Reaching 1m Deep on Mars: The Icebreaker Drill -...

Reaching 1 m Deep on Mars:The Icebreaker Drill

K. Zacny,1 G. Paulsen,1 C.P. McKay,2 B. Glass,2 A. Dave,2 A.F. Davila,2 M. Marinova,3

B. Mellerowicz,1 J. Heldmann,2 C. Stoker,2 N. Cabrol,4 M. Hedlund,1 and J. Craft1

Abstract

The future exploration of Mars will require access to the subsurface, along with acquisition of samples forscientific analysis and ground-truthing of water ice and mineral reserves for in situ resource utilization. TheIcebreaker drill is an integral part of the Icebreaker mission concept to search for life in ice-rich regions on Mars.Since the mission targets Mars Special Regions as defined by the Committee on Space Research (COSPAR), thedrill has to meet the appropriate cleanliness standards as requested by NASA’s Planetary Protection Office. Inaddition, the Icebreaker mission carries life-detection instruments; and in turn, the drill and sample deliverysystem have to meet stringent contamination requirements to prevent false positives.

This paper reports on the development and testing of the Icebreaker drill, a 1 m class rotary-percussive drill andtriple redundant sample delivery system. The drill acquires subsurface samples in short, approximately 10 cmbites, which makes the sampling system robust and prevents thawing and phase changes in the target materials.Autonomous drilling, sample acquisition, and sample transfer have been successfully demonstrated in Mars an-alog environments in the Arctic and the Antarctic Dry Valleys, as well as in a Mars environmental chamber. In allenvironments, the drill has been shown to perform at the ‘‘1-1-100-100’’ level; that is, it drilled to 1 m depth inapproximately 1 hour with less than 100 N weight on bit and approximately 100 W of power. The drilled sub-strate varied and included pure ice, ice-rich regolith with and without rocks and with and without 2% per-chlorate, and whole rocks. The drill is currently at a Technology Readiness Level (TRL) of 5. The next-generationIcebreaker drill weighs 10 kg, which is representative of the flightlike model at TRL 5/6. Key Words: Drilling—Sampling—Mars—Mars drilling—Subsurface exploration—Ice—Search for life. Astrobiology 13, 1166–1198.

Table of contents

Abstract 11661. Introduction 11672. The Martian Near Subsurface: Experience from Past Missions 11683. Considerations When Designing a Drill for Mars Surface Operations 1171

3.1. Science drivers 11713.2. Environmental drivers 11713.3. Planetary protection drivers 11733.4. Technology drivers 1173

4. The Icebreaker Mars Drill 11744.1. Drilling depth 11744.2. Selecting the best drilling method 11744.3. Sample type 11754.4. Components of the Icebreaker drill 1176

4.4.1. Deployment boom 11764.4.2. Z-stage 11764.4.3. Drill head 1177

1Honeybee Robotics, Pasadena, California.2NASA Ames Research Center, Moffett Field, California.3Space Exploration Technologies Corporation, Hawthorne, California.4SETI Institute–Carl Sagan Center Mountain View, California, and NASA Ames Space Science Division, Moffett Field, California.

ASTROBIOLOGYVolume 13, Number 12, 2013ª Mary Ann Liebert, Inc.DOI: 10.1089/ast.2013.1038

1166

4.4.4. Drill auger 11774.4.5. Drill bit 11784.4.6. Brushing station 1179

4.5. Drilling software 11794.5.1. Mission-critical events 11804.5.2. Mission-noncritical events 1183

4.6. Drill as a science instrument 11835. Sample Acquisition 11836. Sample Delivery 1184

6.1. Five-DOF sampling arm and a scoop 11846.2. Pneumatic sample transfer 11856.3. Three-DOF-DOF arm and drill 1186

7. Icebreaker Drill Tests 11877.1. Test environment 11887.2. Testing in a Mars environmental chamber 11887.3. Testing in Mars analog sites of Antarctica 1191

8. Next-Generation Icebreaker: The Icebreaker2 Drill 11949. Conclusions 1195Acknowledgments 1195Abbreviations 1195References 1195

1. Introduction

After the Moon, Mars is the most extensively studiedextraterrestrial body. Thus far, the examination of the

near surface has been mainly achieved by the scooping ofsurface regolith. This was done on the Viking 1 and 2, theMars Phoenix (Smith et al., 2009), and the Mars Science La-boratory missions. Penetrating rocks was also achieved to adepth of a few millimeters by the Rock Abrasion Tools on theMars Exploration Rovers (Gorevan et al., 2003), as well as to adepth of *6 cm by the Powder Acquisition Drill System onthe Curiosity rover (Okon, 2010). Martian ice was also pene-trated by a small drill bit, called the rasp, which was an in-tegral part of the Icy Soil Acquisition Device (Chu et al., 2008).The 2016 InSight mission will have a percussive mole that willattempt to penetrate up to 5 m into martian regolith and pull athermal tether behind it. The 2018 ExoMars rover will have a2 m drill with a goal of acquiring samples from target depthsand delivering them to an onboard crusher (European SpaceAgency, 2012). Although the details about the Mars2020 roverpayload are still unclear, there is a high probability that themission will include a coring drill to acquire core samples upto 10 cm deep and cache them for sample return.

None of these future missions, however, target thenorthern latitudes as landing zones. The latitudes above43�N are of particular interest to astrobiology since it wasdetermined that water ice might exist in the near-surfaceregolith (Byrne et al., 2009). The ice-rich ground is a prom-ising target because the ice may have been habitable in thepast and also might have protected organic remains fromnear-surface oxidization and cosmic radiation. Amid the coldtemperatures and its radiation-absorbing properties, the top1 m of ice-rich regolith on Mars could be the best location tosearch for organic biosignatures that would represent evi-dence of life (McKay et al., 2013).

This paper describes the development of a sample acquisitionsystem for the Icebreaker mission, targeting the ice-rich per-mafrost of the northern latitudes of Mars (McKay et al., 2013).

The science goals for the mission are as follows:

(1) Search for specific biomolecules that would be con-clusive evidence of life;

(2) Perform a general search for organic molecules in theground ice;

(3) Determine the processes of ground ice formation andthe role of liquid water;

(4) Understand the mechanical properties of the martianice-cemented soil;

(5) Assess the recent habitability of the environment withrespect to required elements to support life, energysources, and possible toxic elements; and

(6) Compare the elemental composition of the northernplains with midlatitude sites (McKay et al., 2013).

The Icebreaker mission could be based on the 2007 MarsPhoenix and 2016 InSight landed platforms (Fig. 1). The Ice-breaker sample acquisition system includes a 1 m class drill,which will be deployed by way of a 3-degree-of-freedom(DOF) arm. An integral part of the drill system is a triplyredundant sample delivery subsystem that includes a 5-DOFrobotic arm with a delivery scoop (Dave et al., 2013), end-to-end pneumatic sample transfer, and the drill deployment armthat would position the tip of the drill with its captured sampleabove an instrument sample inlet cup (Zacny et al., 2012a).

Since the drill mission targets Mars Special Regions asdefined by NASA’s Planetary Protection Office (Conley,2011), the drill has to meet required cleanliness standards.The Icebreaker mission carries life-detection instruments, sothe drill and sample delivery system must also meet strin-gent contamination requirements to prevent false positives.

In addition, the drill’s telemetry such as penetration rate,power, and temperature could be used to infer subsurfacestrength, ice content, and downhole temperature. The latterdata combined from different depths could be used to map athermal gradient and heat flow, if subsurface thermal con-ductivity is known or can be estimated. As such, the drill canbe viewed as an integral part of the sample delivery chain

ICEBREAKER: THE MARS DRILL 1167

and is more akin to an instrument than mission-enablinghardware (e.g., robotic arm).

2. The Martian Near Subsurface:Experience from Past Missions

Thus far, there have been only six missions that have de-ployed excavation tools on Mars. These are Viking 1 and 2,Mars Exploration Rovers (MER) Spirit and Opportunity,Mars Phoenix, and the Mars Science Laboratory (MSL).Table 1 lists these six missions along with the type of exca-vation tool deployed.

The first excavators deployed on Mars were the scoops onthe Viking 1 and 2 landers. The purpose of the scoop, calledthe Viking Surface Sampler Assembly (SSA) and shown inFig. 2, was to acquire, process, and distribute samples tovarious instruments (Seger and Gillespie, 1974). The samplerconsisted of a 3 m long rolled-up tubular boom with a col-lector head at its end. The extendable/retractable boomcombined with the integrated azimuth/elevation gimbalallowed the collector head to be placed to any locationwithin the articulation limits of the boom. The collector head,with its solenoid-operated lid, backhoe, and 180� rotationcapability, was designed to acquire samples from a variety of

FIG. 1. The Icebreaker missionwould search for organic bio-markers at the northern latitudesof Mars by drilling to at least 1 mdepth below the surface and trans-ferring sample to life-detectioninstruments (Glass et al., 2011).Color images available online atwww.liebertonline.com/ast

Table 1. Summary of Excavation Tools Deployed on Mars

Date Mission Tool Tool Description Depth

1976 Viking 1 and 2 Scoop: the SurfaceSamplerAssembly (SSA)

The SSA was designed to acquiretop martian regolith, sieve it, anddeliver it to lander-mountedinstruments such as the GCMS.

Tens of cm

2003–now

Mars ExplorationRoversOpportunityand Spirit

Grinder: RockAbrasionTool (RAT)

The RAT was designed to grindthrough the weathered andoxidized crust on martian rocks.The RAT leaves an abraded area45 mm in diameter and a fewmillimeters deep.

mm

2008 Phoenix Scoop and a drillbit: Icy SoilAcquisitionDevice (ISAD)

The ISAD acquired top regolithusing a scoop and icy soil usinga small drill bit called the rasp,mounted underneath the scoop.

Scoop: tensof cm

Drill: mm

2012–now

Mars ScienceLaboratory

Drill: PowderAcquisitionDrill System(PADS)

The PADS is a rotary-percussivedrill. It can penetrate up to 6.5 cmand acquire rock cuttings between1.5 and 5 cm depth range.

Up to 6.5 cm

2012–now

Mars ScienceLaboratory

Scoop: Collectionand Handling forIn situ MartianRock Analysis (CHIMRA)

CHIMRA has a scoop for acquiringsurface regolith.

mm

1168 ZACNY ET AL.

potential surface materials and to deliver raw samples or2000 lm sieved samples to the various experiments. Thebackhoe permitted surface trenching operations, collection ofmagnetic surface materials, and the brushing of surfacematerials overlaying bedrock.

Additional soil processing mechanisms were integratedwith the gas chromatograph–mass spectrometer (GCMS) andthe biology instruments. For the GCMS, the sample was sievedthrough a 2000-micron sieve, crushed to less than 600 microns,sieved through a 300-micron sieve and finally metered 1 cm3 anddelivered to the GCMS instrument. For the biology instrument,the sample was sieved through a 1500-micron sieve, metered to7 cm3, and finally delivered to the biology instrument.

Although the Rock Abrasion Tool (RAT) deployed on MERSpirit and Opportunity is not technically an excavator but agrinder, it is nevertheless a tool that has successfully pene-trated dozens of martian rocks (Gorevan et al., 2003). Theprimary goal of the 680 g 10 W RAT shown in Fig. 3 was togrind away a 45 mm diameter and a few millimeter–deep holeand expose the virgin rock to arm-mounted instruments suchas the Alpha Particle X-Ray Spectrometer (APXS), MossbauerSpectrometer, and Microscopic Imager (MI). Removing thefirst few millimeters was imperative to the success of themissions, since martian rocks are covered in weathered andoxidized crust and these crusts are not representative of theunderlying rock. As such, the RAT is akin to a geologist’shammer that breaks the rocks open to reveal fresh surfaces.

To date, the RAT on MER Spirit performed 92 brushingsand 15 grinds, while the RAT on Opportunity has performed31 brushings and 41 grinds. It should be noted that rocks atGusev Crater, where MER Spirit landed, are much harderthan rocks in Meridiani Planum, where MER Opportunitylanded. This explains the greater number of grinds at Mer-idiani Planum. Since the MER Spirit mission ended, that RATcannot be used any longer. However, the RAT on MER

Opportunity is still being used as a brushing and a grindingtool.

The 2007 Phoenix mission was the first mission to land atthe northern polar regions of Mars (Smith et al., 2009). Thelander carried two instruments that required solid samples: theThermal Evolved Gas Analyzer (TEGA) and the Microscopy,Electrochemistry, and Conductivity Analyzer (MECA). To

FIG. 2. Components of the Viking Surface Sampler Assembly (SSA). Credit: NASA.

FIG. 3. Mars Exploration Rovers’ Rock Abrasion Tool(RAT) and a 45 mm diameter and few millimeter–deepRATed hole. Credit: NASA. Color images available online atwww.liebertonline.com/ast


enable sample acquisition and delivery, the mission included ascoop mounted at the end of a 4-DOF robotic arm (Fig. 4). Thescoop, called the Icy Soil Acquisition Device (ISAD) also in-cluded a 6.35 mm diameter tungsten carbide drill bit called therasp, to enable penetration of icy regolith (Chu et al., 2008).During the mission operations, the scoop was initially used toremove the layer of loose regolith covering ice-bearing mate-rial, and the flat-edged blade underneath the scoop was thenused to scrape away icy regolith and form a more uniformsurface for the rasp tool. Once the icy surface was prepared,the robotic arm preloaded the ISAD against the surface with40 N in such a way that the rasp tool would press firmlyagainst the icy regolith. The rasp cutting tool was engaged forapproximately 30–60 s, and during this time icy-regolithshavings ballistically fell to the back of the scoop. The majorproblem the sampling system experienced during the missionoccurred when delivering icy regolith. The regolith was foundto be sticky and not only had a hard time flowing out of thescoop but also did not easily flow through the screens abovethe instrument inlet ports. The rasp subsystem included a setof features for vibrating the scoop; vibration helped but did noteliminate the problem.

The regolith at the Phoenix site was found to contain 0.4–0.6% wt perchlorate (ClO4) (Hecht et al., 2009). It has beenpostulated that, if temperatures are high enough, it is pos-sible for the ice and the perchlorate-rich regolith to form awet super-eutectic solution (Chevrier et al., 2009). This couldexplain the cohesive nature of the regolith. This finding hastremendous implications for future missions requiring icy-sample acquisition and delivery subsystems. In particular, ifa sample’s temperature is high enough, relying on gravity tomove the sample will make delivery either difficult or vir-tually impossible to accomplish. In addition, cross contami-nation between samples will be much worse for stickysamples.

There are three factors that could make martian regolithsticky: the presence of ice, perchlorate salts, and a largetemperature differential between the sample acquisitionsystem and the sample itself. It is impossible to eliminate thefirst two factors; however, it might be possible to control thethird. For example, the sample acquisition tool could beplaced within the shadow of the spacecraft, or the entiresampling operation could be performed at night, which

might be up to 100�C cooler than the day. In addition, thesampler itself would need a built-in temperature sensor tocontinuously monitor the temperature during sample ac-quisition and either slow down or stop the process to preventpossible warming of the sample.

A recent study of new impact craters at the martianmidlatitudes revealed the presence of water ice much closerto the surface than initially suspected. In particular, water icecould be present as close as 34 cm below the surface at alatitude of 43�N and within the top 10 cm at a latitude of45�N (Byrne et al., 2009). This of course makes future mis-sions to such low latitudes more enticing because of near-surface ice but also puts more pressure on the robustness ofthe sample acquisition system to deal with potentially ice-rich material.

The first mission to carry a drill to Mars is the ongoing2011 MSL rover mission. The MSL mission consists of a 1-tonrover called Curiosity, the goal of which is to traverse mar-tian landscapes and characterize past habitable environ-ments recorded in sediments and rocks. In addition tocontact instruments, it houses two of the most advancedanalytical instruments ever built: Sample Analysis at Mars(SAM) and Chemistry and Mineralogy (CheMin). Both in-struments contain 74 sample cups or cells and require sam-ples in the form of rock powder or regolith. The powder forCheMin has to be smaller than 150 microns, while thepowder for SAM needs to be smaller than 1 mm.

To address critical sample acquisition requirements, theCuriosity rover contains an end-of-arm mounted systemcalled the Sample Acquisition, Sample Processing andHandling (SA/SPaH) system (Jandura, 2010). The SA/SPaH’s two main elements responsible for sample acquisi-tion include the drill, which is called the Powder AcquisitionDrill System (PADS), and a scoop. The PADS shown in Fig. 5is a rotary-percussive drill, with percussion provided by avoice coil mechanism (Okon, 2010). The PADS drill bit canspin at up to 150 rpm, while the percussive system can de-liver up to 0.8 J per blow at up to 1800 blows per minute(BPM). To acquire a sample, the drill is first placed against arock surface by a robotic arm with an axial preload of up to300 N. This initial preload virtually eliminates the require-ment for any subsequent motion of the robotic arm. The drillbit is protected within a sleeve that has a dual purpose of

FIG. 4. Components of the MarsPhoenix 2007 Icy Soil Acquisition De-vice (ISAD).

1170 ZACNY ET AL.

guiding the drilled cuttings (also referred to as rock powder)up the auger and also serving as an anchor for the rover incase of wheel slip on steeper terrains.

