Picorover Feasibility Study

From TeamFrednetWiki

Slope test results

Can a PicoRover reach GLXP requirements?

This study is based on Mission Requirements and considers if the PicoRover system can reach those requirements. PicoRover 2.0 degrees can't record a video while run but PicoRover 1.3 degrees does.
A self-contained robot has a big difficult to show more than 60% of it self. We propose an easy solution to overcome this fact: to have a periscope. This solution consist of to separate counter-weight from the camera and use camera shaft to have a vertical mast and a camera in the top. The bearing have the cooler and mast is uset as dissipator as well. In addition of that, we can use the mast to put the communications antenna, improbing the link-range. Other issue is to have a payload of 10% of rover weight. Payload are placards, marks, etc. PicoRover have a large unused area where to rib placards and logos. See the following picture:

PicoRover periscope can see 99% of itself showing attached logos

Can a Picorover climb a slope?

You can see three cases in the following picture:

Right case: The ball rolls downhill; purple arrow points downhill.
Center case: The ball holds on a slope; purple arrow points downhill but is canceled by the moment produced by the unbalance. Normal arrow is smaller than before.
Left case: The ball climbs a slope; purple arrow points downhill but is overcome by the high moment produced by the high unbalance. Normal arrow is much smaller than before. Ball friction has a limit when the ball slips.

Study cases	100 grams specimen (0.2 mm wire)

72 spikes specimen (Delta = 0.85)	maximum static angle 37 degrees

292 spikes specimen (Delta = 0.45)	maximum dynamic angle 10 degrees

-18 to 250 ºC shield specimen	-17 to 102 ºC eZ430 board inside

Unbalance ratio (Delta)
We define the unbalance ratio to the coefficient between the center of gravity radius and the center of geometry radius. Radius reference is in the shield and geometry radius is equals to the sphere radius.

Delta = r_gravity / r_geometry



Normalized equation: 1 - (Mass_Shield/Mass_Total)

Delta = 1 All the mass is located at the surface of the ball.
This is the unbalanced case.

Delta = 0 All the mass is located at the center.
This is the usual balanced case.

Vectors
In mathematics, given a vector at a point on a surface, that vector can be decomposed uniquely as a sum of two vectors, one tangent to the surface, called the tangential component of the vector, and another perpendicular to the surface, called the normal component of the vector. reference

In the following we present three study cases where Picorover is rolling over a slope. This study is also valid for Jaluro's longitudinal mode. The diameter is a wheel instead of a ball.
Also, we use these vectors:

W vector (GREEN) is the weight vector. Because this vector is located at a different point from the geometry center, we have an unbalanced counterweight.
F vector (YELLOW) is the force applied in the contact surface. This force is due to the momentum produced by the unbalanced counterweight with respect to the contact point.
N vector (CYAN) is the normal vector applied in the contact surface which is in charge of adherence. When the slope is very high, the wheel loses adherence. This phenomenon is not considered in these study cases.
R vector (PURPLE) is the tangential vector which pulls the Picorover down-hill.
T vector (RED) is the resultant thrust vector. This vector is the resultant vector between the force vector F added to the tangential vector R but applied in the geometry center.

Roll down-hill

This study case shows a wheel rolling down-hill on a 30-degree slope. In this case we use a T thrust (RED) vector corresponding to -120 degrees of relative pitch to the contact point of the N normal (CYAN) vector. This is the maximum negative momentum. The R tangential (PURPLE) vector and the F force (YELLOW) vector are added in the same direction and sense. The Force vector is produced due to the momentum of unbalanced counter-weight with respect to the application point. Values of vectors are approximations because adherence coefficient and shield thickness are not considered.

Picorover video with stability problems without spikes YouTube

Sand-test: One year of PicoRover evolution

Climb a slow slope

For this study case, the wheel is placed on a 10-degree slow slope. In this case we reach to the maximum T thrust (RED) vector corresponding to 90 degrees of relative pitch to the contact point of the N normal (CYAN) vector. The thrust vector is the resulting vector of tangential vector and the force vector. This is the maximum momentum produced by the unbalanced counter-weight with respect to the contact point. Values of vectors are approximations because wheel adherence coefficient and shield thickness are not considered.

Picorover video rolling up the hill 10º slope in YouTube

Hold still on a slope

In this study case, the wheel is placed on a 30-degree slope. In this case we use the F force (Yellow) vector in order to compensate the R tangential (PURPLE) vector and hold still on a slope. The N normal (CYAN) vector is applied in the contact point which is lower than the Climb a slow slope study case because the higher slope angle. This theoretical static case is the maximum slope angle for a Picorover. Values of vectors are approximations because adherence coefficient and shield thickness are not considered.

