The Design and Engineering of Curiosity

by Emily Lakdawalla

  5.2.4 Using the arm

  The arm was tested and qualified to be able to reach targets within a cylinder-shaped region called the “primary workspace” (often referred to as the “magic cylinder”), shown in Figure 5.3.2 But it can readily reach targets beyond that, in a wider region called the “work volume.” The arm functions slowly and deliberately, its tip moving at a maximum speed of 1 centimeter per second. The arm’s heft imposes limits on the accuracy of its placement: gearbox backlash, thermal expansion and contraction, and the massive weight of the turret combine in difficult-to-predict ways with the tilt of the rover to make its position in space uncertain within about a centimeter.3 The uncertainty is slightly larger for the instruments because the amount and direction that they sag on their wire rope vibration isolators depends on the orientation of the turret.

  Figure 5.3. The “magic cylinder” is centered 105 centimeters in front of the rover body, 100 centimeters tall and 80 centimeters in diameter. If the rover is on a level surface, the workspace extends 20 centimeters below the surface. Modified from Billing and Fleischner ( 2011 ). The CAD model of the rover shown here predates several design changes.

  Once the arm has been deployed to a location, it can be repositioned to the same pose over and over again with surprising precision. The precision holds even when a different tool is selected on the turret. Rover planners take advantage of this precision to get closer to targets than the arm’s initial placement inaccuracies allow. The rover planners commonly use the contact plate on APXS to test exactly where a target is, by advancing APXS toward the target and waiting for the contact plate to record contact. In circumstances where the APXS can’t be used in this way (soils, loose rocks, or very uneven surfaces), they can use MAHLI. MAHLI’s autofocus distance is very sensitive to the distance to the target, so arm engineers can use the MAHLI autofocus distance to upgrade their knowledge of the position of the target to get closer on a later sol. Rover planners can even use APXS in “proximity mode” (see section 9.​3.​2) as a test of where the surface lies.

  5.3 THE DRILL

  Curiosity’s drill is a percussion instrument that hammers its rotating bit, boring holes 1.6 centimeters wide and up to 6.5 centimeters deep.4 The drill has four motors: a drill feed mechanism for moving the drill bit up and down; a drill spindle mechanism to rotate the bit; a percussion mechanism; and a drill chuck mechanism that can release the drill bit assembly and exchange it for a new one from one of two bit boxes located on the front of the rover (Figure 5.4).

  Figure 5.4. Parts of the drill. Images from Okon ( 2010 ), annotated by Emily Lakdawalla.

  5.3.1 Drill bit assembly

  The drill bit assembly consists of a drill bit, collection sleeve, and sample chamber (Figure 5.5). The spade-shaped steel bit is a commercial off-the-shelf component that has been modified, with two deep flutes machined into it to help move powdered rock up the collection sleeve and into the sample chamber (Figure 5.6). The steel collection sleeve covers all but the last 1.5 centimeters of the drill bit. After 16 drill sites, Curiosity is still using the original drill bit. Although it is not as shiny as it once was, it has not dulled dramatically (Figure 5.6).

  Figure 5.5. Parts of the drill bit assembly. Photos from Anderson et al ( 2012 ), annotated by Emily Lakdawalla.

  Figure 5.6. Condition of Curiosity’s drill bit over time as observed using ChemCam RMI. Top row: sol 172, before the first drill site at John Klein. Bottom row: sol 1528, after drilling at Sebina. NASA/JPL-Caltech/CNES/CNRS/LANL/IRAP/IAS/LPGN.

  5.3.2 Drilling

  To prepare to drill, Curiosity places two projecting contact sensors against the rock, and then continues to drive the arm motors even after the contact sensors are in contact with the rock target (Figure 5.7). This is called “preloading”; the arm can press onto the rock with up to 300 newtons of force. Front Hazcam images taken before and after preloading usually document a tiny upward motion of the rover body as the arm pushes against the rock. Once the arm is placed and preloaded for drilling, the arm doesn’t move; all drilling motion is performed by the drill itself using the drill feed mechanism.