During a drilling operation, the drill penetrates the rock,and at the same time, the drilled cuttings travel up an augerand into a bit chamber that is connected to a powder pro-cessing unit, Collection and Handling for In situ MartianRock Analysis (CHIMRA). The auger ends with a 1.6 cmdiameter tungsten carbide bit. The drill can penetrate up to6.5 cm depth and can capture cuttings from between theupper 2 cm and *5 cm depth. The top 2 cm is discarded ontothe rock surface (this is desirable since the top layer of rockhas been weathered and oxidized), while the bottom-mostpowder is always left in the hole by the nature of the sam-pling system that requires powder to move up some distancebefore it enters the auger tube section. Figure 5 shows thefirst two holes drilled into a rock called John Klein in theYellowknife Bay area of Mars’ Gale Crater. The upper holewas drilled on Sol 180 (Feb. 6, 2013) to a depth of 2 cmwithout collecting any rock powder as a test. The lower holewas drilled to 6.4 cm on Sol 182 (Feb. 8, 2013), and somepowder was successfully delivered to the instruments.

Samples of regolith and other unconsolidated materialsfrom depths of up to 3.5 cm are acquired with a clamshellscoop that is part of CHIMRA (Sunshine, 2010). The scoopwas also designed to acquire unconsolidated samples fromrover wheel–dug trenches. This would allow access to ma-terial as deep as 20 cm below the surface. In the nominaloperation, the scoop can collect somewhere between 1 and 30cc of material.

3. Considerations When Designing a Drill for MarsSurface Operations

There are several aspects that need to be considered whendesigning a drill with a goal of acquiring samples for anal-ysis by very sensitive instruments. These aspects could bedivided into the following categories: the science driven, theenvironment driven, the planetary protection driven, andthe technology driven (Zacny et al., 2008; Zacny and Bar-Cohen, 2010).

3.1. Science drivers

The science drivers include the volume of the sample (sincesample mass is very difficult to measure accurately), the par-ticle size, the maximum allowable temperature during sample

acquisition, whether the sample can be exposed to the atmo-sphere or particular gasses/moisture, and the target depthfrom which the sample needs to be acquired. Knowing theexact sample volume is critical to derive concentration values,since underfilling the instrument intake could lead to under-estimated concentrations. Particle sizes are important for in-struments that are sensitive to the largest particles. Forexample, the X-ray diffraction instrument (CheMin) onboard the Curiosity rover requires particles smaller than150 lm. Other instruments might require particles to passthrough internal filters or that would not block internalconduits (Parro et al., 2008). The thermal environment of thesample is extremely important. Some minerals changecrystal structure if certain temperatures are exceeded. Inaddition, in ice-rich permafrost increasing temperature maylocally melt the ice.

3.2. Environmental drivers

Environmental drivers include gravity lower than that ofEarth, very low temperatures, high diurnal thermal fluctua-tions, low atmospheric pressure, dust storms, and largedistance from the Sun.

Low gravity is particularly challenging since drilling reliesheavily on the weight on bit (WOB) or the vertical forceapplied on the drill bit. In general, the larger the force is, thefaster the penetration rate. The Icebreaker mission (McKayet al., 2013) is based on the Phoenix platform; hence, themaximum possible preload will be the weight of the landerat approximately 350 kg scaled down by the martian gravityof 3.7 m/s2. This equates to 1300 N. However, it is very un-likely that the drill will be deployed along the center ofgravity of the lander. In the most likely scenario, the drill willbe initially in a horizontal position strapped to the deck ofthe lander. It can be deployed upon touch down by a 3-DOFarm and touch the martian surface some distance from thedeck (because rocket plumes might contaminate the regolithin the immediate vicinity of the lander, it would be morefavorable if the drill was deployed as far from the lander aspossible). Assuming a lever arm ratio of 1:2, the maximumload will then be half of the original or 650 N. In addition, inpractice a 3-DOF robotic arm will initially preload the drill Z-structure with a certain force; then the drill will start pene-trating the subsurface along the Z-axis. The maximum WOBavailable for drilling must therefore be much lower than theforce the drill Z-structure was preloaded with against theground. To be safe, the maximum WOB must be at most 50%

FIG. 5. The Powder Acquisition DrillSystem (PADS) on the Curiosity roverand the first holes drilled (1.6 cm diam-eter) on Mars. Photo courtesy of NASA.Color images available online at www.liebertonline.com/ast


of the preloaded force applied to the Z-structure by the ro-botic arm. This further drops the maximum WOB from 650to 325 N. If a factor of safety of 2 is added, this further re-duces the WOB from 325 to *160 N. The factor of safety isrequired since a lander could, for example, touch down on aslope with one of its legs resting on a large boulder, causingthe lander to tilt down slope. If a drill were deployed up theslope, the maximum WOB would then be much lower than ifa lander was on perfectly flat ground. In summary, a landerthat weighs 3500 N on Earth will be able to provide ap-proximately 160 N WOB for the drilling system on Mars.

Low temperature and high thermal fluctuations affect themechanical and electrical design of the spacecraft as well asits operation. In the case of mechanical design, it is importantto avoid any instances where two materials with differentcoefficients of thermal expansion might contact each other.This would lead to high stresses and possible failures. Elec-tronics need to be kept within a warm electronics box withintight temperature limits (for example - 40�C to 40�C). Anyother electronics outside the warm electronics box must havethe capacity to withstand low temperatures or be integratedwith heaters to keep them warm at night. Keeping thespacecraft warm could consume much, and in some in-stances almost all, of its stored electrical energy. Hence, otherscience instruments and tools such as drills need to be asefficient as possible when in use. Ice-rich regolith and ice-saturated rocks (i.e., porous rocks filled with water and thenfrozen) could be up to three times stronger at - 100�C thanat - 10�C (Mellor, 1971; Zacny and Cooper, 2006). Even dryrocks get stronger at lower temperature, but the increase instrength is not as significant as in ice-rich materials (Heinsand Friz, 1967; Kumar, 1968; Zacny and Cooper, 2007a).Higher formation strength will directly lead to higher dril-ling energy and WOB (Zacny et al., 2007).

Mars’ atmospheric pressure brackets the triple point ofwater. In the northern lowlands of Mars, atmospheric pres-sure is above the triple point; hence, liquid water would bevery unstable. When drilling aggressively (or inefficiently)into ice-rich regolith, the heat generated during the drillingprocess could be enough to sublime all the water ice un-

derneath the drill bit cutters (Fig. 6). The water vapor gen-erated in this process acts as a very effective drilling fluid,blowing the cuttings out of the hole in a similar way air drillswork on Earth (Zacny et al., 2004). The major difference isthat on Earth the air is provided from a compressor, while onMars it is a result of ice sublimation. The benefit of suchaggressive drilling is lower drilling power and higher pen-etration rate. In addition, the formation temperature remainsat or near zero, since most of the heat is used up by the latentheat of sublimation. However, if the mission requires sam-pling water ice, such a sublimation effect has to be limitedthrough efficient drilling and carefully planned drillingprotocols that allow for ample cooling time in case the drillbit temperature gets closer to freezing. It should also benoted that heat generated during drilling could increase thetemperature of ice-cemented ground underneath the drill bitand subsequently reduce its strength. This will result in ahigher penetration rate and increased drilling efficiency.

Low atmospheric pressure also reduces sliding frictionbetween the drill bit and the rock as shown in Fig. 7 (Zacny

FIG. 6. Drilling data under Mars andterrestrial atmospheric pressure in icyformation. Sublimed ice blows the cut-tings out of the hole, which improvesdrilling efficiency (Zacny et al., 2004).Color images available online at www.liebertonline.com/ast

FIG. 7. Coefficient of friction between a drill bit and a rockas a function of WOB and atmospheric pressure (Zacny andCooper, 2007b).

1172 ZACNY ET AL.

and Cooper, 2007b). This is an important consideration fordrilling since less heat will be generated during the drillingprocess and drilling energy will be lower. When interpretingdrilling data, this will need to be accounted for as well, sincelower drilling energy could incorrectly indicate lower rockstrength. Sliding friction between two of the same or differ-ent metals needs to be investigated on a case-by-case basis.Coefficient of friction changes as a function of material typeand atmospheric gas (Buckley, 1971; Niebuhr, 2007). Niebuhr(2007), for example, found that the coefficient of friction be-tween titanium 6Al4V and hardened steel is 0.41 in air, 0.39in nitrogen, and 0.33 in carbon dioxide atmosphere (all at 10–15 torr pressure). For 7075T-6 aluminum on steel, coefficientsof friction are 0.45 (air), 0.41 (nitrogen), and 0.55 (carbondioxide). Hence, nitrogen and air behave in a statisticallyidentical manner for both metals, while CO2 increases fric-tion for aluminum but decreases friction for titanium. Coef-ficients of friction of UHMWPE and PTFE on steel are *0.2and 0.1, respectively, and are not significantly influenced byatmospheric composition.

Eolian processes are the most important transport modefor micron-sized particles (Pike et al., 2011). This fine dustposes many problems for mechanical systems on Mars. Inparticular, any mechanical system that has moving parts(e.g., a Z-stage) needs to be protected from abrasive dust orrobust enough to work when coated with dust. The dust mayalso have a high electrostatic charge due to tribochargingthrough contact with other particles or materials, or photo-ionization by the intense UV radiation. The settling of thisdust can significantly reduce the efficiency of solar panels.

Sample transfer systems also need to be protected from dustto avoid dilution and contamination. Protection of a sampletransfer system with wind guards is one way to addressthe problem of dilution or contamination, or even a loss of thesample if wind speed suddenly increases. The Viking landersand the MSL Curiosity rover implemented wind guardsaround the sample drop-off points to the instruments.

The large distance from the Sun affects not only the sur-face temperature but also energy density. The solar irradi-ance at Mars is 586 W/m2, which is approximately 40% thaton Earth at 1353 W/m2. Therefore, solar arrays have to bemuch larger and ideally include a tracking device to pointthe arrays toward the Sun as much as possible. The lowirradiance means that the spacecraft will have limited energyfor housekeeping and science operations. Hence, any sam-pling system needs to be very efficient in reducing the energyburden on the spacecraft. Mars Phoenix had an average en-ergy balance of 2 kWhr per sol; however, 50% of that wasrequired for housekeeping.

In addition, given light-speed time delays in the range ofseveral minutes, Earth-based teleoperation is not feasible fordrilling operations on Mars. Drilling must be autonomous.Unlike terrestrial holes whereby subsurface is either knownfrom prior drilled holes or could be evaluated through geo-physical surveys, the subsurface of the drill site on Mars willbe largely unknown.

3.3. Planetary protection drivers

Ice-rich permafrost on Mars is considered a Special Re-gion; therefore, the mission qualifies as category IVc as de-fined by NASA’s Planetary Protection Office (Conley, 2011).

The Icebreaker lander might require complete sterilization byusing, for example, dry heat microbial reduction (DHMR) ina similar manner as it was used for the Viking landers. Fromthe standpoint of drill design, sterilizing an entire spacecraftwould mean that all parts of the drill, including the elec-tronics, would need to be designed to withstand DHMR, thatis, baking at over 110�C for more than 24 hours (the Vikinglanders were baked at 111.7�C for 30 hours). This wouldsimplify surface operations, since the drill could come incontact with any part of the spacecraft without the worry ofpossible contamination from ‘‘unsterilized’’ surfaces.

The alternative is sterilizing only the hardware that comesin contact with the subsurface, as was done during the MarsPhoenix mission. This would drastically simplify spacecraftdesign and cost since it would allow the use of materials,electronics, and heat-sensitive equipment that would other-wise be damaged by high-temperature treatment. However,the drill auger would have to be protected in some kind ofbiobarrier after sterilizing. In addition, any other hardwarethat comes in contact with a sample (e.g., a scoop) would alsoneed to be sterilized. Since the Icebreaker mission wouldcarry life-detection instruments, a high degree of cleanlinesswould be required to prevent false positives. The sampletransfer sequences as well as proximity operations wouldhave to be thoroughly tested out to avoid potential touchingof the sample transfer hardware with unclean surfaces (e.g.,instrument outer housing).

3.4. Technology drivers

Drilling power is not as much of an issue as drilling en-ergy (though it still needs to be considered). Energy is themain factor because batteries store a finite quantity of elec-trical energy. Once that energy is used up, the drilling taskwould have to be delayed until the batteries are chargedagain (by a primary power source such as solar or nuclear).To better illustrate that point, consider two cases: (1) a 200 Wrotary-percussive drill that takes 1 hour to drill a 1 m holeand (2) a 100 W rotary drill that takes 10 hours to drill to thesame depth. The first, high-power drill requires 200 Whr,while the second, low-power drill requires 1000 Whr of en-ergy, 5 times as much energy.

A comparison between a high-power Hilti TE-7A rotary-percussive drill and a low-power custom-built CRUX rotary-percussive Lunar/Mars drill in saddleback basalt rock isshown in Table 2. Although the Hilti drill required morepower (720 W vs. 180 W for the CRUX), it drilled 3 timesmore efficiently and 12 times faster. There are other tech-nology drivers associated with drilling; these are describedin the context of drill development in the next sections.

Table 2. Comparison between High-Power

and Low-Power Rotary-Percussive Drill Systems

CRUXdrill

HiltiTE-7A

Ratio: Hilti/CRUX

Total power (W) 180 720 4WOB (N) 100 100–130 1Rate of penetration (cm/min) 0.24 3 12.5Energy to reach 1 m (Whr) 1170 360 0.3


4. The Icebreaker Mars Drill

The Icebreaker drill represents a state-of-the-art drillingsystem specifically designed to meet the science requirementof retrieving aseptic samples from ice-rich regolith. To meetthis requirement, several trades were performed, evaluatingvarious drilling and sample delivery approaches. The finalselection of the drilling and sampling approach was drivenby achieving the goal while minimizing the mission risk, andeliminating forward contamination (primary) followed byreducing system mass and increasing drilling efficiency(secondary). The following sections outline various tradesperformed in deciding optimum sampling system design.

4.1. Drilling depth

Drilling regimes could be divided into four distinct areas:surface, 1 m, 10 m, and deeper than 10 m (Bar-Cohen andZacny, 2009). This division is based on the drill complexity,size, and mass rather than science goals.

Surface regime refers to a drill that is deployed from arobotic arm and penetrates a surface rock or ground to adepth from several centimeters to tens of centimeters. Ex-amples of such drills include the MSL drill, the Luna 16 and 20drills, and the Venera 13 and 14 drills (Bar-Cohen and Zacny,2009). However, the goal of the Icebreaker mission is to searchfor past or present life, and this requires reaching depths be-low oxidized and irradiated regolith. It is believed that theminimum depth is 1 m, though greater depths will no doubtbe more favorable (McKay et al., 2013). Therefore, a surfacedrill would not be applicable for the Icebreaker mission.

A 1 m class drill is characterized by using a single drillstring to penetrate into the subsurface. In this regard, it issimilar to a surface drill; however, the depth of penetration isalmost an order of magnitude greater. As such, the drill ac-tuators have to be more powerful so that the drill can beretracted in the event it is getting stuck, and the drillingalgorithm would need to be designed such that it couldmanage a larger range of adverse drilling conditions. Ex-amples of such drills include the Luna 24 drill and the SD2drill (Drill, Sample and Distribution) on the Rosetta landerPhilae (Bar-Cohen and Zacny, 2009).

A 10 m class drill is similar to a 1 m class system exceptthat it includes a system for additional drill strings (e.g., acarousel) to enable the reaching of greater depths (Zacnyet al., 2013a). The greater depth capability comes at a price ofadditional mass for drill strings and a drill string handlingsystem. Drill strings would also require dust-tolerant me-chanical and potentially electrical connections. Examples ofsuch systems include the ExoMars drill, SCAD, and MARTE(Bar-Cohen and Zacny, 2009). A 10 m depth seems to be themost a drill can achieve when using an auger-based cuttingsremoval system. The limiting factor is the inability to effi-ciently move cuttings out of the hole. Drilling to greaterdepths will produce excessively high auger torque due toparasitic losses (auger rubbing on borehole wall) and frictionproduced by the movement of cuttings up the flutes. Oneapproach to the removal of cuttings is to use compressed gasin the same manner as in terrestrial air drilling (Zacny andCooper, 2007c).

For deep drilling in consolidated formations (e.g., ice orrock) where borehole collapse is not an issue, a wire linesystem can be used. Here, a drill in the shape of a tube is

suspended by a tether and is periodically lifted out of thehole to empty the catch basket or drop off cores (if present).Examples of wire line drills include the Mars Deep Drill(MAD) and the Autogopher (Bar-Cohen and Zacny, 2009;Zacny et al., 2013b). If the formation is unconsolidated andthe borehole is likely to collapse, a drill system would need acasing and a coil tubing approach.

Deep drilling is not possible in the near future unlesssubstantial funding is set aside to start demonstrating such atechnology and developing entry, descent, and landing(EDL) systems capable of landing > 1 ton on the surface ofMars. A 10 m class drill is possible but carries substantial riskto the mission due to the requirement for connecting anddisconnecting drill strings multiple times in the presence ofdust. However, a single-string, 1 m class drill is feasible. Asingle string could be as long as the volume of the spacecraftallows it to be. For the Phoenix-sized lander, a drill couldpotentially reach 1.2 m. The SpaceX Red Dragon mission,which emphasizes placing the drill inside its Dragon capsule,could accommodate at least a 2 m long drill string (Stokeret al., 2012).