PicoRover video, with spikes, hold still on a slope of 37º in YouTube

Sand-test: A PicoRover hold still on a 33 degrees sand slope
PicoRover video, with long hair, hold still on a sand slope of 33º in YouTube

It is interesting to see that dynamic performance depends on gravitational field but static performance is independent of gravity. Therefore, these angles theoretically are the same for the Earth or the Moon.

In the following graph, slope angle vs normalized Delta ratio is presented. We have used 3 specimens:

Red: Plastic ball without spikes, 89 mm in diameter and XXX grams of dry mass
Cyan: Plastic ball with spikes, 89 mm in diameter and XXX grams of dry mass
Silver: Aluminum ball with hair, 100 mm in diameter and XXX grams of dry mass

No_Spikes
0gr
10gr
20gr
50gr
100gr
Max

Delta
0.0%
26.8%
41.7%
64.1%
78.1%
100%

Slope
5.0º
24.0º
28.1º
28.9º
29.0º
29.0º

-
.
.
.
.
.
.

With_Spikes
0gr
10gr
20gr
50gr
100gr
Max

Delta
0.0%
23.0%
36.8%
59.2%
74.4%
100%

Slope
6.0º
21.0º
28.0º
29.9º
30.0º
30.0º

-
.
.
.
.
.
.

With_Hair
0gr
10gr
20gr
50gr
100gr
Max

Delta
0.0%
14.0%
24.1%
44.2%
61.2%
100%

Slope
9.8º
16.5º
21.3º
30.6º
33.8º
34.0º

Conclusion

Doesn't matter how much weights the rover, the only factor is the unbalance ratio (Delta)
Maximum continuous slope able to climb by a rover depends on wheel adherence, not depends on rover weight.
Maximum slope able to hold still by a rover on sand depends on umbalance mass (Delta ratio) and less than 29º.
Adherence increases the maximum slope able to hold still by a rover on sand but less than 10º.
Generally, the higer the adherence on sand, the higher the friction.

You can see a real video about these experiments performed in the Polytechnical University of Catalonia (UPC) by the Picorover group.

Picorover video attached to the experiment level in YouTube

In this picture we see the same case hold still on a slope. Those is in the Earth but this is in the moon. Less momentum but also less tangential force.

Can a Picorover survive the lunar night?

In the lunar night, the Picorover minimizes losses of thermal energy. Core temperature is maintained at the minimum range. Main systems are turned off. Additionally, thermal feeding recovers some energy using the thermo-electric effect but when the night is so long there is not option that shut down the PicoRover.
When solar cells provide solar energy, it restarts all systems and day mode is entered. There are batteries able to support very low temperatures.

Power feeding

The other main issue I see with a rolling ball design is where do you put the solar cells? (air.command)
Solar cells are small and are installed inside the camera window. Since Picorover1.3degrees does not opens, deployable solar cells are not allowed. Picorover uses the thermo-electric cells in order to transform thermal energy into electrical power. This feeding subsystem is very effective in both the lunar day and the lunar night. Solar cells are not so critical with this approach.
See this Cryogenic Freezing video about battery test: http://www.youtube.com/watch?v=bqfDZnKecto

Geometry

Anyway, a participan in the GLXP forum (nick: simplex) has presented a thermal study which is extended by us. We use the same parameters but comparing a Sphere vs Box shaped Lunar Rover. In the example we want survive the lunar night using batteries in order to keep the internal pats of the Lunar Rover at 0ºC when we assume -180ºC of outside temperature.
In the following case study we present some advantages when the Think-Outside-The-Box phylosopy is applied for a Lunar Rover.

Sphere vs Box shaped Lunar Lander. There are thermal advantages for spheric shape

In this figure two shape cases are presented. For the same Lunar Rover volume, the Box shaped has a smaller side than the sphere shaped which has a bigger diameter. The wall thickness, which is very thin, is considered 1/20 times the diameter. The yellow area represents the volume occuped by the insulator with thermal conductivity k = 0.004 W/(m·K) like FOAM. For the same volume, the Box shaped has a larger irradiation area than the sphere shaped having less thermal losses. In the previos figure the critical point is also presented when the PAYLOAD volume is full of batteries for survive the lunar night.

