and the mystery of why their modeling was not working.
Even before the delay, there had been insufficient power to run the rover. The delay gave them time to address the power problems by replacing the rover’s battery with one that had double the original capacity. Finding space for the new battery in the cramped interior of the warm electronics box required them to partially disassemble the rover. But the battery upgrade gave the rover about 1600 watt-hours of usable capacity, and ensured that power was rarely a limiting factor in running the rover for any but the most power-hungry activities, at least during the prime mission.54
1.7 FINAL PREPARATIONS (2010–2011)
1.7.1 ATLO, again
The redesigned motors finally began to arrive at JPL in 2009, with the flight models being delivered in 2010. Assembly of the rover began in June 2010 and proceeded rapidly. The legs and wheels were attached at the end of June, as was the mast with all its cameras and ChemCam (Figure 1.17). The rover first spun all six wheels on July 9, and drove on its wheels for the first time on July 23 (Figure 1.18). The arm was attached in August, and put through its paces with the rover sitting at tilts of up to 20° (Figure 1.19).
Figure 1.17. Mobility system installation, 29 June 2010. NASA/JPL-Caltech release PIA13234.
Figure 1.18. The rover’s first drive, 23 July 2010. The corner wheels are steered into position to allow the rover to turn in place. NASA/JPL-Caltech release PIA13308.
Figure 1.19. Testing out the arm’s positioning from a tilted rover, 16 September 2010. NASA/JPL-Caltech release PIA13389.
With the spacecraft finally coming together, systems engineers could begin to test how all the subsystems functioned together – and make sure that the spacecraft safely handled problems. Engineers performed a series of system tests, each focused on a different mission phase. In parallel, engineers took all components of the spacecraft into JPL’s thermal vacuum chamber to perform environmental tests. The chamber simulates the harsh environments of deep space and the Mars surface (Figure 1.20). It can be pumped down to near-vacuum, cooled with liquid nitrogen, and lit with thirty-seven 25,000-watt xenon lamps to simulate direct solar heating in deep space.55
Figure 1.20. Thermal vacuum testing. Top: The cruise stage and backshell, August 24, 2010. Bottom: The rover during Surface Test 8, March 8, 2011. NASA/JPL-Caltech releases PIA13359 and PIA13806.
Despite various hiccups, the tests were mostly successful, demonstrating that the spacecraft could function through entry, descent, and landing, and in surface operational scenarios. The final system test, which took place after the spacecraft had been shipped to Florida, simulated three days of typical science activities on the surface, proving that the flight system was ready for Mars. But only just. The tight schedule and complexity of interacting systems meant that the system software was not particularly mature.
Public information officers began to stir up public interest in the mission. On June 24, 2011, NASA released a computer animation depicting the launch, cruise, and landing of the rover. The animation had been directed by JPL visualization artist Doug Ellison and produced by animator (and former JPL artist) Kevin Lane, who packed the thrilling video with accurate details of the landing sequence, informed by research and interviews with engineers (Figure 1.21).
Figure 1.21. The moment of touchdown as depicted in the landing animation. NASA/JPL-Caltech release PIA14840.
1.7.2 Going to Gale
The fifth and final community workshop on landing site selection opened on May 16, 2011. Engineers remained confident that their entry, descent, and landing system could handle any of the four landing sites under discussion. Scientists pored over the data collected by Mars Reconnaissance Orbiter and Mars Express on the four sites and debated the scientific questions that could be answered at each one. The workshop produced no specific recommendation, but gave the MSL science team the background they needed to discuss the merits of the four landing sites at a team meeting on May 18.
Based on the wider community and narrower team discussions, the project science group (a committee of all of the instrument principal investigators) selected Gale crater as its top choice. In addition to its compelling scientific targets of layered rocks and alluvial fans, it was also the lowest-elevation and most equatorial of the four proposed sites, the most benign landing case. It had some possibility for science within the landing ellipse, but most of its science would be on a go-to mission, driving beyond the ellipse to reach mineral-bearing rocks at the base of the crater’s central mountain (Figure 1.22).
Figure 1.22. Oblique view of Gale crater. The black landing ellipse is 25-by-20 kilometers in extent. The blue line shows a possible route for Curiosity to ascend the mountain. NASA/JPL-Caltech/ESA/DLR/FU Berlin/MSSS release PIA15292.
The project science group recommended Gale and Eberswalde as potential landing sites to NASA, preferring Gale. The final decision would be made at NASA Headquarters. The landing site selection committee gave a presentation to an independent landing site certification board on June 9 and 10, 2011. The NASA Planetary Protection Office reviewed the four sites for compliance with planetary protection provisions on June 23. The next day, the MSL project presented the four sites to NASA Headquarters representatives, who concurred with the downselection to Eberswalde and Gale. They reconvened a month later and recommended Gale. NASA announced the selection of Gale crater as Curiosity’s landing site on July 22, 2011.