Because of the relative simplicity of the system and theability for a single drill string to reach the required depth thatmight preserve organic material, it was decided that theIcebreaker drill should be a single-string system.

4.2. Selecting the best drilling method

There are at least three approaches to penetrating theground: rotary, rotary-sonic, and rotary-percussive. Thedifference between the sonic and percussive systems is that,in sonic, the drill string vibrates, while in percussive systemsa stress wave travels at the speed of sound from the top,where it is generated by an impact hammer to the drill bit.

Vibratory drills use piezo systems (Bar-Cohen et al., 2010)or a set of unbalanced wheels spun by a motor (Paulsen et al.,2012). In the latter approach, two wheels need to be used inorder to cancel out side vibrations and loads. The principlebehind vibratory drills is that they aim to put the drill stringin resonance. Resonant vibration increases penetration forceby fluidizing the regolith ahead of a drill bit and around adrill pipe. The latter has a tendency to reduce sleeve friction(friction between the drill pipe and regolith). The resonantfrequency of a drill string is usually found by doing a fre-quency sweep and observing penetration rate. At a resonantfrequency, the rate of penetration (ROP) drastically increases;hence, a driller continuously adjusts the frequency whilemonitoring the ROP. Such drills are used every day forregolith sampling (see for example the Boart Longyear LS600Sonic Rig) and could also drill through weaker rocks, thoughat much lower penetration rates.

Percussive drills are ideally suited for hard rocks. Here,the drill string is struck by a hammer accelerated throughvarious means such as a voice coil (Okon, 2010), pneumatic(as used in Hilti drills), magnetic (developed by Flexidrill),ultrasonic (Bar-Cohen et al., 2010), or a spring-cam (Paulsenet al., 2011). The selection of the most suitable percussiveapproach depends on the environment (in vacuum, airhammer will not work), whether variable impact energy orimpact frequency is required, the complexity and robustnessof the mechanism, and the greatest ratio of impact energy permass of the system.

1174 ZACNY ET AL.

To determine the best drilling approach for Icebreaker,initially three drills with different drilling approaches weretraded against each other. These include the DAME drill,which has a rotary drilling approach (Paulsen et al., 2006;Glass et al., 2008), the SONIC drill with a rotary-sonic ap-proach (Paulsen et al., 2012), and the CRUX drill with arotary-percussive approach (Zacny et al., 2010a; Paulsen et al.,2012). Figure 8 shows a summary of three different drillingapproaches in 45 MPa Indiana limestone. Since the rotarysystem does not use additional sonic or percussive actuators,the required drilling power was lowest. On the other hand,SONIC requires the most power. In terms of energy, the mostefficient system was found to be rotary, followed by rotary-percussive and rotary-sonic. Rotary-percussive required thelowest WOB and achieved the highest ROP.

Additional tests were also performed in other materialssuch as soils, ice, and ice-cemented regolith (Zacny et al.,2013a). It was found that rotary systems work well in softand non-abrasive formations, while percussive systemsdominate in hard formations. Rotary-sonic systems did notperform as well as percussive and were only marginallybetter than rotary. In addition, WOB required to achieve asignificant penetration rate in any formation was lowest forthe rotary-percussive drills.

Based on these findings, it was decided to select rotaryand rotary-percussive approaches for the Icebreaker drill. Toenable an option of having the two different approaches,two independent actuators were employed for a rotary augerand the percussive mechanism. It is possible to use a sin-gle actuator with some kind of a clutch to disengage per-cussion, but that would mean the indexing (i.e., impact blowsper every bit revolution) would have to be set at the begin-ning and could never be changed. In addition, a percussivesystem could not be used by itself (without rotation), andthat flexibility might be useful during a sample drop-offoperation.

The flexibility of separating rotary from percussive in-creases drilling efficiency. In soft rocks, only a lower-powerrotary approach could be used, while in harder rocks a morepowerful (and potentially more energy intensive) rotary-percussive system would be engaged. However, flexibility ofusing both approaches also puts an additional requirementon the drill bit cutter material and its shape. For percussive

systems, cutters are made of softer and hence toughertungsten carbide to prevent potential fractures. Cutters alsohave a rounded cutting edge to reduce stress concentrations.For rotary systems, where drilling mechanisms include theshearing of rock, cutters need to have a sharp edge. To retainthe cutter’s sharp edge for as long as possible, much harder(and in turn less tough or more brittle) tungsten carbidematerial is used. Alternatively, a polycrystalline diamondmaterial could be used, though polycrystalline diamondcompact cutters that use polycrystalline diamond materialneed to be set at a negative rake angle, and this increasestorque and WOB requirements for the drill system.

The Icebreaker drill employs tungsten carbide rotary-grade (i.e., harder) cutters. This is somewhat counterintuitivesince percussive drills use softer carbide. The decision to useharder carbide was driven by several factors. First, the per-cussive energy of *2.5 J/blow is relatively low. Second, if acutter were to fracture, the failure would not destroy anentire tooth; a large fraction of a tooth would still remain,and that piece would be sufficient to cut the formation, albeitat reduced efficiency. Third, the drill, in all likelihood, willpenetrate icy regolith rather than hard basalt, hence stressesexperienced in carbide teeth would be lower. Fourth, afraction of percussive energy will be lost because the drillstring is long, it has flutes, and it includes two drill stringconnections: drill head to auger and auger to drill bit.

4.3. Sample type

Science data could be obtained from analyzing a core,regolith, cuttings, liquid, or gas, or directly from downholeinstruments. The Icebreaker mission requires cores, cuttings,or downhole instrument data.

To acquire a core, the drill has to reach the target depth,break the core, and capture it. Since the formation will not beknown a priori, the core capture system has to have the ca-pacity to retain different types of samples, ranging fromunconsolidated regolith to ice, icy regolith, and rock. Thedrill must be successfully retracted from the hole before acore can be ejected into a sample receptacle. Once the corehas been ejected, it can be analyzed and processed (e.g.,crushing/sieving). The ExoMars drill scheduled to launch in2018 on the ExoMars rover uses this approach (EuropeanSpace Agency, 2012). However, even though a core is themost desirable sample for scientific analysis, a process of coreacquisition, capture, transfer, and processing is extremelycomplicated and very difficult to accomplish robotically.

The ‘‘downhole instrument data’’ approach uses an in-strument that is either integrated inside a drill string or awire line instrument that is lowered into the hole after thedrill string has been pulled out. Since one cannot be certainthe borehole will remain open after the drill sting has beenpulled out, the wire line approach carries a substantial risk.On the other hand, from the operational standpoint, a drill-integrated sensor is the lowest-risk approach to planetarysubsurface investigation. Examples of drill-integrated in-struments include a neutron spectrometer (Elphic et al., 2008)and a laser-induced breakdown spectroscopy system (Mor-eschini, 2011). As the drill penetrates into the subsurface, theinstrument can be turned on to perform real-time sensing.Since the instrument is inside the drill string, the drill has toreach a target depth for the mission to be successful. The drill

FIG. 8. Selected drill telemetry in 45 MPa Indiana limestone(Paulsen et al., 2012).


does not have to come out of the hole, and there is also noneed to acquire and transfer samples. From the complexitystandpoint, this approach is very simple and robust andhence ideal for ground-truthing, but there is a limit on thequality and quantity of acquired data, since an instrumenthas to be designed to fit within the narrow confines of a drillstring. For the Icebreaker mission, currently there are noexisting or under-development life-detection instrumentsthat could fit within the drill string. However, the Icebreakerdrill string will have an integrated thermocouple to measurethe ground and borehole temperature and the temperatureduring the drilling process (see Section 4.6). In addition, adownhole camera and electrical resistivity sensor are beingconsidered as additional downhole instruments (see Sections4.4.4 and 4.5.1.1).

Any drilling process generates cuttings, and these cuttingscould be a viable sample for a number of instruments. Forthis reason, the Icebreaker team decided to develop a drillthat would acquire cuttings and transfer them to a range ofinstruments. It should be pointed out that the MSL drill alsoacquires cuttings for CheMin and SAM instruments, and thissampling method has already been demonstrated on Mars(see Fig. 5).

4.4. Components of the Icebreaker drill

The Icebreaker drill is shown in Fig. 9. It consists of adeployment boom, Z-stage, rotary-percussive drill head,auger with a bit, and sampling system. The following sec-tions describe each of these subsystems in detail.

4.4.1. Deployment boom. Icebreaker deploys from thespacecraft deck by way of a 3-DOF boom. This cantilevered

boom enables the drill to reach multiple holes, avoid obsta-cles, and operate at an angle, if necessary. A camera on thelander deck would guide position of the drill. The boomneeds to counteract the drill’s Z-stage force with a preloadapproximately two times larger than the maximum allow-able WOB. During the drilling operation, the Z-stage appliesa WOB, pushing the drill downward but in turn raising theZ-stage, reducing the boom preload. For example, if theboom initially preloads the Z-stage with 200 N and the drillWOB reaches 80 N during the course of drilling, the netboom preload on the Z-axis will be reduced from 200 to120 N. Once the boom preloads the Z-stage, no additionalboom actuation is required during drilling operations. Thedrill is a stand-alone device. This split of functions simplifiessurface operations and could also minimize control elec-tronics. When the boom is not moving, the three motorcontrollers that govern the boom rotary actuators could befree to govern the drill actuators.

It should be noted that the ability of the drill to be de-ployed by a 3-DOF boom offers an alternative sampletransfer option. The boom can lift the drill with a samplecaptured at the bit and position the end of the bit above theinstrument inlet port. This option is described further inSection 6.3.

4.4.2. Z-stage. The Z-stage is the main structuralmember of the drill and is preloaded against the surface by aboom as shown in Fig. 9. The part that touches the ground,called the preload shoe, is made of a tube slightly larger thanthe bit diameter (hence, the auger and the bit could easilypass through). Its bottom edge contains sawtooth featuresto enable better grip on smooth surfaces. The preload shoeis part of an auger tube, which is a natural extension of a

FIG. 9. Components of the Ice-breaker drill. The insert shows de-tails of the sampling system. Colorimages available online at www.liebertonline.com/ast

1176 ZACNY ET AL.

borehole. As cuttings move up the borehole to the surface,they continue to climb up until they reach the opening in theauger tube (see Section 6 for more details).

The drill head moves up and down the Z-stage via acarriage connected to a set of cables on both sides. Unlikeprior drills such as the CRUX, which used a balldrive (i.e., ascrew) to advance the drill head and in turn the auger drill,the Icebreaker drill uses cables and pulleys. This approachhas several benefits. It is more dust tolerant than a ballscrew,it reduces vibrations from the drill head to the drill structureand in turn the spacecraft, and it is also much lighter. Thetwo cables on either side of the carriage contain load cells. Itis a differential from these two load cells that provides WOBfeedback to the drill controller.

4.4.3. Drill head. The drill head is a two-actuatorsystem—the first actuator rotates the auger while the secondactuator drives the percussive mechanism. The percussivemechanism is based on a cam-spring principle. Voice coiland ultrasonic approaches were rejected because of thecomplex electronics required to drive them, relatively highratio of the blow energy to system mass, potential thermalissues, and very low Technology Readiness Level (TRL) atthe time of making the decision. Pneumatic and hydraulicsystems were rejected because of very low TRL and addedcomplexity required for fluid or gas management. A ‘‘dogclutch’’ was rejected because of low efficiency, heating due tosliding friction of two clutches against each other, and lim-ited range of blow energy.

The cam-spring system was selected because it is robust,can be actuated by existing motors, offers relatively highenergy per system mass, and has Apollo Lunar Surface Drill(ALSD) heritage. The main difference between the ALSD andIcebreaker is that the ALSD percussive mechanism was en-closed in a pressurized cylinder filled with nitrogen gas. Thisenabled lubrication and heat dissipation. Although theALSD drill head leaked nitrogen gas when in vacuum, theleak rate was low enough to ensure successful operation onthe Moon for at least 3 days (the duration of the Apollosurface mission). For Mars applications, such an approachwould not work, since the drill head would lose all its ni-trogen en route to Mars (a 7–9 month journey). For thisreason, the mechanism has been redesigned with new ma-terials and low friction coatings. The system has been thor-oughly tested at Mars pressure conditions to evaluate itstemperature and wear characteristics (Zacny et al., 2010b). Todate, the cam-spring percussive system successfully com-pleted over one million cycles at Mars pressure with novisible damage. This is equivalent to a continuous operationof over 8 hours, sufficient to drill eight, 1 m deep holes.

The drill head contains a six-channel slip ring that enablesintegration of sensors. The slip ring is used as a conduit forthermocouple wires to a temperature sensor embedded in-side a drill bit, an optional downhole camera, and an op-tional regolith resistivity measurement sensor (see Section4.4.4).

4.4.4. Drill auger. A single-string auger is made up ofthree parts as shown in Fig. 10: top auger, sampling auger,and drill bit. The shorter sampling auger is directly above thedrill bit. Its flutes are 6 mm deep and 10� steep designedspecifically for retaining cuttings. The long top auger, on the

other hand, has 2 mm shallow and 30� steep flutes for pre-venting borehole collapse and for efficient movementof cuttings and fallback material out of the hole. It shouldbe noted that this dual auger system has been specificallydesigned for the bite sampling approach, whereby the drillis periodically lifted out of the hole to deposit cuttingsfrom the lower sampling auger. Although the sampling au-ger is not ideal for conveying of cuttings out of the hole, itis an effective way to retain sample from the bottom of thehole.

The Icebreaker drill auger is required to move ice chips tothe surface, where they can be delivered to science instru-ments. Augers work very well if the friction coefficient be-tween auger material and cuttings is much lower than thefriction coefficient between the same cuttings and the bore-hole wall (Zacny and Cooper, 2007c). The friction betweenice chips and aluminum or steel (some of the candidatematerials for the Icebreaker drill string) increases by at least afactor of 5 as temperature falls from near zero to -65�C (Ta-lalay, personal communication, May 26, 2013). At these lowtemperatures, friction of ice on ice is lower than friction of iceon steel or aluminum (Schulson and Fortt, 2012). Hence, it isexpected that the auger will not move ice chips to the surfaceas well at low temperatures on Mars. The bite samplingapproach compensates for the effect of the high friction co-efficient on auger conveyance.

The auger has a diameter of approximately 25 mm. Thisparticular size was selected as a compromise between twoconflicting requirements: drilling energy/WOB and samplevolume. In general, drilling energy and WOB decrease as au-ger diameter gets smaller. However, auger strength and stiff-ness as well as sample volume increase as diameter gets larger.

The hollow auger allows the wires for the bit temperaturesensor, the downhole conductivity sensor, and the downholecamera to pass through. A heat pipe could also be integratedinto a hollow auger. The pipe could transfer heat generatedduring the drilling process at the bit to the section above thesurface. Alternatively, a long heater could be embedded in-side the auger, or resistive heater patterns could be depos-ited directly onto the outside of the auger (see http://www.mesoscribe.com) to unfreeze the drill string in case it freezesin place (see Section 4.5.1.1). The field-tested auger is 1.2 m

FIG. 10. Triple-stage auger: top auger, sampling auger, anddrill bit. The bit has tungsten carbide cutters and an em-bedded temperature sensor. The sampling auger capturesand retains the cuttings while the top auger efficiently movescuttings and fallback material out of the hole.


long; however, the final length (or upper limit) is a functionof the space available on the lander.

4.4.5. Drill bit. The drill bit design as shown in Fig. 11has been specifically designed to perform well in ice and icysoil (i.e., cutters are placed at zero rake angle) and also inhard, competent rocks (cutter tips are slightly rounded offso as not to break during hammer drilling). The drill bitincludes borehole reamers—these are additional cutterspointed upward to help the drill during retraction from theborehole in case of borehole collapse or rock inclusions.

The drill bit uses tungsten carbide cutters to efficientlybreak the rock. The carbide cutters have been selected basedon a compromise between hardness and toughness. Hard-ness is required to avoid premature cutter wear when dril-ling hard and abrasive materials. Toughness is required toprevent cutter fractures when the percussive system is em-ployed in hard rocks.

The cutter rake and side rake angle were designed to re-duce drilling torque and to sweep the drilled cuttings to thestrategically placed junk slots. Junk slots that are incorrectlydesigned or placed can substantially increase drilling torque,temperature of the bit and the formation, and in extremecases even halt the drill advancement entirely. The Ice-breaker junk slots are conduits for cuttings movement fromthe bottom of the hole and into the auger and hence arecritical to the bottom hole cleaning process and in turn effi-cient and cool drilling.

The drill bit also has an embedded temperature sensor.The sensor allows continuous monitoring of the thermalenvironment around the drill bit. This is a critical telemetrythat is used as input in the drilling algorithm to preventpossible warming of the sample. The thermal data, alongwith drilling telemetry such as power and penetration rate,are used to evaluate drilling efficiency (Zacny et al., 2008;Szwarc et al., 2012). Temperature data are especially criticalwhen drilling icy formations. In that case, the temperaturesensor is used to estimate the temperature of the sur-rounding ice. If the ice temperature gets close to melting,the following three actions could be done: (1) drill rota-tional speed can be reduced; (2) the drilling can be stoppedaltogether, the bit retracted 1 cm above the bottom hole,and the auger slowly rotated; (3) the drill can be com-pletely pulled out of the hole until the borehole coolsdown.

