 ICEBREAKER-3 MARS-ANALOG SAMPLE ACQUISITION TESTS. B. Glass1, B. Mellerowicz2, A. Davis3, K. Zacny2, C. McKay1, A. Dave1, P. Lee1. 1NASA Ames Research Center, Moffett Field, CA 94305, USA, Email: brian.glass@nasa.gov, 2Honeybee Robotics, Pasadena, CA 91103, USA, 3California State Polytechnic University, Pomona, CA 91768, USA.  Abstract: Looking for organics, biomarkers and signs of past or extant life on Mars will require sample acquisition there below the desiccated and irradiated surface.  A decade of evolutionary development of integrated automated drilling and sample handling at analog sites and in test chambers has made it possible to go deeper through hard rocks and ice layers on Mars. The latest Icebreaker-3 drill has been tested with a Phoenix mockup and at the Haughton Crater Marsanalog site in the Arctic (with sample transfer arm) and in a Mars chamber, with successful sample acquisition under  automated control.   Introduction: Exploring and interrogating the shallow subsurface of Mars from the surface will require some form of excavation and penetration, with drilling being the most mature approach. In 2013 the Icebreaker-2/LITA drill, designed for use at the Atacama Desert Chilean Mars-analog site, was tested at the Haughton Crater Mars-analog site [1]. Unlike previous tested prototypes since 2004, the 10kg/1m depth LITA did not demonstrate a capability of penetrating hard rock or ice-consolidated material. Earlier Marsdrill prototypes could penetrate these materials, but are too massive to fly on a Phoenix-sized payload [2] (Table 1).   Unlike terrestrial drills, Mars exploration drills must work dry (without drilling muds or lubricants), blind (no prior local or regional seismic or other surveys), and light (very low downward force or weight on bit, and perhaps 100W available from solar power).  Given the lightspeed transmission delays to Mars, an exploratory planetary drill cannot be controlled directly from Earth. Drills that penetrate deeper than a few cm are likely to get stuck if operated open-loop (the MSL drill only goes 5cm, and the MER RATs 5mm by comparison), so some form of drill automation is required. [3] Icebreaker-3 Approach: While the earlier LITA drill worked well in the Atacama desert [4] and in lab tests, it did not perform satisfactorily in Haughton Crater's impact-breccia permafrost [1].  A stiffer drillstring material was needed to improve rotarypercussive operation in when there was significant torsional drag on the drillstring. Post-test analysis of the LITA drill performance issues in ice-cemented ground and ice lenses indicated that a stiffer shaft design with more drill-head rotary power (applied torques) would be needed to break those layers and overcome parasitic auger drag on the sides of the drill string.  The ability to reverse-rotate would be also needed in order to extricate the drill by backing out when sensors indicate that a jam or sticking is imminent. And a temperature sensor in the drill bit would be very useful to monitor when the frictional heating of cuttings is approaching the triple point (and hence sample alteration, and Phoenix-like clumping of materials). Figure 1 shows the design changes.   The resulting Icebreaker-3 drill, completed in June 2014, added back 2 kg mass (to 12 kg) compared to LITA, but was still 3-5x less massive than earlier prototypes.  Power consumption is comparable to LITA (typically 30-40 W, 200W max during 5-10 min drilling sequences).  It was tested at the Haughton Crater analog site in the Canadian Arctic, as have a series of drill prototypes (shown in Table 1). Study Objectives: Objectives for the Icebreaker-3 (IB-3) drill tests in 2014 were: (a) to confirm nominal operations and the efficacy of the design changes under lab conditions, (b) test drilling operations and be Fig. 1. Icebreaker-3  changes to the LITA drill design, addressing the 2013 field test results [1]. Table 1. Mars-prototype drill tests at Haughton Crater.  havior under the same permafrost conditions at Haughton Crater that hampered the LITA drill, and confirm satisfactory performance there with the IB-3, (c) confirm the interoperability of the IB-3 with a sample transfer robotic arm (into instrument inlets [5]), and (d) test the IB-3 in a Mars chamber to confirm performance and to study the grain size of the icy cuttings produced (see [6] for that description and results).  Objectives for the Haughton Crater tests in 2014 were to test the IB-3 drill in frozen impact breccia; to surpass the maximum single-hole depth drilled by the earlier LITA Mars-prototype drill design (92 cm); to demonstrate the expected fault modes of this drill, for use in developing failure detection and automated control software; and to compare the reliability, required energy and the downward forces needed to make headway, compared with other drill designs (Table 1). For initial lab testing of IB-3 prior to Arctic deployment in June 2014, a Phoenix deck mockup platform at Honeybee Robotics was set up adjacent to the drill with a sample transfer arm and a mockup SOLID instrument with an inlet target for automated delivery of sample. Fig. 2(a) shows IB-3 deployed over the deck alongside integrated sample transfer arm.  The Haughton Crater deployment is shown in Fig. 2(b).  Field Test Site: The Haughton Crater planetaryanalog drilling site is a high-fidelity analog for Mars landing sites with subsurface ice (as at the Martian higher latitudes) and the broken, depth-graded textures similar to impact regolith. The active-layer boundary in 2014 at Drill Hill was observed to be at 51cm depth.  Results: The IB-3 drill was tested at the Haughton Crater site in August 2014. IB-3 drilled >5m, in six boreholes, and with sufficient power (torque) and shaft stiffness to break and penetrate hard rock or iceconsolidated material (a typical "bite" shown in Figure 3). IB-3 drilled >2m in ice or ice-consolidated material (compared to 30cm by the LITA drill in 2013). Unlike prior prototypes (Table 1) IB-3 drilled rapidly and experienced almost no fault conditions. Prior meltrefreeze and binding faults seen often with LITA were not experienced by IB-3. A corkscrewing fault recovery upon drill retraction is shown in Figure 4.  Interoperability testing at Honeybee in June 2014 demonstrated that the IB-3's sample (cuttings) was handed-off from the drill to a sample transfer arm and thence to instrument inlets ("dirt-to-data" from [5]).   Fig. 4. Corkscrewing fault recovery on 2 August 2014 at Haughton Crater by Icebreaker-3 drill. Conclusions: The very lightweight, low-power, low weight-on-bit, Icebreaker-3 drill with its design improvements was capable of penetrating hard rock and ice-consolidated material in lab, chamber and field tests. The stiffer drillstring, stronger motors and reverse rotation capability improved drilling operations notably over prior designs and with sample transfer interoperability demonstrates a technology readiness level suitable for consideration as a flight sample acquisition instrument.   References: [1] Glass, B., et al, (2014) LPSC XLV. [2] Glass, B. et al, (2014) J. Field Robotics. [3] Glass, B. et al. (2008) Astrobiology. [4] Zacny, K., et al, (2013) LPSC XLIV. [5] Dave, A., et al, (2013) Astrobiology. [6] Bergman, D. et al, (2015) LPSC XLVI.   Fig. 2. IB-3 tested (a) with sample transfer arm on Phoenix deck mockup (June 2014) and (b) at the Haughton Crater Mars-analog site (August 2014).   Fig. 3. Nominal 15 cm (150mm) drilling "bite" by IB3 during Haughton Crater tests. 
