William O'Neil managed the Galileo Jupiter mission at the Jet Propulsion Laboratory (JPL) from 1990 to 1998, but he had cut his teeth on the Surveyor lunar landings and the Mariner 9 and Viking Mars missions of the 1960s and 1970s. It was thus appropriate that he became Chief Technologist for JPL's Mars Exploration Program after he stepped down as Galileo manager.
One of O'Neil's first initiatives in his new role was a pair of workshops aimed at generating fresh ideas for JPL's Mars Sample Return (MSR) mission, which had become mired in fiscal and engineering problems. In its April 1998 iteration, a single launch vehicle boosted the MSR spacecraft to Mars in late 2004. By the start of the first MSR workshop in July, a redesign effort begun in June had yielded a baseline MSR plan that split the orbiter and lander between two rockets launched in August and September 2005. The lander would carry the Mars Ascent Vehicle (MAV), which would launch the Mars sample to orbit for retrieval by the orbiter and return to Earth. The 512-kilogram liquid-propellant MAV was overweight, contributing to mass limitations which meant that only a small, short-range sample-collecting rover could be included in the mission.
In his presentation to the first MSR workshop, Brian Wilcox, a JPL rover engineer and former model rocketry enthusiast, described a possible alternative to the baseline mission's liquid-propellant MAV based on the U.S. Navy's 1958 PILOT microsatellite launcher design. His "MicroMAV" was a 20-kilogram solid-propellant rocket with no moving parts. Wilcox noted that, unlike liquid propellants, solid propellants would not freeze during the frigid martian night.
Wilcox envisioned an MSR similar to one he proposed in 1989, in which a rover with six wheels and a top-mounted solar array would carry the three-stage MicroMAV with it while exploring Mars. The MicroMAV would ride slung horizontally along one of the rover's sides. The rover would collect an unspecified quantity of rocks and dirt and load them into the sample canister in the MicroMAV's third stage, then would pivot the MicroMAV onto the top of the solar array and point its nose skyward. The MicroMAV would then ignite its first stage motor.
The first stage, which would loft the MicroMAV above most of Mars's atmosphere, would have a total mass at ignition of 9.75 kilograms, of which 7.8 kilograms would comprise solid propellant. It would include four fins and a horizon sensor. The fins would be canted slightly so that the thin martian air rushing past them during ascent would spin the MicroMAV about its long axis for gyroscopic stabilization.
After first stage burnout, the MicroMAV would coast upward, still spinning about its long axis. As it neared the top of its trajectory, its nose would begin to tip downward toward the horizon. The horizon sensor would alternately "see" the sky above and the ground below.
When the sensor tallied a pre-set number of rotations, it would trigger second stage ignition. This would also discard the first stage. The second stage, which would supply most of the MicroMAV's orbital velocity, would have a mass of 9.4 kilograms with 7.8 kilograms of propellant. After second stage burnout and separation, the MicroMAV third stage would be in Mars orbit; its periapsis (orbit low point) would, however, remain within Mars's atmosphere. Second stage burnout would thus trigger a timer designed to ignite the third stage motor.
The tiny 0.85-kilogram third stage would include 0.05 kilograms of propellant and the Mars sample. During first and second stage flight, its motor would point forward. Because it would be spinning like a gyroscope, it would remain pointed in the same direction following second stage separation. This would mean that, half a Mars orbit later, the motor would point away from its direction of motion. At that same moment, the MicroMAV would attain apoapsis (orbit high point) and the timer would reach zero. The third stage engine would then ignite to raise the MicroMAV's periapsis to a safe altitude.
Third stage ignition would also ignite a "pyrotechnic layer" that would heat its exterior "white-hot for an instant." This would destroy any martian microbes that might have hitched a ride on the third stage and would also solder shut the sample canister to prevent the escape of any contaminants inside.
The grapefruit-sized MicroMAV sample canister would be entirely passive, with neither a radio beacon nor a flashing light to aid the orbiter in locating it. The orbiter would begin looking for the MicroMAV from a position about 100 kilometers above its orbit. For 18% of its orbit, the canister would be sunlit but set against Mars's nightside as seen from the orbiter. At such times, the orbiter would point its wide-angle imager toward the MicroMAV's predicted position and image the area several times to enable controllers on Earth to determine the MicroMAV sample canister's orbit. Wilcox estimate that controllers using orbiter images would need no more than 31 hours to locate the MicroMAV.
The MicroMAV concept excited much interest among JPL engineers. Though further study revealed the MicroMAV MSR scenario to be unworkable in the form Wilcox described - for example, JPL quickly abandoned rover launch in favor of a more conventional launch from a fixed lander (image above) - the concept of a simplified solid-propellant MAV profoundly influenced subsequent JPL MSR planning.