Because martian soils have clay minerals and salts, thesoils can be partially or fully thawed at temperatures below0�C. Chevrier et al. (2009) determined theoretical eutecticvalues for sodium and magnesium perchlorate solutions aspotential liquid brines at the Phoenix landing site to be236 – 1 K for 52 wt % sodium perchlorate and 206 – 1 K for44.0 wt % magnesium perchlorate.

The electrical conductivity of regolith depends on manyproperties such as the degree of saturation, porosity, com-position (conductivity) of the pore water if present, miner-alogy (fraction of charged particles), salt content, andtemperature (Mitchell, 1992). Since the electrical conductivitywill largely depend on the mobility of ions and in turn on thefraction of unfrozen water content, measuring conductivitywill enable determination of the physical state of the material(liquid or solid). Measuring resistance (inverse of conduc-tivity) will no doubt give a much better indication of theregolith physical state than measuring its temperature. Thisis because the temperature at which regolith becomes stickyor soft might vary depending on many factors such as ionconcentration and water saturation. A more direct indicationof the change in physical state (solid–liquid) is a change inthe formation’s electrical resistance. An optional soil resis-tivity measurement sensor may be fitted into the drill bit.

Clay-rich regolith exhibits three regimes: low conductivity( <-6�C), transition (-6�C to -3�C), and high conductivity( >-3�C) (Zacny and Cooper, 2005). Hence, identifying thetransition zone where regolith starts to thaw would be ofparamount importance (this is because the regolith thaws at- 6�C and not 0�C). If the drilling software is set up to slowdown drilling at 0�C (or just below) and not at - 6�C, thedrill might freeze in the hole.

To measure the resistance of the cuttings during drilling,two electrical prongs can be embedded inside the bit in sucha way as to protrude slightly below the bit shank as shown inFig. 12 (note that the drill in Fig. 12 is an old prototype bitand is not used in the Icebreaker system). Figure 13 showsthe temperature and resistance of clayey regolith during adrilling test. The resistance at a bit temperature of - 5�C,

FIG. 11. Icebreaker tungsten carbide drill bit. Color imagesavailable online at www.liebertonline.com/ast

FIG. 12. A diamond-impregnated drill bit with two elec-trodes sticking out. The electrodes are used to measure re-sistance of the cuttings during drilling and in turn infer theirphysical state (frozen or thawed). Color images availableonline at www.liebertonline.com/ast

1178 ZACNY ET AL.

- 4�C, and - 2�C was 1 MU, 300 kU, and 150 kU, respectively.At the onset of drilling and when the bit was pulled out ofthe hole after the test, the resistance was too large to measure( > 40 MU). The rate of resistance decrease per degree Celsiuswas 700 kU (-5�C to - 4�C) and 75 kU (-4�C to - 2�C). Thisindicates that ion mobility was sharply increasing as thetemperature was approaching 0�C. This test validated theresistivity measurement approach as a viable method to beemployed in future planetary drills to monitor the state ofthe drilled material during the drilling process.

It might be feasible for pockets of brines to be present insubsurface ice. In fact, in the Siberian Arctic there are lensesof sodium-chloride water brines (called cryopegs) that haveconstant temperatures in the range of - 9�C to - 11�C. Thesecryopegs are sandwiched within permafrost marine sedimentsthat are 100–120 thousand years old (Gilichinsky et al., 2003).A number of bacteria were isolated from these cryopegs thatexhibited the ability to survive and develop under harshconditions, such as subzero temperatures and high salinity(Shcherbakova et al., 2004). This resistivity approach will iden-tify such a condition and trigger software to pull the drill out ofthe hole immediately, to prevent freezing the drill in place.

In addition, Fletcher (1968, 1970) predicted the existence ofa quasi-liquid surface within about - 6�C to - 1�C, withthickness of a few molecular layers, increasing to about 10layers at - 1�C. Although these results, based on free energycalculations, are speculative, they do fit a similar tempera-ture profile observed in formations containing clay minerals.

4.4.6. Brushing station. The brushing station uses a pas-sive brush with long steel bristles to ‘‘scrape’’ drilled cuttingsoff the auger flutes (Fig. 9). The brush wheel forms a worm-gear configuration with the auger. Hence, as long as the augerrotates, so does the brush. If space is limited, a linear brushcould be used instead. The brush is also used to clean theauger before it is inserted back into the hole to minimize crosscontamination. Since the brush is in direct contact with thedrill, it will have to be sterilized and stored inside a biobarrier.

4.5. Drilling software

Before actual drilling commences, the drill will need to bedeployed from its stowed and locked horizontal position. Oncedeployed, the drill will have the following modes of operation:

(1) Idle: All motors are standing idle awaiting commands.(2) Warm-up: Heaters are activated (if required) to warm

the motors up to a temperature where they are safe tooperate.

(3) Seek: The drill searches for the ground level or thebottom of the hole. This operating mode accounts forpotential fallback of material into the hole to protectagainst stalls and potential corkscrewing. During thisstep, the bit temperature would measure subsurfacetemperature.

(4) Drill: The drill penetrates the subsurface in short in-crements called bites, with all necessary automatedsafety checks and fault recovery procedures. In addi-tion, temperature and potentially formation electricalresistivity data are acquired and analyzed in real time.

(5) Retract: The drill retracts from the hole.(6) Sample transfer: The sample is transferred into an in-

strument. There are three options as described inSection 5.

The first steps that lead to preloading of the drill structureagainst the surface are relatively easy to accomplish andhence will not be dealt with in this paper. However, theactual drilling processes (Seek, Drill, Retract, etc.) requiredevelopment of robust software and implementation of low-risk drilling protocols. This section describes the drillingsoftware.

Large communication delays between Earth and Mars (upto 12 hours) justify robust drilling algorithms and protocols.If the drilling system relies entirely on software to makedecisions and continues drilling as fast as possible withoutretracting the auger out of the hole, there is a risk that thesoftware will not be able to deal with all drilling conditions.As a consequence, the drill could get damaged or perma-nently stuck and prematurely end the mission. An alterna-tive approach is to limit the drilling software to a set of well-defined subroutines. Here, any anomaly, however small,would require the drill to be pulled out of the hole and re-main in its ‘‘home’’ position for further commands fromEarth. This approach lowers drilling risk at the expense ofdrilling speed and energy.

The first approach has been developed by NASA AmesResearch Center while the second approach has been tackledby Honeybee Robotics. Most probably, the final Icebreakersoftware will be a compromise between the two approaches;however, for now these two approaches are being pursued inparallel.

The NASA Ames software includes diagnostic softwarefor monitoring the state of the drill, and contingent executionsoftware for guiding the drill through a daily drilling planwhile at the same time helping the drill recover from off-nominal situations (Glass et al., 2005, 2011). The software has

FIG. 13. Electrical resistance and the drillbit temperature during drilling in clayeysoil. (Zacny and Cooper, 2005).


been undergoing extensive field testing in the Arctic, theAntarctic Dry Valleys, and in a Mars environmental cham-ber. The Honeybee Robotics approach is divided into mis-sion-critical and mission-noncritical routines. Criticalroutines deal with the conditions that would lead to the endof the mission, while noncritical routines deal with drillingoptimization during nominal conditions.

In both cases, the software relies on drill telemetry to inferthe state of the drill and to control the drill operation viarule-based and model-based algorithms and subroutines. Ifan off-normal condition is detected, in all cases the drill isrequired to go back to its safety mode (retracts from the holeand empties the auger by moving the auger past the brush)and awaits further instructions from mission control.

4.5.1. Mission-critical events. Mission-critical events arethose events that could lead to a mission failure (e.g., drillstuck in the hole). To avoid them, the Icebreaker drill relieson its robust drilling software and drilling protocols. Ingeneral, the two environments that should be considered areformations with ice and without ice. The next sections de-scribe various events that could occur in both of these en-vironments. It should be noted that, although only some ofthe events described below were observed either in the field,in a lab testing, or in the Mars chamber testing, all the inci-dents could in fact occur on Mars. It needs to be emphasized,though, that in some cases, it might also be difficult to de-termine what type of event occurred since two or moreevents may result in the same outcome (e.g., drill stuck).

4.5.1.1. Formation containing ice. If a formation contains ice,there are several conditions that can occur that would pre-maturely end the mission. These conditions include drillfreeze in a hole, bit icing, auger flutes icing up, and drillchoked up with cuttings.

If drilling is aggressive (uses a lot of power), the ice occursnear the surface, and atmospheric pressure is near the triplepoint of water, it is possible that the energy released duringthe drilling process as heat is used up almost entirely by thelatent heat of fusion plus vaporization. As ice turns intovapor, it not only uses up all the heat generated duringthe drilling process (and this in turn keeps the borehole andthe drill bit at the freezing point or below) but also blows thecuttings out of the hole (Zacny et al., 2004). This scenario,however, only occurs if the drill starts drilling aggressivelyright from the beginning and cuttings are constantly clearedout of the hole. If for some reason sublimation stops andcuttings fill the auger flutes, restarting this process will bealmost impossible. Water vapor forming near the rotatingdrill bit will create a pressure-cooker effect; that is, the vaporpressure will increase above the triple point, which leads tovapor changing its state to liquid. That water would bethermodynamically unstable and could revert to ice. Thisscenario may also occur if icy regolith or ice is covered by athick layer of dry regolith. That regolith will form a tightrestriction and prevent any water vapor from escaping ex-plosively out of the hole. There are a number of possibleoutcomes, and each of them has been observed during lab-oratory and field tests.

In the first instance, the ice can stick to the drill bit andprevent cutters from biting into the formation and ceasepenetration (see Fig. 14). Although the drill would not be

stuck, further progress would be stopped. The drill wouldhave to be pulled out of the hole and the ice allowed tosublime away. Pointing the drill bit toward the Sun wouldspeed up the process (see Fig. 15). Ice sublimation on Mars isnot a fast process but was observed by the Phoenix missionin the trench informally called ‘‘Dodo-Goldilocks’’ over thecourse of 4 days (Mellon et al., 2009).

In the second scenario, the ice could refreeze within thecuttings. These cuttings could be augered out of the hole anddumped onto the ground. This scenario would not be mis-sion-ending, but the sample would most likely be lost sincelarge chunks of regolith would be impossible to deliver to aninstrument. Regolith clumping was observed during thePhoenix mission (Arvidson et al., 2009). In this case, regoliththat might have had some water ice and perchlorate saltsbecame sticky and formed clumps that did not pass easilythrough a screen above instruments’ inlet ports. It was ob-served that regolith on the sunlit side of a scoop stuck to thescoop, while regolith on the shadowed side successfully fellout of the scoop onto the screens (Fig. 16). Hence, granular,loose sample can be maintained by keeping the formationand the sample as cold as possible.

In the third scenario, ice can also form between the augersurface and the cuttings; that is, icy cuttings would stick tothe auger surface and would not move up the auger (seeFig. 17). This condition is serious because further conveyanceof cuttings from the part of the hole below the restrictionwould not be possible. In this case, the penetration ratewould drop, which would trigger an increase in WOB. Thetorque would also increase since the cuttings below the re-strictions would be pushing against the borehole and theauger at the same time, causing the auger to slow down (thisis referred to as choking). If this event is identified fast en-ough by the drilling algorithm, the drill should be immedi-ately pulled out of the hole. As before, if the auger is exposedto the Sun, the ice could sublime away while regolith wouldeither gravity fall or could be brushed off with a passivebrush, which is part of the sample delivery system within theIcebreaker drill.

The fourth scenario pertains to liquid water freezing thecuttings and freezing onto the borehole. Hence, the borehole

FIG. 14. Ice at the bottom of the hole could thaw and re-freeze around the cutters, thereby stalling penetration rate.Color images available online at www.liebertonline.com/ast

1180 ZACNY ET AL.

turns into a nut, while the drill is a screw (Fig. 18). Thisscenario could occur if the auger surface is coated by a non-stick coating. The drill could potentially be moved a little bitup or down (and this has been observed), but screwing thedrill out of the hole entirely is prevented by the sample-retaining auger of the Icebreaker drill and could never beachieved. This is because the lower flutes are designed toretain sample, not to transport it upward along the flutes asthe top section of the auger does.

If this occurs, whether the drill would have the capacity torecover or not depends on the strength of the formation (i.e.,water content and temperature) and on the drill power (ro-tary and percussive). Trying to get the drill out of the holewould be essentially like trying to get a rebar out of concrete.This is in fact a good analogy, since the strength of frozenregolith is as high as or even higher than the strength ofconcrete (Zacny et al., 2007). It might be possible for thepowerful drill to deliver enough torque and percussive en-ergy to shear away the iced-up ‘‘thread,’’ but this approachmay not work in most, if not all, cases. In the most likelyscenario, the drill would never be recovered.

In the fifth scenario, the water ice can freeze within thecuttings and bind onto the borehole and the auger. This is theworst case, and such an event would most likely perma-nently trap the drill in a hole. Such occurrences are commonin the polar regions. Figure 19 shows a set of tools that were

used to recover a stuck drill (bottom of the picture) in theAntarctic Dry Valleys. It took two people over 6 hours to getthe frozen drill out of the hole. It should be noted that any ofthe four previous conditions could lead to the fifth one (i.e.,drill frozen in a hole).

If a drill freezes in a hole on Mars, the only way out is towarm up the drill string and melt the ice. Assuming a 1 mlong and 25 mm diameter drill is stuck in ice at - 55�C andhas a thermal conductivity of 2.5 W/mK, the power requiredto reach 0�C at the surface of the auger (i.e., melt ice stickingto the surface of the auger) would be approximately 300 Wand take *24 minutes (Fig. 20). This is a theoretical ap-proximation, and tests would have to be conducted to de-termine the actual power and time required (and in turnrequired energy). This contingency plan would require in-tegration of a line heater inside the auger capable of with-standing the hammering action of the drill.

To prevent any of the above conditions from occurring,the drill bit has an integrated thermocouple for monitoring

FIG. 15. The boom can point a drilltoward the Sun to speed up de-icing viasublimation. Color images available on-line at www.liebertonline.com/ast

FIG. 16. Mars Phoenix’s scoop after it had been invertedand vibrated to jar the sample of icy soil loose. The imageshows that much of the sample was clumpy and remainedstuck inside the scoop. Credit: NASA.

FIG. 17. Ice (left) and icy soil (right) refroze within theauger flutes. Color images available online at www.liebertonline.com/ast


the bit temperature. The temperature data are then fed into adrilling algorithm that monitors the absolute temperatureand the temperature increase. If the temperature starts rap-idly rising or rises slowly but approaches the upper bound ofthe temperature limit (currently set at 0�C), the drillingsoftware reduces rotational speed and lowers WOB. If thisdoes not result in a drop in temperature, the drill is lifted offthe hole bottom (i.e., WOB is zero) while the auger is rotatedat slow rpm. If the temperature drop is still not satisfactory,the final step would be to pull the drill out of the hole andallow the subsurface to cool down.

If a formation contains clay minerals or salts, temperaturealone cannot be used to infer physical state of ice. This isbecause brines can form at a temperature much lowerthan 0�C. Clayey regoliths also exhibit freezing point de-pression. The only certain method of determining the onsetof water ice changing phase to liquid is to continually mea-sure the cuttings’ conductivity. This method is described inSection 4.4.4.

4.5.1.2. Dry formation (no water ice present). In formationsthat have no ice (dry formations), the biggest problems in-clude an auger choked up with cuttings or fallback material(hole collapse) and loose pebbles either within a boreholewall (leading to a binding fault) or at the bottom of the hole.

Choking may occur because the auger has not been de-signed properly to efficiently move the cuttings to the sur-

face, or the auger is not rotated at high enough rotationalvelocity, or the drill bit has poorly located junk slots. Thechoking due to borehole collapse is much more difficult todeal with. In both cases, having a more powerful rotarymotor with ample torque might be enough to recover thedrill. However, a better approach is to implement a drillingprotocol that would keep the hole open by frequent reaming.This will not require cuttings to be moved along the entirelength of the auger. The bite sampling approach solves theauger-choking problem by implementing both solutions.This approach is similar to peck drilling in machining pro-cess, where a drill bit is periodically removed to clear metalshavings. Since the drill is moved down and up severaltimes, it reams the hole, which reduces the chances of itgetting stuck and helps reduce parasitic losses due to augerflutes rubbing against the borehole.

If a rock is protruding from the borehole wall and causinga binding fault (this will be identified by a torque spike oc-curring two times per revolution for two-flute augers), itmight be possible to try to ream through this area. However,if a rock is relatively small, it might completely dislodge andeither fall to the bottom of the hole or continue rubbingagainst the auger. Unfortunately, there is no robust methodto deal with loose rocks. It might be possible to try to drillthrough them to reach more consolidated formation; how-ever, this could cause cutter fractures due to the impact ofcutters on loose rocks.