1.7.3 Journey to Florida
As Earth crept toward Mars on its inside track around the Sun, the once-every-26-months launch period approached. NASA had missed the intended window in 2009; they could not afford to miss the 2011 window. Between May and November 2011, all of the pieces of the interplanetary spacecraft would come together at Kennedy Space Center in Florida.
On May 12, 2011, a C-17 transport plane carried the heat shield, backshell, and cruise stage from California to Florida. A scary event happened during processing of the spacecraft on May 20. A crane operator accidentally lifted the backshell and the support cart that the backshell was attached to off of the ground, exerting unplanned loads on the backshell. There was no obvious damage, but engineers scrambled to figure out whether the weight of the cart suspended from the backshell could have damaged it. They determined that the backshell was designed robustly enough to withstand the additional load of the cart, and assembly could proceed as normal.56 But the incident ratcheted up the tension on the launch preparations.
Meanwhil
e, tests continued on the rover back at JPL at a feverish pace. It was finally buttoned up and transported to Florida, along with the descent stage, on June 23. The Department of Energy delivered the MMRTG on June 30. Components of the Atlas V rocket arrived at Kennedy aboard the Delta Mariner barge a month later.
1.7.4 Planetary protection jeopardy
Problems with the drill bits dogged the mission right up until launch. Early in 2011, the project discovered that residue from oil used in manufacturing the test drill bits was contaminating drilled samples. If there was similar oil on the flight drill bits, it could invalidate SAM’s attempts to search for Martian organic compounds. By March, they had figured out how to re-manufacture the drill bits, and planned to deliver them just in time for the rover to be shipped to Florida.57
The SAM team had to know whether the flight drill bits also carried traces of hydrocarbons. In Florida, a team of contamination control engineers opened the sterile bit enclosures to check for the oil residue. They found no contamination – but in checking for oil contamination, they had broken the sterile seal on the bits, and the bits were not re-sterilized. Even if they had not checked for the oil contamination, they would have had to break the sterile containment of one of the bits to install it on the rover, because of a change in mission plan. Originally, Curiosity was to have launched to Mars with all three bits enclosed in sterile boxes, and only break that seal upon arriving at Mars and loading the first bit in the drill. The mission decided it was too risky to launch without one bit preloaded. That bit was not in sterile wrapping throughout the rover’s delivery to Florida and assembly operations, in violation of the mission’s planetary protection plan.
The wheels represented another vector by which Earth microbes could contaminate Mars. They were heat-sterilized and wrapped to prevent recontamination during final assembly. But the planetary protection requirements weren’t specific about when it was permissible to unwrap the wheels, with ambiguous requirements like “The flight wheels shall be mounted on the rover as late as possible and will be covered as much as possible to prevent recontamination.”58 Photos from the Kennedy Space Center show that the wheel wrappings were removed just after delivery, on June 27. The final system tests were performed in July and August with the wheels unwrapped and resting on the floor (Figure 1.23). They were re-wrapped before the spacecraft was stacked to the descent stage and backshell in late September, and the wrappings removed for the final time before heat shield encapsulation in early October.
Figure 1.23. Engineers at Kennedy Space Center test out the rover’s motor functions, July 18, 2011. NASA/KSC release KSC-2011-5919.
The lack of sterile containment of both wheels and drill bits was “a clear violation” of the mission’s planetary protection requirements.59 The Office of Planetary Protection did not learn of the violation until less than three months before the launch, in mid-September. On September 20, Peter Theisinger formally requested permission to deviate from the requirements, and the Office of Planetary Protection rejected that request. The violation of planetary protection requirements had the potential to jeopardize Curiosity’s on-time launch.
Not enough time remained to remove, sterilize, and replace the drill bits and wheels before the launch. The only way out was a bureaucratic one: to reclassify the mission to a lower planetary protection standard. Fortunately, with its low elevation and equatorial location, Gale crater is a site where no ground ice was expected to be anywhere close to the surface, so a landing failure had no potential to create a spacecraft-induced special region (see section 1.4.5). That made it possible to reclassify the mission from planetary protection category IVc down to category IVa, reserved for “lander systems not carrying instruments for the investigations of extant martian life” that will not come into contact with any special region.60
Catherine Conley, NASA’s planetary protection officer, formally recategorized the mission to category IVa on November 1. “Violation of NASA planetary protection requirements represents a significant risk to the MSL project,” she wrote in a letter to mission and NASA management. “Requesting a planetary protection deviation of this magnitude, so late in the project lifecycle, is improper.” She singled out the wheels as representing the greatest risk for contaminating Mars, and imposed restrictions on the rover's future exploration activities:
The project is prohibited from introducing any hardware into a Mars Special Region, as defined in NASA Procedural Requirements document NPR 8020.12D. Fluid-formed features such as Recurring Slope Lineae are included in this prohibition. Any evidence suggesting the presence of Special Regions or flowing liquid at the actual MSL landing site shall be communicated to the Planetary Protection Officer immediately, and physical contact by the lander with such features shall be entirely avoided.