  Figure 5.7. Drill use on Mars. The two contact sensors/stabilizers are pressed against a rock, and the drill feed has extended to place the drill in contact with the ground. Mastcam image 0174ML0006380000105184E01, NASA/JPL-Caltech/MSSS/Emily Lakdawalla.

  Once the drill feed mechanism has advanced the drill bit to contact the rock, drilling begins with percussion from a 400-gram mass striking a spring-loaded anvil rod 30 times per second. The team can select the initial energy of the blows within a range from 0.05 to 0.81 joules. At first, the drill uses no rotation, only percussion, to strike a small asterisk-shaped divot in the rock at the location of the desired drill hole – like a carpenter using a nail or awl to set the starting location for their drill. These initial taps dig no more than 0.8 millimeter into the rock.5

  For the first 1.5 centimeters of drilling, the powdered material piles up around the drill hole, making a tailings pile. After the first 1.5 centimeters, the collection tube contacts the surface, and powdered material that passes by the spade tip of the bit climbs up the auger into a two-chambered sample collection area within the drill bit assembly. After drilling, the feed retracts and the arm lifts the drill off the rock. The rover then photographs the drill bit with Mastcam to verify that it is none the worse for wear after the drilling activity.

  At many sites, Curiosity performs a “mini-drill” test before the full drill, penetrating less than the unsleeved 1.5 centimeters into the rock, in order for the tactical team to assess rock and rock powder properties before committing to gathering powdered rock sample. A rock with an extremely unusual water-rich mineral composition could liquefy under the vibration of the drilling mechanism, which would be catastrophic for the ability to acquire samples. The team can choose to skip mini-drilling to save time if they determine from a rock’s appearance (from Mastcam, MAHLI, and ChemCam RMI) and composition (from ChemCam LIBS and APXS) that it is similar to previously drilled rocks.

  5.3.3 Drill bit assembly replacement

  What if the drill bit gets worn out, or worse, stuck in a rock? If the rover slips during drilling, it could leave the drill bit stuck. To avoid the situation, before drilling, the rover drivers make sure that the rover is in a stable position, with all wheels firmly in contact with the ground, and no small rocks under the wheels. If there is any question of wheel stability, they may sequence a set of MAHLI wheel images in order to be sure the wheels are stable. If the rover should slip, binding the drill bit, the drill feed mechanism is capable of pulling upward with a force of nearly 10,000 newtons.6 If the drill remains stuck, they can try pulling the feed while percussing and/or rotating, which would reduce the friction between the drill and the rock but could also result in the loss of the acquired sample. If the drill bit remains stuck after that, they can try motion of the arm to counteract whatever motion of the rover had caused the drill bit to bind.

  If all of these efforts fail, the rover can detach its bit and leave it behind in the rock, exchanging it for one of two more bits located in bit boxes on the front of the rover (Figure 5.8; another good view of a drill bit box is in Figure 5.18). Because the drill bit has not yet needed swapping, the drill chuck mechanism has not been used since a brief test wiggle in the first weeks after landing.7

  Figure 5.8. A test of the positioning of the turret for drill bit replacement. If the drill bit really were being replaced, Curiosity would already have used the drill chuck mechanism to release the drill bit it is holding. Four round-tipped posts surrounding the drill bit assembly would advance into four funnels at the corners of the bit box, aligning the drill with the replacement drill bit assembly. To the right of the turret is the sample playground. Navcam image NLA_400696022RAS_F0040000NCAM00110M1 taken sol 34. Inset: a view of the interior of the bit box taken by MAHLI from a similar position. Image 0036MH0000490010100065E01. NASA/JPL-Caltech/MSSS/Emily Lakdawalla.

  5.3.4
Drill problems

  Several issues have affected the drill both before and after launch. One was the potential contamination of the drill bit that caused the reclassification of MSL’s planetary protection status (see section 1.​7.​4). The others are: the possible presence of Teflon debris in drilled samples; a short in the percussion mechanism; and a serious problem with the drill feed mechanism. A new problem was diagnosed in the drill chuck mechanism as this book was going to print around sol 1800. It is similar in character to the drill feed problem. It will not be further discussed here.