FIG. 18. The grooves in icy soil are animpression of auger flutes. Color imagesavailable online at www.liebertonline.com/ast

FIG. 19. A set of tools used to recover a stuckdrill: a vacuum for moving frozen chips out ofthe hole, a long drill for over-drilling aroundthe stuck drill string, a long corer for drillingover the stuck drill string, and a brush. Colorimages available online at www.liebertonline.com/ast

1182 ZACNY ET AL.

4.5.2. Mission-noncritical events. The mission noncriti-cal drilling aspects pertain to optimizing drilling parametersto minimize drilling specific energy. Specific energy is theenergy required to drill a unit depth of the hole. This termtakes into account power and penetration rate. Hence, forexample, the more efficient drill will require 100 Whr to drillto 1 m depth, while the less efficient drill will need 200 Whr.It is worth pointing out that minimizing drilling specificenergy also helps reduce the mission-critical events in icyformations, since less heat is being generated per depth ofthe hole.

For the Icebreaker system, the software uses a so-called‘‘energy routine.’’ In this rule-based routine, the drillingproceeds in rotary mode as long as a penetration rate above10 cm/hr is being maintained for WOB < 40 N. If the pene-tration rate drops and maintaining it would require in-creasing WOB to above 40 N, the percussive actuator isengaged. The percussive system increases the penetrationrate but at the expense of higher power. In soft rocks andregolith, high penetration rate can be achieved with rotarydrilling alone. However, in harder formations, percussivemechanisms need to be engaged.

4.6. Drill as a science instrument

The primary purpose of the drill is to provide samples forthe science instruments. The drill, however, could also beused to augment the instrument data with information aboutformation strength and temperature.

In particular, the drill bit includes an integrated sensor formonitoring temperature during drilling operations. Initially,when the drill is lowered to the ground for the first time, thebit could be used to obtain the temperature of the ground.That temperature is used as a reference point and to calibratethe thermal model for the drilling algorithm. Every subse-quent time the drill is lowered into the hole to take another‘‘bite,’’ it could acquire a borehole ambient temperature atdifferent depths. These single data points could then beplotted as a function of depth to reveal a thermal gradient.The thermal gradient combined with an estimate of forma-tion thermal conductivity could then be used to determineheat flow properties of Mars.

Many researchers have already demonstrated that drillingtelemetry such as power, penetration rate, and WOB couldbe used to infer formation strength and in particular un-

confined compressive strength or UCS (Teale, 1965; Mellor,1972; Reddish and Yasar, 1996; Zacny et al., 2006; Paulsenet al., 2010). In addition, Mellor (1971) demonstrated thatUCS increases with an increase in water-ice content (peakingat maximum saturation) and with a decrease in temperature(peaking at approximately - 80�C). The strength of pure icealso increases with a decrease in temperature, reaching70 MPa at -100�C (Tsytovich, 1975). Hence, by estimatingUCS from the drilling data and combining it with measuredsubsurface temperature data, it is possible to determine thefraction of ice in the formation. Once the fraction of ice andthe ice temperature are known, a thermal conductivity esti-mate could be narrowed down.

Because the total drilling power is a sum of power re-quired to break the formation, convey the cuttings up theflutes, and overcome parasitic losses due to auger rubbingagainst a borehole, the determination of formation strengthcan only be done initially with each hole, when the flutes areempty. Parasitic losses could be measured (and in turn fac-tored out) by rotating the auger in a hole *1 cm above thebottom of the hole (i.e., making sure no drilling takes place).Once the drill starts drilling, it will be more difficult to inferformation strength due to the added contribution to the totalpower from the auger conveying the cuttings.

5. Sample Acquisition

There are two possible cuttings acquisition approaches. Inthe first one, the drill can continually drill down to its targetdepth while another sample delivery system transfers thecuttings to appropriate instruments as soon as they reachthe surface. This approach has a major drawback in that theauger torque due to cuttings being augered to the surfacewould become higher with depth. In addition, the materialthat appears on the surface would not represent the currentmaximum depth but rather a shallower depth. In a 1 m longand 25 mm diameter auger that has 2 mm deep flutes, thevolume occupied within the flutes is 144 cm3. Assuming afluff factor of 2 (i.e., volume of cuttings is twice as large asvolume of originally compacted and consolidated material),the initial volume of the sample occupying the 144 cm3 is72 cm3. Since the auger diameter is 25 mm, the depth of holewith that 72 cm3 volume is 15 cm. Therefore, material on thesurface once the drill reaches 1 m depth represents a depth of85 cm (and not 1 m). To recover those additional samples, the

FIG. 20. Temperature of a 25 mm drillstring frozen in a - 55�C ice, being heatedby a 300 W integrated line heater. The radialposition is measured from the center of thedrill string. The vertical line at 0.012 m rep-resents the auger surface. Color imagesavailable online at www.liebertonline.com/ast


drill has to be retracted out of the hole, and material on theflutes that represents the bottom 15 cm formation must bescraped away into a sample transfer system.

The alternative to drilling to the maximum depth in onego is to employ a peck drilling approach, referred to as ‘‘bite’’drilling. In this approach, the drill is periodically retractedfrom the hole, the cuttings are transferred to an instrument,and the drill is lowered back into the hole to acquire the nextbite as shown in Fig. 21. This approach requires the lowersection of the auger to have deep flutes to accommodate asmuch sample as possible and also flutes at a shallow angle toretain as much sample as possible.

It should be noted that, if for some reason the hole col-lapses while the drill is retracted, the next time the drill islowered back into the hole, the fallback material will beaugered out first; then the new material will be acquired onthe sampling auger. Hence, there is no danger in samplingthe same material twice.

This bite sampling approach has several advantages. Forexample:

� The stratigraphy is more likely to be preserved sincesamples are delivered from known depths. However,the degree of ‘‘stratigraphy preservation’’ will largelydepend on the properties of the subsurface such as po-rosity and, in the case of regolith, also a regolith cohe-sion and friction angle. It should be emphasized,though, that there will always be some uncertainty re-garding the depth from which any given sample is ob-tained. The best that can be said is that if some newmaterial appears, then it must have come from thenewly drilled interval. However, materials previouslyobserved may not still be present at the bottom of thehole even if they are found in the cuttings (since theymay have fallen in from higher up). Bite sampling does,however, help.

� The drill power to reach 1 m is lower since there isminimal drag due to cuttings being augered to thesurface (cuttings are captured in the lower ‘‘sampling’’auger and do not climb up the flutes to the surface).

� If desirable, samples from a target depth can be deliv-ered while the rest can be discarded onto the ground.

� Periodically removing the drill from the hole also allowsthe subsurface and the drill itself to cool down. Thiscooling period is critical since the biggest danger in

drilling ice-rich regolith is thawing followed by re-freezing of the ice around the drill.

� The method also offers ‘‘graceful failure,’’ in that unlessthe drill breaks before acquiring its first sample, thesystem will always be able to deliver some sample foranalysis.

� Finally, every time the drill is lowered back into thehole, the borehole temperature would represent theambient subsurface temperature. Adding up thesetemperature data can reveal the thermal gradient and inturn heat flow properties of Mars, if thermal conduc-tivity is known. It should be pointed out that in order tomeasure the temperature of the hole wall on reentry thedrill stem would need to thermally equilibrate, and thiscould take some time. In addition, because the drillauger will be made out of some metal, and metals areinherently good conductors, a thermal gradient willexist between the top of the auger above the groundlevel and the tip of the auger touching the bottom of thehole. The effect of this thermal gradient would also haveto be taken into account.

It should be noted that there is going to be some level ofcross contamination due to the smearing of cuttings on theflutes against the borehole as the auger is being pulled out.The extent of cross contamination will depend on the cut-tings’ frictional and cohesive properties.

6. Sample Delivery

To reduce the risk related to sample delivery, the Ice-breaker drill enables a triple redundant sample deliverysystem compatible with a bite sampling approach. Theseinclude a scoop mounted at the end of a 5-DOF robotic arm,pneumatic transfer, and a 3-DOF deployment arm. The lasttwo sampling systems may be viewed in a YouTube video(Honeybee Robotics, 2012).

6.1. Five-DOF sampling arm and a scoop

The scoop is the most straightforward method of deliv-ering a sample and has Mars Viking 1 and 2, Mars Phoenix,and MSL heritage. Since the Icebreaker mission may have todeal with sticky material, the scoop has been designed withan integrated scraper and resembles an ice cream scoop(Dave et al., 2012). The scoop is deployed by a 5-DOF arm.

FIG. 21. The ‘‘bite’’ sampling ap-proach. Color images available onlineat www.liebertonline.com/ast

1184 ZACNY ET AL.

When the sample is ready for transfer, the scoop is posi-tioned next to the sample drop-off point as shown in Fig. 22.It can retrieve samples during drilling, drill retraction, andfrom the shallow-angle flutes as the end of the auger isbrought up. After the scoop has been filled with a sample,the arm moves the scoop and positions it above an instru-ment inlet port. The scoop scraper blade is connected to alever, and to discharge the sticky sample the lever has to bemoved through 90�. To achieve that, the arm preloads thelever against a hard stop just above the sample inlet pointand then moves the scoop forward. This forward motion ofthe arm/scoop holds the blade in place and scrapes thesample into the instrument. The implementation of the scoopoffers sampling redundancy. If for some reason the drill fails,the scoop could be used to acquire surface material foranalysis.

The scoop has been tested in the Mars analogs of theCanadian High Arctic and the Dry Valleys of Antarctica. Inaddition, the scoop has been tested in a Mars environmentalchamber and delivered a variety of materials, ranging fromnoncohesive to highly cohesive, that is, sticky (Dave et al.,2013).

6.2. Pneumatic sample transfer

Gas is a very effective medium for lofting particles in avacuum (Zacny et al., 2005). In the best-case scenario, 1 g ofgas could lift close to 6000 g of regolith at velocities ap-proaching 10 m/s in tests conducted at Mars pressures(Zacny et al., 2010c, 2010d). Such high efficiencies are possi-ble because gas exit velocity in a vacuum is proportional topressure ratio rather than pressure difference. Of course, ifthe mission requires capturing a subsurface sample, the drillsystem would have to include some kind of cuttings catch-ment subsystem or use reverse circulation drilling.

The low atmosphere not only makes air drilling possiblebut also allows implementation of a pneumatic sampletransfer. Pneumatic sample transfer is a point-to-point sam-ple delivery method (Fig. 23). The sample is first acquiredinto a small chamber while the door is in the open position.Once the chamber is full (this could be triggered by an op-tical sensor inside the chamber), the door closes. A puff ofgas is then directed into the chamber on one side, and the gaswith the content of the chamber (i.e., the sample) is movedthrough the sample delivery tube and directly into a cycloneseparator. The sample drops to the bottom of the sampledelivery hopper, while the gas escapes to the atmospherefrom the top of the cyclone separator. A hopper could thendistribute the sample via a carousel underneath it to variousinstruments or deliver a sample to a sample-processingsystem such as a crusher (Zacny et al., 2012b). Alternatively,the pneumatic system could contain several cyclone separa-tors, each one mounted above the instrument inlet port. Agas manifold would then direct the dusty gas to the appro-priate port. In each case, an additional couple of gas puffscould be used to clean up the sample delivery hardware toprevent cross contamination before the new sample is ac-quired. This sample delivery approach has been tested at 760torr while drilling Indiana limestone (Fig. 23) and inside aMars chamber while drilling ice-cemented materials (Fig. 24).

The major benefit of this sample delivery approach is thata sample can be moved directly from the drill to the instru-ment, and this could be accomplished without the need topull the drill out of the hole or even the need to stop thedrilling operation. In fact, the sample could be transferredevery few minutes while the drill penetrates into the sub-surface. To recover the sample from the bottom of the hole,the drill, however, would need to be pulled all the way out ofthe hole and the sample from the sampling auger sectionscraped into the sample chamber.

FIG. 22. Diagram stepping through the sample acquisition and delivery operation (Dave et al., 2013). Color images availableonline at www.liebertonline.com/ast


The pneumatic system of course requires gas supply. Gascould be supplied from a dedicated gas cylinder, or it couldcome from the pressurant tank. A lander’s propulsion systemcontains either a helium or nitrogen-filled high-pressure tankfor pressurizing propellant. Upon touch down, this gas is nolonger required and hence could be used for the pneumaticsample delivery system. It should be noted that MSL SAMuses helium gas as the carrier gas in the gas chromatographinstrument. Helium is used because it is inert; hence, it willnot react with the sample. Compressed helium was also usedin the Viking biology experiments. It is also suitable for a lifesearch mission as it cannot give a false positive for organicscontent.

6.3. Three-DOF-DOF arm and drill

The third approach to sample transfer uses the drill itselfand its 3-DOF deployment arm. As mentioned earlier, thedrill stores sample within the deep flutes just above the bit.Once the drill is retracted from the hole, the sampling portion

of the auger is protected inside the auger tube (Fig. 25). Thearm then lifts the entire drill off the surface and positions thetip of the bit above the instrument inlet port (Fig. 26). Posi-tioning the drill above the cup ensures there is an air gap,which satisfies NASA Planetary Protection requirements inthe case of an unsterilized lander (Dave et al., 2013). Once thesample drop-off position has been verified, the drill can ad-vance forward while slowly rotating. The sample thengravity falls directly into a cup. For more cohesive or stickysamples, the drill can engage percussion, which would alsohelp to dislodge small rocks.

An alternative approach would be to mount a brushwithin the drill docking station above the instrument inletcup as shown in Fig. 26. Hence, the brush could help withsample discharge from the deep flute auger while the augeris being rotated. An instrument could have individual inletports with individual docking stations and brushes. This willminimize cross contamination between samples since thebrush will most likely be full of cuttings after the dischargeoperation. It would also require DHMR of the individual

FIG. 23. Pneumatic sample delivery system. Initially, cuttings are acquired into a chamber (left). A puff of gas transfersthe content through a tube and into a cyclone separator on a lander deck (right). Color images available online at www.liebertonline.com/ast

FIG. 24. Pneumatic sample delivery systemundergoing tests in a Mars chamber. Colorimages available online at www.liebertonline.com/ast

1186 ZACNY ET AL.

docking station because the drill would make direct contactwith it.

7. Icebreaker Drill Tests

Drill testing is one of the critical aspects of drill develop-ment. It is the results of drill tests that drive any futurechanges to the drill design, the drilling protocols (the drillingsteps or approach), and the drilling software. Normally, thetechnology development progresses via a continuous loop ofbreadboarding, testing, modification to the breadboard ormanufacturing of a new breadboard, and further testing,until the final prototype is ready to be manufactured for aflight mission.

One important factor that needs to be emphasized is thatin order to determine the drill nominal operation, the initial

test formation needs to be very homogeneous. As such, rockswhose properties have been evaluated should be used. Ex-amples of such rocks include Indiana limestone and Bereasandstone, offering a range of strength and abrasiveness. Thedrill should also be tested at relevant vacuum conditions andtemperatures, since these would affect the way mechanicalsubsystem and actuators work. For example, drilling in hardvacuum may reveal abnormally high rotary torque due towearing off of dry lubricant on gears. If the rock is not ho-mogeneous, such a torque increase would be incorrectly at-tributed to changes in rock properties. Only once the drillsystem has been evaluated under relevant conditions is thedrill ready for tests in a range of analog environments.

Because of limitations imposed on drilling by the envi-ronment, as well as by the mission itself, the drilling processneeds to be efficient and effective. For this reason, Icebreaker

FIG. 25. Sample delivery using thedrill and 3-DOF arm. Color imagesavailable online at www.liebertonline.com/ast

FIG. 26. A brush on a dockingstation above the instrumentinlet port could assist in cuttingsdischarge. Color images avail-able online at www.liebertonline.com/ast


was designed to drill fast with lower power and low WOB.Icebreaker has been targeting a penetration rate of 1 m/hr atan average power level of 100 W and with WOB less than100 N. Hence, a term 1-1-100-100 was coined, which refers todrilling at 1 m in 1 hour with 100 W and 100 N WOB. Thesefour are also primary drilling parameters from which othermetrics such as drilling energy or penetration rate can becalculated. Meeting the 1-1-100-100 level would mean thatthe Icebreaker drilling mission to Mars fits within any ex-pected mission limitations or requirements.

7.1. Test environment

Tests in relevant environments advance the technology toa higher TRL. For the drill to achieve the highest possibleTRL on Earth, that is, TRL 6, the drill has to be tested insimilar conditions as it will experience en route to a targetplanetary body and while performing tasks on that body.TRL 6 is defined as a system/subsystem model or prototypedemonstration in a relevant environment (ground or space)(Mankins, 1995).

Testing of drills or any other hardware is an essential butvery expensive element of the technology developmentprocess. For that reason, testing is a fine balance betweendeciding what needs to be tested, how it should be tested(under what conditions), and what budget is needed. Inpractice, the project budget drives the first two. In addition,testing itself is not the only major task. Building of experi-mental setup (chambers, ground support equipment, testbeds), as well as analyzing the data, could take as much time(or more) as performing the actual tests.

For the Icebreaker drill targeting Mars, the relevant envi-ronment includes ambient temperature, pressure, atmo-spheric gas, humidity, formation type and temperature,gravity, and vibration and thermal environment duringlaunch from Earth and during EDL at Mars.

