1. Pneumatic Sampling in Extreme Terrain with the Axel Rover
Yifei Huang
Frank W. Wood SURF Fellow

2. Overview Motivation Pneumatic Sampling Design & Testing Conclusions
Concept, and feasibility
Design & Testing
Nozzle
Cyclone
Sample Container
Pressure Container
Instrument Deployment
Conclusions

3. Sampling in Extreme Terrain
Satellite images suggest liquid brine flow
Spectroscopy images – negative results for water
Difficulties in sampling
Newton Crater: degree slopes
MER:15 degree slopes
Curiosity: 30 degree slopes
Solution
Axel rover: vertical slopes
Figure: Sources:

4. The Axel rover Goal: Develop a sampling system on Axel DuAxel rover
Traversing cliffs
Instrument deploy
Goal: Develop a sampling system on Axel

5. What is pneumatic sampling?
1. Release pressurized air
Actuator opens and closes a cylinder of pressurized air
2. Air flows down the outer tube of the nozzle
3. Air enters inner tube, carrying soil with it
Nozzle is already embedded in dirt
Up is the path of least resistance
4. Soil and air flow up into sample container
Dirt in tube
Figure: Zacny et al. (2010)

6. Why Pneumatics?
Fewer moving components, low number of actuators, less risk for failure
Closed tubing: low instrument contamination
Energy efficient
A small amount of air can lift a large amount of dirt
1 g of gas lifted 5000g of soil [Zacny and Bar-Cohen, 2009]
Easier soil transportation

7. Design: Nozzle
Round #1
Nozzle #2
Soil Level
Nozzle #3
Nozzle #1

8. Design: Nozzle Nozzles built on the 3D printer (ABS plastic)
Tests with loose sand (400um size)
25psi air was released for 2 sec

9. Design: Nozzle
Round #2
Sand:
Dirt:
Nozzle #4
Nozzle #5

10. Design: Cyclone Separator
Used to separate air and soil
Dusty air will enter tangential to cyclone
Larger particles have too much inertia
Hit the side of cyclone and fall down
Smaller particles remain in the cyclone
Pushed up into the Vortex Finder by pressure gradient
Vortex Finder
Cylindrical portion
Conical portion
Small Particle
Large Particle
Design by Honeybee Robotics
Figure: DB Ingham and L Ma, “Predicting the performance of air cyclones”

11. Design: Sample Container
Objective: Minimize actuation with springs
Concept:
Design:
Cyclone
Sample Container
Spring

12. Design: Instrument Deployment
Second 4-bar linkage attached to original 4-bar
Motion of 2 4-bars are coupled
Advantages: No actuator on deployed plate
Nozzle is attached here

13. Benchtop test stands
Instrument deploy
Sample Caching

14. Design: Pressure Container

15. Benchtop Test Tests with loose sand (400um size)
25psi air was released for 2 sec

16. Contamination In sand In dirt
Weighed cyclone, tubing, and nozzle before and after tests
Negligible mass: ~0.2% of lifted mass remained in cyclone/tubing/nozzle
In dirt
Soil is stuck inside nozzle and cyclone
Cyclone: % of lifted mass
Nozzle: % of lifted mass

17. Effects of Pressure Tests with loose sand (400um size)
Air from wall was released for 2 sec

18. Conclusions Pneumatics is feasible Improvements needed:
Successfully acquired 2g of soil
Improvements needed:
Acquiring moist soils (dirt)
Taking multiple samples
Placing system inside Axel

19. Acknowledgements Kristen Holtz, co-worker Funding: Mentoring:
Keck Institute for Space Studies
Caltech Summer Undergraduate Research
Fellowship (SURF)
Mentoring:
Melissa Tanner, Professor Joel Burdick, Caltech
JPL Axel Team
Kris Zacny, Honeybee Robotics
Prof. Melany Hunt, Prof. Bethany Elhmann
Paul Backes, Paulo Younse, JPL

20. Initial Calculations for Earth conditions
Estimate velocity of air:
Estimate the mass that can be lifted
Assume dirt is inert, ρ=2000kg/m3
Mass = ~12g/s
Mars conditions:
Requires less canister pressure