  5.3.4.1 Teflon debris

  Shortly before launch in November 2011, engineers doing testing of drilling operations found that seals inside the engineering model of the drill bit assembly were slipping during drilling, which generated Teflon debris that mixed with the drilled rock powder.8 This was a potentially serious source of contamination that could compromise SAM’s ability to detect organic materials within Mars’ rocks. Although the possibility of Teflon contamination of the drilled material had been recognized early in development, when the drill was actually tested, it generated more debris than expected. It was too late in the process to effect any kind of design change, of course. The mission ultimately determined that the amount of contamination was small enough that it would not likely affect SAM results, and suggested that the mission avoid using the drill in a way that generated the most debris – “minimizing the low-rate-of-penetration operations.” No sign of Teflon contamination has been noticed in drilled samples since landing.

  5.3.4.2 Battle short and the sol 911 percussion anomaly

  Another potentially serious problem was discovered during Earth testing of a testbed version of the drill mechanism in 2011. A broken bushing caused a short circuit in the test drill that could have fried the rover’s motor controller if engineers had not acted swiftly. The consequences of such an event happening on Mars would be dire. It was too late to make any changes to the flight drill. Engineers in Florida opened the belly pan of the rover to install a “battle short” that would route half of the excess current to ground if such a short circuit developed in flight.9

  On sol 911, sensors detected current flowing through the battle short as Curiosity was using drill percussion to transfer sample from the drill to CHIMRA, halting the operation.10 There is no way to know if the cause is the same as the problem discovered on Earth, but the effect is similar. The shorts have recurred since sol 911, but are intermittent and extremely brief. If they remain that way, the battle short adequately protects the electronics. The engineers have instructed the rover to tolerate very brief shorts without faulting and terminating the drilling process.11 At the same time, the mission has shifted to relying less on the percussion mechanism. They now avoid using drill percussion for sample transfer, relying on CHIMRA vibration. They have also changed the way they operate the drill: originally, they began drilling with a medium percussion level and made adjustments according to the penetration rate, but they now begin with very light percussion and only increase the rate as needed. Engineers have also developed a new rotary-only drilling technique, made possible by the softness of the rocks within Gale crater, but rotary-only drilling has not yet been used on Mars because of a different drill anomaly.

  5.3.4.3 Sol 1536 drill feed anomaly

  On sol 1536, the engineers attempted rotary-only drilling at a site called Precipice. The operation did not complete, because the drill feed mechanism stalled immediately. Current flowed to the drill feed motor, but the motor produced no motion. Like the problem with the percussion mechanism, it is intermittent, so has been difficult to troubleshoot, but it appears to reside in the drill feed brake mechanism. As of sol 1800, the rover hasn’t done any drilling.

  The drill feed motor has a power-off brake: when no electricity is flowing to the brake, a disk (the “moveable brake”) is pressed against another disk (the “fixed brake”) by a set of springs. The pressure holds the drill feed firmly in position even when percussion, vibration, and rotation mechanisms are operating. Energizing a solenoid pulls the moveable brake away from the fixed brake, allowing the drill feed motor to spin a worm drive that slowly translates the drill feed out or in. The brake has two solenoids for redundancy.12

  Engineers troubleshooting the issue found that energizing either solenoid with the normally commanded current failed to produce any feed motion. Commanding with tweaked parameters (like higher current, energizing both solenoids instead of one, multiple attempts to disengage the brake, and so on) produced some motion, but not reliably. The team strongly suspects that a displaced component or piece of foreign debris is interfering with motion of the movable brake, preventing it from fully disengaging when commanded.

  From December 2016 through March 2017, engineers tested and performed diagnostics in an attempt to recover the full use of the drill feed. After developing several innovative techniques, they achieved the full range of feed motion, albeit at speeds too high to drill into rocks. However, after using CHIMRA to sieve a sand sample at Ogunquit Beach on sol 1651 (March 29, 2017), engineers found that the behavior of the drill feed had deteriorated.

