 A Sample Delivery System for Planetary Missions, That Excavates, Filters and Dispenses Sample D. Willson1, , C.R. Stoker2, L. G. Lemke2, A. Duncan2 1KISS Institute for Practical Robotics/NASA Ames, Bld 245, NASA Ames M.S. 245-3, Moffett Field CA, david.willson@nasa.gov. 2NASA Ames Reserch Center Bld 245, NASA Ames M.S. 245-3, Moffett Field CA   Introduction: Capturing and transferring samples of Mars soil or drill cuttings to instruments is a deceptively difficult problem.  The 2008 Phoenix mission planned to dig into icy soils and deliver them to instruments for analysis using it's robot arm with Icy Sample acquisition device (ISAD) (Fig 1) [1]. The sample transfer was unexpectedly difficult. Samples clogged screens that were placed on instrument inlets to limit particle size, and stuck into the scoop even when it was tipped over to dump them out. Sample delivery was fouled by wind blowing away the sample and poor control over where the scoop dumped sample so that the deck was covered with soil by the end of the mission. Most in situ analysis instruments require limits on particle size and sample volume limits. A complex system for sample handling (CHIMRA) was developed for the MSL rover robot arm [2] that used vibrator devices and robot arm manipulation to deliver definable quantities of 150 micron size dry sample, but this system would not work well with icy samples. Thus a simpler system applicable for both dry and icy samples is needed and since 2014 we have been developing a Sample Delivery System (SDS) similar to the Phoenix mission ISAD type scoop that can excavate and deliver into instruments definable quantities of filtered icy sample.   Concept Development Trials: Numerous ideas were considered for the SDS which included food strainers, some with with rotating blades to push sample through a screen, and a rotating Trommel with chain and malice to breakup icy sample (Fig 2). These ideas either rely on gravity to push particles through the screen or incorporate metal plates passing over a screen, which works well with malleable food materials but becomes a pinch point with small pebbles and grains.  Finally a prototype SDS scoop incorporating a rotating wire brush (Fig 3) positioned over a grate located in the back of an excavator shaped scoop was developed and tested in Mars conditions [3]. The grate slot size was set to the proposed Icebreaker instrument maximum particle size requirement of 1mm. The brush pushes particles through the grate and prevents the grate from clogging. The prototype avoided the use of vibrator devices as they rely on gravity and do not actively push material through screens. The performance of vibrators is not easy to define in varied conditions and soil types and do not work well with sticky sample as shown with the Phoenix mission experience.  Prototype Test Results: Our tests involved a range of soils, drill cuttings and mixtures of ice & soil consisting of micro-meter to 30 mm sized particles that included Antarctica soils, crushed basalt, Mars simulant soil, and ice drill cuttings. The SDS operation was fully characterized with sample dispensing rate, particle size distribution both before and after sifting, and rotating brush motor current (Fig 4).  Our results showed sample volume was dispensed at a linear rate dependent on the portion of particles <1mm, and brush rotation speed.  However in icy and ice/soil mixtures, the sample became much cohesive and the wire brush hollowed out a 'cave' under the sample that bridged over the brush. Only a small volume of sample was dispensed.  Fig (1): The ISAD Scoop Mars Phoenix mission Fig (2): Concepts: (A) Trommel with chain/malice, (B) Food strainer. Rotating Brush Grate Fig (3): SDS Scoop and rotating wire brush positioned over grate A key lesson learned from these tests is that since the sample is small, it's "weight" has little influence on getting it to pass through funnels and filters. Soil cohesiveness forces dominate easily over gravity forces, thus devices are needed to minimize the need for gravity. Also small portions of particles dispensed were > 1mm, due to long or elliptical shaped particles. Thus the grate/filter size will need careful consideration to meet instrument particle size requirements. Overall the results showed that definable sample volumes dispensed could be estimated but an additional device was required to push cohesive or icy sample onto the brush.   Current SDS prototype: The new prototype SDS (Fig 5) is under development, which has: 1) an improved hopper/excavator scoop geometry to encourage sample flow to the brush with no pinch points, and 2) a plunger device designed actively push cohesive sample into the brush. The plunger minimizes the need for gravity to push cohesive sample to the brush, is positioned external to the hopper when retracted and has large clearances when in the hopper to mitigate mechanism jamming with sample pebbles. Sample is dispensed out scoop rear similar to a pepper grinder operation providing better precision delivery. This avoids the Phoenix problems of sample dumping that blocked instrument inlets and covered the deck with dirt. Prototype testing will in March 2016.  Summary: A sample delivery system consisting of a Phoenix mission ISAD type scoop that can excavate, filter and dispense icy and dry sample at a definable rate is being developed employing a rotary brush activily pushing sample though a grate. Results thus far have determined that definable sample volumes dispensed could be estimated, consideration of the grate size is required as a small portion of particles filtered are larger than the gate size, and a plunger device is needed to breakup cohesive sample and feed. Both mechanisms minimize the dependence for gravity for the sample filtering and dispensing process.  References: [1] Chu P., Wilson J., Davis K., Shiraishi L., Burke K. (2008) Proceedings of the 39th Aerospace Mechanisms Symposium, NASA Marshall Space Flight Center. [2] Sunshine D. (2010) Proceedings of the 40th Aerospace Mechanisms Symposium, NASA Kennedy Space Center. [3] Willson D., Lemke L.G., Stoker C.R, Dave A., McKay C.P. (2014) 46th Lunar and Planetary Science Conference, .        University Valley Soil  Sieved Mass Vs Time Sample Particle Size Distribution Filtered Sample Particle Size Distribution Fig (5): Current SDS development showing brush, grate, plunger and diepensing sample. Scoop rotates 70 degrees for dispensing into instruments Fig (4): Prototype SDS Results (top left): With dry Antarctic soil, (top right) dispensing rate,  (Bottom left) particle size distripution befor filtering and (bottom right) particle size distribution after filtering. Grate Sample  Dispensing Into instrument inlet Brush Plunger SDS 3D print in plastic - for initial testing. 