Gravity, though an important aspect to drilling, is im-possible to simulate on Earth. Hence, the gravity was dealtwith by scaling down the maximum drilling forces by a ratioof martian to terrestrial gravities. In some cases, certain as-pects of drilling such as pneumatic cuttings transfer could betested during reduced-gravity flights. These flights allow fora target gravity level (e.g., Moon, Mars, and microgravity) tobe maintained for approximately 20 s. If vacuum is a criticalaspect, a vacuum chamber could be placed inside a reduced-gravity plane. Hence the effect of vacuum and the effect ofgravity could be tested at the same time (Zacny et al., 2011). Ifa longer-duration test is required, a possible option could beoffered by commercial suborbital flights. Private companiessuch as Virgin Galactic, XCOR, Armadillo Aerospace, BlueOrigin, and Masten Space Systems are all developing sub-orbital spacecrafts, and these will be available for testingscience payloads. If a few minutes of reduced gravity is notsufficient, the last resort, albeit an extremely expensive one,is to test the system within a centrifuge on board the Inter-national Space Station.

To meet the vibration requirements, the drill would needto be tested across frequencies it will experience during thelaunch and EDL. To meet the thermal requirements, the drillwould need to be tested in a thermal vacuum chamber acrossa range of temperatures it will experience en route to Marsand while on Mars’ surface. Designing a drill to such re-

quirements and performing actual tests is very costly; hence,it was decided that the Icebreaker drill would not undergovibration and thermal tests, but instead all tests will be fo-cused on drilling performance.

For the Icebreaker drill development with application tomartian ice-cemented ground, a testing sequence was di-vided into two complementary efforts. One part involvedtesting in a vacuum chamber, while the other part involvedtesting in an analog environment on Earth. The reason forsuch a split was to take advantage of the major learning andimprovement opportunities each environment had to offer.

7.2. Testing in a Mars environmental chamber

The Icebreaker drill was thoroughly tested in a Marschamber for two reasons: to determine whether its criticalpercussive mechanism would perform efficiently in a low-pressure Mars atmosphere and to acquire drilling telem-etry and identify drilling faults in various Mars analogformations.

A box-shaped Mars chamber, shown in Fig. 27, was es-pecially designed and built for testing drill systems to at least1 m depth. The chamber itself is 3.5 m high to maximize headroom for integrating a long drill string and a test sampleunderneath it. The height of 3.5 m could also be extended inboth directions (up and down) by adding 50 cm diametercylindrical extensions to the existing 50 cm flanges at the top

FIG. 27. The Icebreaker drill integrated into a Mars envi-ronmental chamber. Material for drilling is placed inside acylinder and actively cooled with cooling coils. Color imagesavailable online at www.liebertonline.com/ast

1188 ZACNY ET AL.

and the bottom of the chamber. The chamber has a footprint1 m · 1 m to enable integration of large samples with coolingcoils and still have plenty of room for instrumentation. A setof 12 acrylic windows integrated into the two chamber doorsallows viewing of the drill and drill operation. A set of cam-eras is also placed at various strategic locations within thechamber, and movies are taken of the entire drilling process.

The chamber does not have an in-wall cooling system.Instead, only the sample and potentially the drill could becooled (although this was not performed during the tests).Cooling of the sample was achieved via cooling coils con-nected to either a closed-loop refrigeration system or liquidnitrogen. The closed-loop refrigeration has a cooling capacityof 150 W at its lowest temperature of - 25�C and allows closecontrol of the temperature. The liquid nitrogen cooling sys-tem allows reaching much lower temperatures, in the rangeof - 200�C. The sample temperature could be controlled byincreasing or decreasing the flow of liquid nitrogen coolantthrough the coils, but the temperature is not as easy tocontrol as in a closed-loop cooling system.

The Icebreaker drilling parameters that can be controlledinclude WOB, rotational speed of the auger (rpm) and per-cussive frequency. Drilling telemetry includes drilling timeand depth (which are converted to ROP), auger power andpercussive power (which are converted to total power),drilling energy (product of ROP and total power), and bittemperature.

Drilling power is a sum of auger power (rotary torque *rpm) and percussive power (impact energy * impact fre-quency). Since auger rpm is the same at 100, the auger poweris therefore a direct function of auger torque. While testing,the percussive system was activated only if the WOB re-quired to maintain a set penetration rate exceeded 40 N.Percussive power was constant at 35 W since neither impactenergy (2.6 J) nor frequency (810 BPM) was changed. Per-cussive frequency can be changed by changing rpm of theactuator, but during our tests, this parameter was keptconstant. Percussion was either turned on or off; hence forthe purpose of energy calculation, the average percussivepower was calculated as a fraction of time the percussivemechanism was turned on. For example, if the percussionsystem was on for only 25% of the time, the average per-cussive power was 0.25 * 35 W = 8.8 W.

The Mars analog formations used for the Icebreaker drilltesting included

(1) Mars Mojave Simulant regolith with a large fraction ofsmall pebble-sized rocks, which was saturated withwater containing 1–2% perchlorate salt and frozen to- 200�C;

(2) Mars Mojave Simulant regolith saturated with waterand frozen to - 20�C;

(3) Ice at - 20�C;(4) Ice with 1–2% perchlorate at - 20�C.

Mars Mojave Simulant is a basaltic Mars simulant made fromcrushing of saddleback basalt collected in the Mojave Desert(Peters et al., 2008). Perchlorate salt was added because thePhoenix and recently the MSL missions indicated the pres-ence of perchlorate salts on Mars (Hecht et al., 2009).

The first drilling tests were conducted on Indiana lime-stone. Although that rock is not the best Mars analog, it is anideal rock for verifying nominal drill operation and estab-

lishing a benchmark. Indiana limestone is (1) a uniform rock(2) of medium strength (45 MPa UCS) and is (3) non-abrasive.

The first attribute—uniform rock—means that any chan-ges in drilling telemetry can be identified as off-nominal drilloperations or possibly inefficient cuttings removal. Thiscould be dealt with by frequently tripping the drill out of thehole to clean the auger of cuttings. If the rock had layering,fractures, and so on, these would make the interpretation ofdrilling telemetry more difficult. Rock uniformity essentiallyreduces the number of test variables.

Medium rock strength allows sufficient testing of drillmotors, mechanisms, and drill bits without unduly wearingthese components out. Since strength is medium and notlow, the percussive mechanism would be engaged. Softerrocks would not require percussion, while drilling in veryhard rocks would take longer time and wear out the drill bitfaster.

Since the rock is non-abrasive, the bit wear would beminimal. Drilling in abrasive rocks (rocks that containquartz) would significantly wear out the bit, and in turn drillperformance could be different at the beginning and at theend of a 1 m hole.

A reason for establishing the base mark for the Icebreakersystem is that after drilling additional holes in various Marsanalog formations, the drill could be tested in Indianalimestone once again to quantify deterioration of its perfor-mance (e.g., wear of cutting bit, mechanical components)over time and footage of rocks drilled.

Figure 28 shows an example of drilling telemetry acquiredduring a 1 m drill test in Indiana limestone. Plotted telemetryincluded ROP, WOB, drill bit temperature, specific energy,and total power. The auger rpm was set to 100 and percus-sive frequency to 810 BPM. Hence, blows per revolutionwere 8.1 (i.e., the drill hammered 8.1 times every revolution).The test was done at 3 torr vacuum and ambient temperatureof 26�C. The drill bit diameter was 25.4 mm.

It can be noted that ROP and power gradually increasedwith depth while WOB in fact dropped over the same depthrange. This is counterintuitive, since lower WOB should leadto lower ROP and in turn lower drilling power. The increasein ROP with a drop in WOB can be attributed to cork-screwing effect. The rubbing of auger flutes on the boreholewall combined with cuttings drag as they are being movedup the flutes can pull the auger into the hole in a similar waya corkscrew advances into a cork just by rotating it. Thisphenomenon was observed by Apollo astronauts drillinginto lunar regolith on the Moon. They observed that it wasunnecessary to push on the drill to maintain penetration rate.The drill auger would pull itself deeper into the hole. OnApollo 15, the penetration rate was too high, and cuttingswere not cleared out of the hole sufficiently fast. This led toauger choking and precious time lost while the astronautsstruggled to free up the auger. In 2010, we performed drillingtests on the slopes of the Mauna Kea volcano in Hawaii intephra (lunar analog soil). We used a 700 W handheld rotary-percussive drill and a set of 25.4 mm diameter augers thatcould be screwed together to reach a 5 m depth. During thedrilling tests we observed the same corkscrewing tendencyof the auger as the Apollo astronauts did. In fact, initially thedrill did choke, and it took a lot of effort to pull it out of thehole. Subsequently, the drill had to be lifted up to reducepenetration rate and in turn avoid future auger choking.


Figure 28 also shows that values for the drilling power,WOB, and energy are in the range of what a spacecraft couldprovide on the surface of Mars (or the Moon). The powerwas less than 70 W, WOB was less than 90 N, and energy todrill to 1 m depth was less than 100 Whr. The average ROPwas 1 m/hr. Hence, the Icebreaker drill maintained the 1-1-100-100 level in these conditions and formation type.

It should also be noted that the drill bit temperaturenever exceeded 50�C or approximately 25�C above theformation temperature. The bit temperature initially in-creased at a depth of 114 mm due to higher specific energyand then dropped as soon as the specific energy also drop-ped. The specific energy (as opposed to drilling power) givesmuch closer correlation to drill bit temperature (Zacny et al.,2008).

Figures 29, 30, and 31 show ROP, power, and drillingenergy obtained from drilling into various formations at 760torr (Earth atmospheric pressure) and 6.4 torr (average Marspressure at the Mars Phoenix landing site). The x axis showsformation types, with increasing formation strength. Data inFig. 31 are combined from data in Figs. 29 and 30. It shouldbe noted that Phoenix pressure of 6.4 torr is above the triplepoint of water pressure of 4.5 torr, and in turn, it is feasiblefor liquid water to form.

Drilling energy, calculated as power/ROP, gives an ap-proximate value of energy required to drill to 1 m depth. Inall tests, remaining drilling parameters were kept the same.These include rotational speed of 100 rpm, percussive energyof 2.6 J/blow, and percussive frequency of 810 BPM. Itshould be noted that only formation temperature waschanged, while the ambient temperature remained at ap-proximately 25�C. The drill bit diameter was 28.6 mm.

Results indicate that there is a noticeable difference inROP, power, and energy between the two pressure regimes.In particular, the power is higher and ROP lower (leading tohigher energy) at a pressure of 6.4 torr. It is believed that thisdifference is due to the cuttings removal rather than drilling.That is, we did not observe that the strength of rocks, ice, oricy regolith increased with a drop in pressure. However, inour previous studies we observed that the cuttings flow isless effective at 6.4 torr and even less effective when theatmosphere of 6.4 torr has been purged with dry nitrogen todecrease the humidity level.

In general, the drilling power never exceeded 100 W,while ROP ranged from 40 cm/hr for Indiana limestoneto *200 cm/hr for ice at - 20�C. The energy ranged from 20Whr/m to 200 Whr/m. The WOB was software-limited to

FIG. 28. Icebreaker drilling telemetry in Indiana limestone. Color images available online at www.liebertonline.com/ast

FIG. 29. Icebreaker ROP as a function of formation typeand atmospheric pressure.

FIG. 30. Icebreaker power as a function of formation typeand atmospheric pressure.

1190 ZACNY ET AL.

below 100 N. Hence, this set of tests demonstrated drilling atclose to 1-1-100-100 level.

7.3. Testing in Mars analog sites of Antarctica

The high-elevation areas of the Dry Valleys of Antarcticaare an ideal Mars analog site for many reasons. The tem-peratures are always below freezing; hence, ice-cementedground is formed through vapor diffusion from the air to theground, as is the case at the Mars northern polar regions(Lacelle et al., 2013). The top layer, often 20 cm or more deep,is desiccated sand and silt; the ice-cemented ground orsometimes relatively pure ice lies below this dry frozenregolith layer (Fig. 32). This layering is similar to what wasfound at the Mars Phoenix landing site and is expectedelsewhere on Mars (Smith et al., 2009).

The Icebreaker drill was deployed at two sites in Uni-versity Valley (77�51.891¢S, 160�48.029¢E, elevation 1700 m),as shown in Fig. 32. The first site included ice-cementedground, while the second site included massive ice. The drillwas tested three times to a depth of 1 m in ice-cementedground and once to a depth of 2.5 m in massive ice. Theice-cemented ground temperature at the depth of 1 m

was - 19�C, while the temperature of massive ice at thedepth of 2.5 m was - 25�C; the mean annual temperature atthe site is - 25�C (Marinova et al., 2012). In each case, the drillwas placed on a *10� slope to demonstrate off-vertical op-eration. The drill followed the bite sampling routine.

Figure 33 shows drilling telemetry in ice-cemented groundwhen the bite approach was implemented; that is, the drillwas pulled out of the hole every 10 cm to deposit the sampleand clean the auger flutes. The depth of 1 m in ice-cementedground was reached in 50 min, and the average penetrationrate was 1.2 m/hr. The penetration rate is relatively constantwith depth. The occasional dips are due to the drill beingpulled out and lowered back into the hole, and the remnantsof data processing. The maximum power was in the range of100 W, of which 60 W was attributed to the percussive sys-tem, while 40 W to the auger (i.e., drilling and cuttings re-moval). The WOB was software-limited to less than 100 N.The percussive system was triggered once WOB would ex-ceed 40 N. The total drill energy to reach 1 m and samplenine, *100 mm bites (the first bite was never sampled be-cause it was dry regolith) was approximately 100 Whr.Hence, Icebreaker demonstrated drilling at the 1-1-100-100level. Note that the percussive power and the total powerincreased at *100 mm depth. This was the depth to ice-ce-mented ground. Between the surface and 100 mm depth, theregolith was dry and easy to penetrate. Below the 10 cmdepth, the drill encountered icy regolith, which was as hardas *40 MPa rock (Li et al., 2001; Ma and Chang, 2002).Hence, from 100 mm depth onward the percussor was au-tomatically turned on to maintain steady penetration rate atlow WOB. It should be noted that this telemetry couldtherefore be used to determine depth to ice-cemented groundor help with selective sampling (e.g., sampling only dry orice-cemented ground).

The bit temperature graph shows troughs every 100 mmor so. These indicate the depths where the auger was pulledout of the hole to deposit a sample. It can be seen that thereare only nine such events, which is because the first 100 mmsample was not retrieved—it was dry regolith. In all cases,

FIG. 31. Icebreaker drilling energy as a function of forma-tion type and atmospheric pressure.

FIG. 32. Drill sites at the Ant-arctic Dry Valleys (UniversityValley): Site #1 was ice-cementedground, while Site #2 was mas-sive ice. Color images availableonline at www.liebertonline.com/ast


the bit temperature was initially low, and with depth it keptincreasing. However, the bit never reached temperaturesabove freezing, indicating that the temperature control rou-tine worked satisfactorily (the routine changes drilling pa-rameters to maintain drill bit temperature at or below 0�C).The frequent sample and auger retrieval out of the groundallowed both the auger and the formation to cool down.

Figure 34 shows the Icebreaker drill telemetry from drill-ing a hole approximately 50 cm from the hole reported inFig. 33. The main difference is that, in the current test, thedrill continuously drilled to 800 mm depth, while previouslyit followed the bite sampling routine. As before, it can beseen that the depth to ground ice is approximately 223 mm;up to this depth, the power was very low, of the order of 20

W, while penetration rate was in excess of 200 cm/hr. At thedepth of approximately 223 mm, the drill encountered ice-cemented ground, and in turn the penetration rate dropped.As a result of drop in ROP, the WOB increased to try tomaintain penetration rate. Once it crossed the 40 N threshold,the percussive system was engaged to further increase ROP.Engaging the percussive system increased total power by 70W. Additional power increase was attributed to higher augertorque. From the moment the drill encountered ice-cementedground, there was a steady decrease in penetration rate from100 cm/hr at 223 mm to 10 cm/hr at the 800 mm depth. Thiswas mirrored by the increase in auger power from 100 to 180W. The drop in ROP and increase in auger power is attrib-uted to poor cuttings removal out of the hole. Since the hole

FIG. 33. Icebreaker drill telemetry during tests in the ice-cemented ground in the Antarctic Dry Valleys—a Mars analog site.The drill used the bite drilling approach. The test data shows auger power and total power (percussive power = total power -auger power), ROP, WOB, and temperature of the drill bit. Color images available online at www.liebertonline.com/ast

FIG. 34. Icebreaker drill telemetry during tests in the ice-cemented ground in the Antarctic Dry Valleys—a Mars analog site.The drill continuously drilled to 800 mm depth. The test data shows auger power and total power (percussive power = totalpower - auger power), ROP, WOB, and temperature of the drill bit. Color images available online at www.liebertonline.com/ast

1192 ZACNY ET AL.

was being drilled continuously, cuttings occupied the entirelength of the auger within the hole. As the hole got deeper, thedistance cuttings needed to travel to the surface got progres-sively longer, and in turn the auger frictional drag increased. Itis worth pointing out that during the duration of the test, thetemperature subroutine was successful in keeping the bittemperature at or below freezing. The two drill tests, as shownin Figs. 33 and 34, demonstrated that when using a bite rou-tine, drilling power is lower, penetration rate is faster, and bittemperature is lower than when drilling a continuous hole.