  As of this writing, the engineering team is pursuing a new drilling and sample delivery approach that does not require using the drill feed. They successfully extended the feed to its full 110-millimeter distance on sol 1780. On Earth, they are working on developing the ability to perform feed-extended drilling (FED), using arm motion instead of feed motion to advance the drill bit into the rock. Initial testing of feed-extended drilling began on Mars on sol 1848. While this can recover the ability to drill, not using the feed also prevents transfer of sample material to CHIMRA (see section 5.4.2.1). Future feed-extended sample transfer (FEST) may involve reverse augering material from the sample chamber out through the bit and directly into SAM and CheMin.

  5.4 CHIMRA: COLLECTION AND HANDLING FOR IN SITU MARTIAN ROCK ANALYSIS

  CHIMRA (pronounced “chimera”) is a labyrinth of chambers that can sieve and portion out samples for delivery to the science instruments.13 The main parts of CHIMRA are shown in Figure 5.9. There are two main paths by which sample moves around inside CHIMRA: one with a 150-micrometer sieve, and another with a 1-millimeter sieve.

  Curiosity can acquire sample material either through drilling or through scooping loose material with the CHIMRA scoop. CHIMRA uses a combination of gravity and vibration to move sample around: the rover rotates the turret into a direction where the desired direction of sample motion is downward, and then uses its vibration mechanism to encourage the powder to move. CHIMRA’s labyrinthine interior is difficult to imagine even for the engineers who interact with it on a regular basis. Four 3D-printed models of CHIMRA located throughout mission operations enable engineers to twist and turn it and open and close its doors to simulate its movements physically.

  Figure 5.9. Parts of CHIMRA. Left Mastcam image 0032ML0000830000100870E01 of turret from initial checkout on sol 32. NASA/JPL-Caltech/MSSS/Emily Lakdawalla.

  5.4.1 CHIMRA tour

  Engineers designed CHIMRA to avoid clogging. Its interior spaces are as wide open as possible, without sharp corners. Wherever possible, the design avoids forcing sample to move through a narrower space than it has already passed through. The mechanism was also designed to allow engineers to visually inspect every surface within CHIMRA repeatedly over the course of the landed mission.

  CHIMRA has four motorized mechanisms: the vibration actuator, the portion door actuator, and the primary and secondary thwack actuators. The vibration actuator is a self-contained mechanism that rotates an off-center tungsten mass to generate vibrations. It generally vibrates at a speed that encourages the CHIMRA mechanism to resonate, which efficiently shakes the 8-kilogram CHIMRA on its mount while not wasting much energy vibrating the rest of the 34-kilogram turret. The portion door mechanism is a very small motor that rotates a lever that presses up against the open end of the hole out of which CHIMRA drops 150-micrometer-sieved portions. The thwack mechanisms both serve
multiple functions. Each of the two thwack mechanisms is connected to a door that opens up CHIMRA for inspection, sample dumping, and cleaning, and a “thwack arm” that carries a sieve. The primary thwack mechanism is connected to parts of the 150-micrometer sieve path (section 5.4.2). The secondary thwack mechanism is connected to parts of the 1-millimeter sieve path, including the scoop (section 5.4.3). Both can be wound up with a spring to slam the sieve against the rest of the mechanism to clear stuck sediment, hence the “thwack” moniker (section 5.4.4).

  5.4.2 CHIMRA 150-micrometer sample pathways

  This pathway can generate individual sample aliquots amounting to about 75 cubic millimeters each for delivery to SAM or CheMin, or a single “portion plus” aliquot of (very approximately) three times that size.

  5.4.2.1 Drill to CHIMRA reservoir

  After Curiosity has drilled a sample, the sampled powder sits in the forward sample chamber, immediately above the drill bit. The drill reservoir is two-chambered so that the drill can be used at angles of up to 20° without sample spilling out of the sample exit tube prematurely, regardless of drill orientation. Once the drill feed is fully retracted, the drill bit assembly sample exit tube aligns with the CHIMRA sample inlet tube. In the aftermath of the drill feed anomaly described in section 5.3.4.3, this is an important detail. If the drill feed is not available, the only way to transfer material from the drill to CHIMRA will be by dumping the drilled material somewhere and picking it up again with the scoop, a difficult or perhaps impossible proposition.