1.7.5 Final assembly
Finally, the time came to put together all the pieces of the puzzle. The rover and MMRTG met for the first time for a fit check on July 12. The MMRTG produces a prodigious amount of heat, so it wasn’t safe to install it permanently until the last possible moment. Handlers removed the MMRTG to storage in its own cavernous, climate-controlled room.
Assembly was interrupted in September by an emergency situation discovered during drill testing back at JPL. The drill percussion mechanism developed a short circuit that could damage the rover’s electronics if it occurred on Mars, jeopardizing the mission. The engineers developed a solution quickly, but implementing the solution required opening up the rover’s belly pan and adding a new wire to ground the rover’s power bus. This “battle short” wouldn’t prevent future shorts in the drill, but would protect the rover’s electronics if it happened again. The project agreed to take the risky step of performing surgery on the rover just weeks before launch.61 It turned out to be a wise decision, as the drill percussion mechanism has indeed experienced shorts on Mars (see section 5.3.4.2).
Stacking of the spacecraft components began inside the Payload Hazardous Servicing Facility on September 23, with the connection of the descent stage to the rover and then the backshell. They topped the stack with the cruise stage on October 10, and lifted the stack onto the base of the heat shield on October 11, completing the assembly of the spacecraft, except for the MMRTG (Figure 1.24).
Figure 1.24. On October 11, 2011, the spacecraft stack was completed. NASA/KSC release KSC-2011-7350.
Two weeks later, with the spacecraft flipped upside down, they enclosed the saucer-shaped craft inside the fairing that would protect the spacecraft during its trip through Earth’s atmosphere (Figure 1.25). MSL needed the full width of the Atlas V’s largest, 5-meter fairing, but little of the length; most of the interior of the tall fairing remained empty. They delivered the spacecraft in its nose cone to the launch pad on November 3, then hoisted it atop the rocket (Figure 1.26).
Figure 1.25. MSL is dwarfed by its fairing, October 25, 2011. NASA/KSC release KSC-2011-7530.
The final step in assembly took place at the top of the tower just a week before launch. The MMRTG was finally installed on November 17 (Figure 1.26). A hatch in the fairing, and a matching hatch in the aeroshell, allowed technicians access to insert the MMRTG, and then to sew on the cloth windbreak over the MMRTG’s cap (Figure 1.27). With the MMRTG in place, cooling the spacecraft became a top priority. An air conditioning system in the launch tower blew chilled air through an inlet in the fairing onto the cruise stage radiators, helping to dissipate the heat for the week that led up to launch.
Figure 1.26. Workers lift the MMRTG, in a protective cage, to the top of the Atlas V rocket in its launch tower on November 17, 2011. NASA/KSC release KSC-2011-7836.
Figure 1.27. Engineers work through a hatch in the rocket fairing and a second hatch in MSL’s backshell to install the MMRTG onto the rover, November 17, 2011. NASA/KSC release KSC-2011-7900.
REFERENCES
Benardini J et al (2014) Implementing planetary protection measures on the Mars Science Laboratory. Astrobiology 14:27–32, DOI: 10.1089/ast.2013.
0989
Billing R and Fleischner R (2011) Mars Science Laboratory Robotic Arm. Paper presented to the 15th European Space Mechanisms and Tribology Symposium, Constance, Germany, 2011
Boeing Rocketdyne Propulsion and Power (2003) Boeing to Build Space-borne Power Generator. Press release dated 1 Jul 2003
Boynton W et al (2002) Distribution of hydrogen in the near surface of Mars: Evidence for subsurface ice deposits. Science 297:81–85, DOI: 10.1126/science.1073722
Caffrey R et al (2004) Initiating the 2002 Mars Science Laboratory (MSL) Focused Technology Program. Paper presented to the 2004 IEEE Aerospace Conference, Big Sky, Montana, USA, 6–13 Mar 2004, DOI: 10.1109/AERO.2004.1367650
Conley C (2011) MSL deviation request. Letter to Peter Theisinger dated 1 Nov 2011
Cook R (2009) MSL Technical and Replan Status. Presentation to the NASA Planetary Science Subcommittee meeting, Washington, DC, USA, 9 Jan 2009
The Design and Engineering of Curiosity Page 7