Figure 35 shows the drilling telemetry in massive ice to adepth of 2.5 m. During this test, initially a bite routine wasimplemented to a depth of 1 m followed by continuous dril-ling. Every zigzag on the temperature curve corresponds to thedrill pull-out event to empty the cuttings. Drilling telemetry to1 m depth with the bite routine was quite uniform. However,past 1 m depth, the telemetry mirrors that in Fig. 34; that is, thepower would increase while penetration rate decreased.

In this test, the auger was quite effective to 2 m depth inclearing the cuttings, but just below 2 m the auger powerstarted to increase almost exponentially. Note that the in-crease in power was not due to drilling into a harder sub-

strate, since the bit temperature remained low at - 20�Cduring the entire drilling process. If the formation becameharder, this would be reflected in higher drilling power andin turn bit temperature, since drilling power is associatedwith formation destruction and in turn heat.

During the course of the test, the WOB was < 90 N, theaverage ROP was 1 m/hr (it took *2.5 hours to reachthe 2.5 m depth), and the average power was 100 W. Hence,the total energy was *250 Whr or 100 Whr/m. This testdemonstrated that it is feasible to drill to 2.5 m in massive icewith relatively low power, but it also illustrated the impor-tance of frequently cleaning up cuttings to keep the power low.

In all three tests, the bit temperature was always at orbelow freezing, and this helped keep cuttings cold. In addi-tion, the air temperature in the Antarctic Dry Valleys wasalways below freezing; hence the entire drill system, in-cluding the auger, was cold. The drill string therefore actedas a heat pipe and helped to dissipate heat formed during thedrilling operation. The cold Antarctic air was also very dry.The low temperature and low humidity helped keep drillcuttings frozen. Cuttings were also shielded by a drill plat-form from direct solar radiation. Figure 36 (left) shows

FIG. 35. Icebreaker tests in massive ice in the Antarctic Dry Valleys—a Mars analog site. The test data shows auger powerand total power (percussive power = total power - auger power), ROP, WOB, and temperature of the drill bit. Color imagesavailable online at www.liebertonline.com/ast

FIG. 36. Left: ice-cemented groundand ice cuttings acquired by the Ice-breaker drill in the Antarctic did notstick. Right: cuttings from drilling inice-cemented ground. Center: icecuttings. Right: Ice cuttings con-tained single ice crystals 5 mm large.Color images available online atwww.liebertonline.com/ast


cuttings acquired by drilling into the ice-cemented ground. Itcan be seen that these cuttings do not stick but rather fall as ifthey were dry granular media (particles). The same wasobserved with ice cuttings, which did not stick and alsopreserved large ice chunks as shown in Fig. 36 (right). Thisobservation is of paramount importance to the Icebreakermission, as it shows that it is possible for icy cuttings not tostick. The major factor is keeping the cuttings temperaturelow during the drilling operation by implementing suitabledrilling protocols and by shielding them from direct sunlightonce on the surface. From an operational standpoint, it mightbe prudent to land with a drill facing away from the morningSun, thermally isolate the sample collection mechanismsfrom the instrument deck, and perform drilling and sampletransfer operations early in the morning (or at night, if en-ergy budget allows).

8. Next-Generation Icebreaker: The Icebreaker2 Drill

The Icebreaker2 drill represents the third generation ofHoneybee Robotics rotary-percussive 1 m class drills (seeTable 3). The first generation, the CRUX drill, was a 2 kWsystem that was incapable of operation in a vacuum (Zacnyet al., 2010e). However, it confirmed the benefit of percussivesystems for planetary applications and in particular lowWOB requirement to achieve nominal penetration in a rangeof soils, icy soils, and rocks (Paulsen et al., 2010). The secondgeneration, the Icebreaker drill, has been thoroughly testedin vacuum chambers and in the Arctic and Antarctica, anddemonstrated drilling at 1-1-100-100 level; that is, 1 m in 1hour with 100 W and 100 N WOB (Paulsen et al., 2011). Toreduce the development cost, the drill was not mass-opti-mized and weighed over 40 kg. The Icebreaker2 drill shownin Fig. 37, which is almost as powerful as Icebreaker (*400W vs. 300 W), has a slightly lower percussive energy (2 J/blow vs. 2.5 J/blow) but weighs only 10 kg (Zacny et al.,2013c). The drill has been mass-optimized to reflect a‘‘flightlike’’ design. The drill has already been deployed inGreenland and Devon Island (Canadian High Arctic) in thesummer of 2013. The drill will also be extensively tested inthe Mars environmental chamber. Figure 37 also shows themock-up lander, which is slightly larger than the size of the2007 Mars Phoenix and 2016 Mars InSight mission landers.

It should be noted that Icebreaker2 has a twin: the LITAdrill. The LITA drill has been integrated with the CarnegieMellon University Zoe rover and was deployed in Atacama,

Chile, in the summer of 2013 as part of the NASA Astro-biology Science and Technology for Exploring Planets–funded project LITA: Life in the Atacama (Cabrol et al., 2013).The drill autonomously drilled to a predefined depth anddelivered sample to a sample manipulation system (a car-ousel with 20 cups) for analysis by the Mars MicrobeamRaman Spectrometer (MMRS) and a UV-stimulated fluo-rescence imager called BUF (Wang et al., 2013).

FIG. 37. The second- and third-generation rotary-percussivedrills: Icebreaker (left) and Icebreaker2 (right). Icebreaker2 hassimilar capabilities as Icebreaker but weighs only 10 kg. Colorimages available online at www.liebertonline.com/ast

Table 3. Comparison between the 1st (CRUX), the 2nd

(Icebreaker) and the 3rd (Icebreaker2)Generation Planetary Drill Systems

CRUX 1st generation Icebreaker 2nd generation Icebreaker2 3rd generation

Technology Readiness Level 4 5 5/6Penetration depth (cm) 100 100 100Mass (kg) 80 40 10Overall stowed length (cm) 190 146Nominal drill diameter (mm) 40 25 13Rotary power (W) 2000 220 150Percussive power (W) 160 175 150Rotational speed (rpm) 250 130 200Percussive energy ( J/blow) 2 2.5 2Percussive frequency (BPM) 1700 1300 2600Max WOB/pull-out force (N) 1000 1000 500

1194 ZACNY ET AL.

Neither Icebreaker nor Icebreaker2 were subjected to vi-bration tests associated with the launch or EDL at Mars. Hence,they are not at TRL 6. However, both drills have enduredvibrations associated with various modes of transportation:airplanes, trucks, ground vehicles on rough terrains (in theArctic), ATV, hand carried, and dropped from a height. Bothdrills were shipped across the continents: Icebreaker went toAntarctica and the Arctic; Icebreaker2 and its sister drill, theLITA drill, went to the Atacama, the Arctic, and Greenland.

9. Conclusions

This paper presents the development and testing of theIcebreaker drill, a critical element of the Mars Icebreakermission (McKay et al., 2013). Its goal is to acquire asepticsamples and deliver them to life-detection instruments on thedeck of the lander. The Icebreaker drill is a 1 m class, arm-deployable drill system. The drill has been under develop-ment since 2005 and currently is at TRL 5/6.

The drill employs a rotary-percussive approach to drilling,which reduces WOB and energy requirements. The drillrated power is 400 W; however, the average drilling power isin the range of 100 W. The drill initially weighed 40 kg;however, the mass has been reduced to 10 kg through use oflighter materials and several design modifications.

The Icebreaker drill has been tested in the Mars analogsites of the High Arctic and Antarctic Dry Valleys, as well asin a vacuum chamber. In all tests, the drill penetrated into avariety of formations with limited power and WOB. As a usefulmnemonic, the drill has been designed to maintain 1-1-100-100performance level—that is, 1 m in 1 hour with 100 W and 100 NWOB. Thus far, all drilling tests maintain this or lower level,with a few exceptions where power reached 180 W.

An integral part of the Icebreaker drill is its drilling soft-ware. The software has been developed via two parallel ef-forts. NASA Ames has been focusing on implementing ahigh-level control and diagnostics, while Honeybee Roboticshas been focusing on rule-based algorithms.

The drill uses a bite sampling approach whereby the drillis retracted every 10 cm or so to deliver the sample to asample transfer system. Frequently retracting the drill out ofthe hole has many advantages. It allows the subsurface tocool down, it reduces the volume of cuttings that need to beaugered out of the hole, and stratigraphy could be main-tained to within 10 cm. In addition, since analysis of eachsample will no doubt take several sols, it is prudent to keepthe drill above the hole during this time rather than riskfreezing it in the hole.

The drill includes three independent sample deliverysystems. The pneumatic sample delivery is a point-to-pointsystem. As soon as the sample is augered out of the hole, it iscaptured inside a small chamber. A puff of gas clears thesampling chamber and delivers the precious cargo directlyinto a target instrument. The 3-DOF arm that deploys thedrill can also be used in conjunction with a drill as a sampledelivery tool. In particular, once the drill is above the surfaceand sample captured around the bit, the arm can position thetip of the bit above the instrument inlet port. The drill canpercuss and move the auger with the cuttings forward todispense the sample directly into a cup. The third approachuses an independent 5-DOF arm with a scoop at the end. Thescoop is positioned close to the sample drop-off point and,

upon capturing the sample, delivers it to a target instrumentin the same way the Mars Phoenix arm delivered a sampleinside a scoop.

Since the Icebreaker mission is a life-detection missiontargeting Mars Special Regions, the drill has been designedin such a way as not to preclude sterilization by DHMR.

The Icebreaker drill could also be used as a science instru-ment. Its bit-embedded thermocouple could provide thermalgradient data, which combined with estimated thermal con-ductivity could be used to estimate martian heat flow. Thedrilling telemetry could be used to estimate formation UCS,and UCS combined with subsurface temperature could beused estimate the fraction of ice present in the subsurface.

Acknowledgments

The research reported in this manuscript was funded bythe National Aeronautics and Space Administration (NASA)Astrobiology Science and Technology for Exploring Planets(ASTEP) and Astrobiology Science and Technology Instru-ment Development (ASTID) programs: NASA ContractsNNA09DA97C ‘‘Mars Umbrella Drilling Scout (MUDS)’’,and NNX11AJ87G ‘‘Robotic Investigation of Subsurface Lifein the Atacama’’. Testing in Antarctica was supported by theUS Antarctic Program as part of the National ScienceFoundation Office of Polar Programs.

Abbreviations

ALSD: Apollo Lunar Surface Drill.BPM: blows per minute.CheMin: Chemistry and Mineralogy.CHIMRA: Collection and Handling for In situ Martian

Rock Analysis.DHMR: dry heat microbial reduction.DOF: degree of freedom.EDL: entry, descent, and landing.GCMS: gas chromatograph–mass spectrometer.ISAD: Icy Soil Acquisition Device.MER: Mars Exploration Rovers.MSL: Mars Science Laboratory.PADS: Powder Acquisition Drill System.RAT: Rock Abrasion Tool.ROP: rate of penetration.SAM: Sample Analysis at Mars.SA/SPaH: Sample Acquisition, Sample Processing and

Handling.SSA: Surface Sampler Assembly.TRL: Technology Readiness Level.UCS: unconfined compressive strength.WOB: weight on bit.

References

Arvidson, R.E., Bonitz, R.G., Robinson, M.L., Carsten, J.L.,Volpe, R.A., Trebi-Ollennu, A., Mellon, M.T., Chu, P.C., Davis,K.R., Wilson, J.J., Shaw, A.S., Greenberger, R.N., Siebach, K.L.,Stein, T.C., Cull, S.C., Goetz, W., Morris, R.V., Ming, D.W.,Keller, H.U., Lemmon, M.T., Sizemore, H.G., and Mehta, M.(2009) Results from the Mars Phoenix lander Robotic Armexperiment. J Geophys Res 114, doi:10.1029/2009JE003408.

Bar-Cohen, Y. and Zacny, K., editors. (2009) Drilling in ExtremeEnvironments: Penetration and Sampling on Earth and OtherPlanets, Wiley-VCH, Weinheim.


Bar-Cohen, Y., Badescu, M., Bao, X., Sherrit, S., Zacny, K., Sa-dick, S., and Ji, J. (2010) Deep drilling and sampling viacompact low-mass rotary-hammer auto-gopher. In Earth andSpace 2010: Engineering, Science, Construction, and Operations inChallenging Environments, edited by G. Song and R.B. Malla,American Society of Civil Engineers, Reston, VA, doi:10.1061/41096(366)94.

Buckley, D. (1971) Friction, wear and lubrication in vacuum.NASA SP-277, Library of Congress Catalog Card Number 72-174581, NASA, Washington, DC.

Byrne, S., Dundas, C.M., Kennedy, M.R., Mellon, M.T., McEwen,A.S., Cull, S.C., Daubar, I.J., Shean, D.E., Seelos, K.D.,Murchie, S.L., Cantor, B.A., Arvidson, R.E., Edgett, K.S., Re-ufer, A., Thomas, N., Harrison, T.N., Posiolova, L.V., andSeelos, F.P. (2009) Distribution of mid-latitude ground ice onMars from new impact craters. Science 325:1674–1676.

Cabrol, N.A., Wettergreen, D.S., and the LITA Project ScienceTeam. (2013) Life in the Atacama: science and technologypathways to the robotic search for life on Mars [abstract 1190].In 44th Lunar and Planetary Science Conference, Lunar and Plan-etary Institute, Houston.

Chevrier, V.F., Hanley, J., and Altheide, T.S. (2009) Stability ofperchlorate hydrates and their liquid solutions at the Phoenixlanding site, Mars. Geophys Res Lett 36, doi:10.1029/2009GL037497.

Chu, P., Wilson, J., Davis, K., Shiraishi, L., and Burke, K. (2008)Icy Regolith Acquisition Device for the 2007 Phoenix MarsLander. In Proceedings of the 39th Aerospace Mechanisms Sym-posium, NASA/CP-2010-216272, NASA Marshall Space FlightCenter, Huntsville, AL.

Conley, C. (2011) http://planetaryprotection.nasa.gov/. Ac-cessed 6 June 2011.

Dave, A., Thompson, S.J., McKay, C.P., Stoker, C.R., Zacny, K.,Paulsen, G., Mellerowicz, B., Glass, B., Willson, D., and Bo-naccorsi, R. (2012) The sample handling system for the MarsIcebreaker Life mission: from dirt to data [abstract 4186]. InConcepts and Approaches for Mars Exploration, Lunar and Pla-netary Institute, Houston, LPI Contribution No. 1679.

Dave, A., Thompson, S.J., McKay, C.P., Stoker, C.R., Zacny, K.,Paulsen, G., Mellerowicz, B., Glass, B.J., Willson, D., Bo-naccorsi, R., and Rask, J. (2013) The sample handling systemfor the Mars Icebreaker Life mission: from dirt to data. As-trobiology 13:354–369.

Elphic, R.C., Chu, P., Hahn, S., James, M.R., Lawrence, D.J.,Prettyman, T.H., Johnson, J.B., and Podgorney, R.K. (2008)Surface and downhole prospecting tools for planetary explo-ration: tests of neutron and gamma ray probes. Astrobiology8:639–652.

European Space Agency. (2012) The ExoMars drill unit. Avail-able online at http://exploration.esa.int/science-e/www/object/index.cfm?fobjectid=43611.

Fletcher, N.H. (1968) Surface structure of water and ice. II. Arevised model. Philosophical Magazine 18:1287–1300.

Fletcher, N.H. (1970) The Chemical Physics of Ice, CambridgeMonographs on Physics, Cambridge University Press, Cam-bridge, UK.

Gilichinsky, D., Rivkina, E., Shcherbakova, V., Laurinavichuis,K., and Tiedje, J. (2003) Supercooled water brines withinpermafrost—an unknown ecological niche for microorgan-isms: a model for astrobiology. Astrobiology 3:331–341.

Glass, B., Cannon, H., Hanagud, S., and Frank, J. (2005) Drillingautomation for subsurface planetary exploration. In Proceed-ings of the 8th International Symposium on Artificial Intelligence,Robotics and Automation in Space (iSAIRAS), Munich Germany,

September 2005, ESA SP-603, European Space Agency,Noordwijk, the Netherlands.

Glass, B., Cannon, H., Branson, M., Hanagud, S., and Paulsen, G.(2008) DAME: planetary prototype drilling automation. As-trobiology 8:653–664.

Glass, B.J., McKay, C., Thompson, S., and Zacny, K. (2011) Au-tomated Mars drilling for ‘‘Icebreaker.’’ In Aerospace Con-ference, 2011 IEEE, Institute of Electrical and ElectronicsEngineers (IEEE), Piscataway, NJ, pp 1–7.

Gorevan, S.P., Myrick, T., Davis, K., Chau, J.J., Bartlett, P., Mu-kherjee, S., Anderson, R., Squyres, S.W., Arvidson, R.E.,Madsen, M.B., Bertelsen, P., Goetz, W., Binau, C.S., andRichter, L. (2003) Rock Abrasion Tool: Mars Exploration Rovermission. J Geophys Res 108, doi:10.1029/2003JE002061.

Hecht, M.H., Kounaves, S.P., Quinn, R.C., West, S.J., Young,S.M., Ming, D.W., Catling, D.C., Clark, B.C., Boynton, W.V.,Hoffman, J., Deflores, L.P., Gospodinova, K., Kapit, J., andSmith, P.H. (2009) Detection of perchlorate and the solublechemistry of martian regolith at the Phoenix lander site. Sci-ence 325:64–67.

Heins, R. and Friz, T. (1967) The effect of low temperature onsome physical properties of rock. In Drilling and Rock Me-chanics Conference, January 25–26, 1967, Austin, TX, Society ofPetroleum Engineers, Richardson, TX, document ID 1714-MS.

Honeybee Robotics. (2012) Ice Breaker Drill in Action. Availableonline at http://www.youtube.com/watch?v=fTNPokiXa0E.

Jandura, L. (2010) Mars Science Laboratory Sample Acquisition,Sample Processing and Handling: subsystem design and testchallenges. In Proceedings of the 40th Aerospace MechanismsSymposium, NASA/CP-2010-216272, NASA Kennedy SpaceCenter, FL.

Kumar, A. (1968) The effect of stress rate and temperature on thestrength of basalt and granite. Geophysics 33:501–510.

Lacelle, D., Davila, A.F., Fisher, D., Pollard, W.H., DeWitt, R.,Heldmann, J., Marinova, M.M., and McKay, C.P. (2013) Excessground ice of condensation-diffusion origin in UniversityValley, Dry Valleys of Antarctica: evidence from isotopegeochemistry and numerical modeling. Geochim CosmochimActa 120:280–297.

Li, H., Yang, H., Chang, C., and Sun, X. (2001) Experimentalinvestigation on compressive strength of frozen regolith ver-sus strain rate. Journal of Cold Regions Engineering 15,doi:10.1061/(ASCE)0887-381X(2001)15:2(125).

Ma, W. and Chang, X. (2002) Analyses of strength and defor-mation of an artificially frozen soil wall in underground en-gineering. Cold Regions Science and Technology 34:11–17.

Mankins, J.C. (1995) Technology Readiness Levels: a white pa-per. NASA Office of Space Access and Technology, AdvancedConcepts Office, NASA Marshall Space Flight Center,Huntsville, AL.

Marinova, M., McKay, C.P., Heldmann, J.L., Davila, A.F., An-dersen, D.T., Jackson, W.A., Lacelle, D., Paulsen, G., Pollard,W.H., and Zacny, K. (2012) Distribution of the depth to ice-cemented soils in the high-elevation Quartermain Mountains,Dry Valleys of Antarctica. Antarct Sci 25:575–582.

McKay, C., Stoker, C.R., Glass, B.J., Dave, A.I., Davila, A.F.,Heldmann, J.L., Marinova, M.M., Fairen, A.G., Quinn, R.C.,Zacny, K.A., Paulsen, G., Smith, P.H., Parro, V., Andersen,D.T., Hecht, M.H., Lacelle, D., and Pollard, W.H. (2013) TheIcebreaker Life mission to Mars: a search for biomolecularevidence for life. Astrobiology 13:334–353.

Mellon, M.T., Arvidson, R.E., Sizemore, H.G., Searls, M.L.,Blaney, D.L., Cull, S., Hecht, M.H., Heet, T.L., Keller, H.U.,Lemmon, M.T., Markiewicz, W.J., Ming, D.W., Morris, R.V.,

1196 ZACNY ET AL.

Pike, W.T., and Zent, A.P. (2009) Ground ice at the Phoenixlanding site: stability state and origin. J Geophys Res 114,doi:10.1029/2009JE003417.

Mellor, M. (1971) Strength and deformability of rocks at lowtemperatures. CRREL Research Report 294, Cold RegionsResearch and Engineering Lab, Hanover, NH.

Mellor, M. (1972) Normalization of specific energy. InternationalJournal of Rock Mechanics and Mining Sciences & GeomechanicsAbstracts 9:661–663.

Mitchell, J.K. (1992) Fundamentals of Soil Behavior, 2nd ed., JohnWiley, Hoboken, NJ.

Moreschini, P., Zacny, K., and Rickman, D. (2011) Downhole el-emental analysis with LIBS [abstract #P43B-1667]. In AGU FallMeeting 2011, American Geophysical Union, Washington, DC.

Niebuhr, D. (2007) Friction and wear behavior of engineeringmaterials in a simulated martian (CO2) environment, a pre-liminary study. Wear 263:88–92.

Okon, A. (2010) Mars Science Laboratory drill. In Proceedings ofthe 40th Aerospace Mechanisms Symposium, NASA/CP-2010-216272, NASA Kennedy Space Center, FL.

Parro, V., Fernandez-Calvo, P., Rodrıguez Manfredi, J.A., Mor-eno-Paz, M., Rivas, L.A., Garcıa-Villadangos, M., Bonaccorsi,R., Gonzalez-Pastor, J.E., Prieto-Ballesteros, O., Schuerger,A.C., Davidson, M., Gomez-Elvira, J., and Stoker, C.R. (2008)SOLID2: an antibody array–based life-detector instrument in aMars drilling simulation experiment (MARTE). Astrobiology8:987–999.

Paulsen, G., Zacny, K., Cannon, H., Glass, B., Chu, P., Mumm,E., Davis, K., Frader-Thompson, S., Petrich, K., Bartlett, P., andGlaser, D. (2006) Robotic drill systems for planetary explora-tion. In Space 2006, American Institute of Aeronautics andAstronautics, Reston, VA, doi:10.2514/6.2006–7512.

Paulsen, G., Zacny, K., McKay, C., Shiraishi, L., Kriechbaum, K.,Glass, B., Szczesiak, M., Santoro, C., Craft, J., Malla, R.B., andMaksymuk, M. (2010) Rotary-percussive deep drill for plan-etary applications. In Earth and Space 2010: Engineering, Sci-ence, Construction, and Operations in Challenging Environments,edited by G. Song and R.B. Malla, American Society of CivilEngineers, Reston, VA, doi:10.1061/41096(366)128.

Paulsen, G., Zacny, K., Szczesiak, M., Santoro, C., Mellerowicz,B., Craft, J., McKay, C., Glass, B., Davila, A., and Marinova, M.(2011) Testing of a 1 meter Mars Icebreaker drill in a 3.5 metervacuum chamber and in an Antarctic Mars analog site. InAIAA SPACE 2011 Conference & Exposition, American Instituteof Aeronautics and Astronautics, Reston, VA, doi:10.2514/6.2011–7236.

Paulsen, G., Szczesiak, M., Maksymuk, M., Santoro, C., andZacny, K. (2012) SONIC drilling for space exploration. In Earthand Space 2012: Engineering, Science, Construction, and Opera-tions in Challenging Environments, edited by K. Zacny, R.B.Malla, and W. Binienda, American Society of Civil Engineers,Reston, VA, doi:10.1061/9780784412190.059.

Peters, G.H., Abbey, W., Bearman, G.H., Mungas, G.S., Smith,J.A., Anderson, R.C., Douglas, S., and Beegle, L.W. (2008)Mojave Mars simulant—characterization of a new geologicMars analog. Icarus 197:470–479.

Pike, W., Staufer, U., Hecht, M.H., Goetz, W., Parrat, D., Sykulska-Lawrence, H., Vijendran, S., and Madsen, M.B. (2011) Quantifi-cation of the dry history of the martian soil inferred from in situmicroscopy. Geophys Res Lett 38, doi:10.1029/2011GL049896.

Reddish, D. and Yasar, E. (1996) A new portable rock strengthindex test based on specific energy of drilling. InternationalJournal of Rock Mechanics and Mining Sciences & GeomechanicsAbstracts 33:543–548.

Schulson, E.M. and Fortt, A.L. (2012) Friction of ice on ice. JGeophys Res 117, doi:10.1029/2012JB009219.

Seger, R. and Gillespie, V. (1974) The Viking Surface Sampler.NASA Langley Research Center, Hampton, VA. Available onlineat https://archive.org/detaile/nasa_techdoc_19740003574.

Shcherbakova, V., Rivkina, E., Laurinavichuis, K., Pecheritsina,S., and Gilichinsky, D. (2004) Physiological characteristics ofbacteria isolated from water brines within permafrost. Inter-national Journal of Astrobiology 3:37–43.

Smith, P.H., Tamppari, L.K., Arvidson, R.E., Bass, D., Blaney, D.,Boynton, W.V., Carswell, A., Catling, D.C., Clark, B.C., Duck,T., DeJong, E., Fisher, D., Goetz, W., Gunnlaugsson, H.P.,Hecht, M.H., Hipkin, V., Hoffman, J., Hviid, S.F., Keller, H.U.,Kounaves, S.P., Lange, C.F., Lemmon, M.T., Madsen, M.B.,Markiewicz, W.J., Marshall, J., McKay, C.P., Mellon, M.T.,Ming, D.W., Morris, R.V., Pike, W.T., Renno, N., Staufer, U.,Stoker, C., Taylor, P., Whiteway, J.A., and Zent, A.P. (2009)H2O at the Phoenix landing site. Science 325:58–61.

Stoker, C., Davilla, A., Davis, S., Glass, B., Gonzales, A., Held-mann, J., Karcz, J., Lemke, L., and Sanders, G. (2012) IceDragon: A mission to address science and human explorationobjectives on Mars, concepts and approaches for Mars ex-ploration, June 12–14, 2012; LPI, Houston, Texas.

Sunshine, D. (2010) Mars Science Laboratory CHIMRA: a devicefor processing powdered martian samples. In Proceedings ofthe 40th Aerospace Mechanisms Symposium, NASA/CP-2010-216272, NASA Kennedy Space Center, FL.

Szwarc, T., Aggarwal, A., Hubbard, S., Cantwell, B., and Zacny,K. (2012), A thermal model for analysis and control of drillingin icy formations on Mars. Planet Space Sci 73:214–220.

Teale, R. (1965) The concept of specific energy in rock drilling.International Journal of Rock Mechanics and Mining Sciences &Geomechanics Abstracts 2:57–73.

Tsytovich, N.A. (1975) The Mechanics of Frozen Ground, McGraw-Hill, New York.

Wang, A., Lambert, J., Sobron, P., and the Life in the Atacama ProjectTeam. (2013) An instrument suite for mineral ID and biomarkerseeking in Atacama [abstract 2586]. In 44th Lunar and PlanetaryScience Conference, Lunar and Planetary Institute, Houston.

Zacny, K. and Bar-Cohen, Y. (2010) Drilling and excavation forconstruction and in situ resource utilization. In Mars: Pro-spective Energy and Material Resources, edited by V. Badescu,Springer, Berlin, pp 431–459.

Zacny, K. and Cooper, G. (2005) Investigation of the perfor-mance of a coring bit in frozen mud under martian conditionsof low temperature and pressure. J Geophys Res 110,doi:10.1029/2004JE002341.

Zacny, K. and Cooper, G. (2006) Considerations, constraints andstrategies for drilling on Mars. Planet Space Sci 54:345–356.

Zacny, K. and Cooper, G. (2007a) Coring basalt rock under sim-ulated martian atmospheric conditions. Mars 3, doi:10.1555/mars.2007.0001.

Zacny, K. and Cooper, G. (2007b) Friction of drill bits undermartian pressure. J Geophys Res 112, doi:10.1029/2005JE002538.

Zacny, K. and Cooper, G. (2007c) Methods for cuttings removalfrom holes drilled on Mars. Mars 3, doi:10.1555/mars.2007.00.

Zacny, K.A., Quayle, M.C., and Cooper, G.A. (2004) Laboratorydrilling under martian conditions yields unexpected results.J Geophys Res 109, doi:10.1029/2003JE002203.

Zacny, K., Quayle, M., and Cooper, G. (2005) Enhancing cuttingsremoval with gas blasts while drilling on Mars. J Geophys Res110, doi:10.1029/2004JE002340.

Zacny, K., Glaser, D., Bartlett, P., Davis, K., and Wilson, J. (2006)Test results of core drilling in simulated ice-bound lunar


regolith for the Subsurface Access System of the Constructionand Resource Utilization eXplorer (CRUX) project. In Earth &Space 2006: Engineering, Construction, and Operations in Challen-ging Environments, proceedings of the 10th Biennial InternationalConference on Engineering, Construction, and Operations inChallenging Environments and Second NASA/ARO/ASCEWorkshop on Granular Materials in Lunar and Martian Ex-ploration, League City, TX, doi:10.1061/40830(188)64.

Zacny, K., Glaser, D., Bartlett, P., Davis, K., and Gorevan, S.(2007) Drilling results in ice-bound simulated lunar regolith.AIP Conf Proc 880, doi:10.1063/1.2437524.

Zacny, K., Bar-Cohen, Y., Brennan, M., Briggs, G., Cooper, G.,Davis, K., Dolgin, B., Glaser, D., Glass, B., Gorevan, S.,Guerrero, J., McKay, C., Paulsen, G., Stanley, S., and Stoker, C.(2008), Drilling systems for extraterrestrial subsurface explo-ration. Astrobiology 8:665–706.

Zacny, K., Paulsen, G., Szczesiak, M., Santoro, C., and Craft, J.(2010a) CRUX and IceBreaker rotary percussive lunar drills. InSpace Resources Roundtable XI, June 8–10, 2010, Golden, CO.

Zacny, K., Wilson, J., Craft, J., Asnani, V., Oravec, H., Creager,C., Johnson, J., and Fong, T. (2010b) Robotic lunar geotech-nical tool. In Earth and Space 2010: Engineering, Science, Con-struction, and Operations in Challenging Environments, edited byG. Song and R.B. Malla, American Society of Civil Engineers,Reston, VA, doi:10.1061/41096(366)19.

Zacny, K., McKay, D., Beegle, L., Onstott, T., Mueller, R.,Mungas, G., Chu, P., and Craft, J. (2010c) Novel method ofregolith sample return from extraterrestrial body using a puffof gas [paper #1082]. In 2010 IEEE Aerospace Conference, In-stitute of Electrical and Electronics Engineers (IEEE), Piscat-away, NJ, doi:10.1109/AERO.2010.5446987.

Zacny, K., Craft, J., Hedlund, M., Chu, P., Galloway, G., andMueller, R. (2010d) Investigating the efficiency of pneumatictransfer of JSC-1a lunar regolith simulant in vacuum and lunargravity during parabolic flights. In AIAA Space 2010, Ameri-can Institute of Aeronautics and Astronautics, Reston, VA,doi:10.2514/6.2010–8702.

Zacny, K., Paulsen, G., and Glass, B. (2010e) Field testing ofplanetary drill in the Arctic. In AIAA Space 2010, AmericanInstitute of Aeronautics and Astronautics, Reston, VA,doi:10.2514/6.2010–8701.

Zacny, K., Paulsen, G., and Szczesiak, M. (2011) Challenges andmethods of drilling on the Moon and Mars [paper #1002]. In

2011 IEEE Aerospace Conference, Institute of Electrical andElectronics Engineers (IEEE), Piscataway, NJ, doi:10.1109/AERO.2011.5747261.

Zacny K., Paulsen, G., Mellerowicz, B., Craft, J., McKay, C.,Glass, B., Davila, A., Marinova, M., Dave, A., and Thompson,S. (2012a) The Icebreaker: Mars drill and sample deliverysystem [abstract 1153]. In 43rd Lunar and Planetary ScienceConference, Lunar and Planetary Institute, Houston.

Zacny, K., Hedlund, M., Herman, J., Craft, J., Smythe, B., andMcKay, C. (2012b) Sample crushing, sieving, metering, anddistribution system. In 2012 IEEE Aerospace Conference, In-stitute of Electrical and Electronics Engineers (IEEE), Piscat-away, NJ, doi:10.1109/AERO.2012.6187062.

Zacny, K., Paulsen, G., Szczesiak, M., Craft, J., Chu, P., McKay,C., Glass, B., Davila, A., Marinova, M., Pollard, W., andJackson, W. (2013a) LunarVader: development and testing of alunar drill in a vacuum chamber and in the lunar analog site ofthe Antarctica. J Aerosp Eng 26, doi:10.1061/(ASCE)AS.1943-5525.0000212.

Zacny, K., Paulsen, G., Mellerowicz, B., Bar-Cohen, Y., Beegle,L., Sherrit, S., Badescu, M., Corsetti, F., Craft, J., Ibarra, Y.,Bao, X., and Lee, H.J. (2013b) Wireline deep drill for ex-ploration of Mars, Europa, and Enceladus. In 2013IEEE Aerospace Conference, Institute of Electrical and Elec-tronics Engineers (IEEE), Piscataway, NJ, doi:10.1109/AERO.2013.6497189.

Zacny, K., Paulsen, G., Mellerowicz, B., Craft, J., Wettergreen,D., and Cabrol, N. (2013c) Life in the Atacama: the drill andsample delivery system [abstract 1332]. In 44th Lunar andPlanetary Science Conference, Lunar and Planetary Institute,Houston.

Address correspondence to:Kris Zacny

Honeybee Robotics398 W Washington Blvd., Suite 200

Pasadena, CA 91103

E-mail: [email protected]

Submitted 7 June 2013Accepted 15 October 2013

1198 ZACNY ET AL.

Date post:	18-Feb-2019
Category:	Documents
Upload:	dinhhuong
View:	224 times
Download:	0 times

Reaching 1m Deep on Mars: The Icebreaker Drill -...

Documents