When the PicoRover has a diameter of 180 mm (15.8 kg) other squared rover requires higher number of batteries (19.6 kg) and a side of 160 mm. These number may vary depending on materials, thickness and type of batteries used.

Remember that PicoRover group is trying to survive the lunar night not warming the rover but using cryogenic compliance materials and stable electronic components over the superconductivity critical point.

Thermal gradient

Currently, the Picorover group is preparing a functional demonstrator with the real shield made of composite materials. This shield will be able to resist temperatures in the range of 73K to 470K approx. ( -200 ºC to 200 ºC or -330 ºF to 390 ºF). In the center picture you can see an aluminum sheet removed in order to inspect the state of the glass fiber after the thermal exposition inside the oven from -18 to 250 ºC (0 to 482 ºF) as a part of the vacuum Test2. In the right hand picture you can see the state of the polyurethane FOAM.

The thermal shield is an insulator composed by three layers:

Inside a very thin aluminium layer protects the CORE zonal subsystem from electromagnetic interferences and for thermal stabilization. This layer works like an internal mirror keeping the heat inside. A Polypropylene epoxy layer gives stregth to the structure and is used like a glue.
The intermediate is a thin layer of polyurethane having good insulating properties, is light and moderate stregth. A Polypropylene epoxy layer gives stregth to the structure and is used like a glue.
The external layer is made of glass-fiver and Polypropylene matrix and it is glued with shield components. Inside there are copper and aluminum sheets for shielding these components.

Vacuum Test

The Firs Test is just put all the shield inside the vacuum chamber in order to test if composite bubbles can explode and make weak the shield. See the videos Test1 for the concept and Test1a for a high vacuum test in YouTube.
In the Second Test I will apply inside the shield as much vacuum as possible until we break the shield. This is an empty shield for a destructive test (Implosion test). I want to know the minimum negative delta pressure allowabe. What will be broken first, the window or the shield? See the video Test2 in YouTube.
The Third Test is to put a MCU inside the Shield an close the window (I am preparing a glass window to cover the hole). First the component alone and then with the MCU inside. The MCU will meassure temperature. I will put inside the fridge and record the difference. Following I will put inside the oven and record the temperature. That way I can see an approximation of the real thermal gradient of the insulator. See video Test3a for the component alone and inside the shield Test3b in YouTube.

Thermal patern between outside temperature and PicoRover MCU temperature

These numbers are required in order to be qualify and match launcher specifications.

Radiation

Radiactive protection has to be implemented by shielding and by software protection. The counterweight is made of lead for thrust the PicoRover. Inside is located the critical components: MCU, Memory, etc. The rest of component are exposed to the radiation because the effects could be overcome.
As per high energy gamma and neutrons particles it is true that a shield is not enought and a robust algorithm has to be installed in the MCU in order to detect wrong actions. Also Forward Error Correction FEC has to be implemented in the communications layer.

Conclusions

In orbit, the challenge is to ensure good thermal conductivity between different parts of the spacecraft so that heat is transfered from the warm side to the shaded sides. There will always be shaded areas and these will see very cold sky (few Kelvin) when facing empty space. The thermal equilibrium is proportional to:

T(eq) ~ (α/ε)^-.25

where α is the absorptivity, the ability of a surface to absorb solar radiation, 0 ≤ α ≤ 1 ε is the emissivity, the ability of a surface to radiate heat, 0 ≤ ε ≤ 1

Thus, if heat is of concern, you will use surface materials with low absorptivity and high emissivity. For example, white paint has α = 0.15 and ε = 0.9 and a white structure in space would reach equilibrium at -61° C in sunlight or -70° C in eclipse. So there is plenty of room to dissipate "electronic heat". Black paint has α=ε=0.9 and the equilibrium temperatures would be +20° C in sunlight and -2° C in maximum eclipse.

The surface of the Moon is a bit more tricky because you also have to take the surface radiation into account, but the same principles apply. See for example why the Astrobotic rover has the shape and color as it has: http://www.youtube.com/watch?v=cMXHYrrNuUI

These were just some simple principle considerations and the engineering part is more tricky, but hey, that's why we have thermal engineers!

--Alex

Can a Picorover carry the Elphel high definition camera?

Picorover diameter is minimized around 0.1 meters in order to put a HD internal camera.


Elphel camera inside the PicoRover1.3degrees

The only modification is the sensor which is separated from the control board. Due to this modification, components are exposed to the radiation. Hence we don't use the camera's computer for critical operations. For this reason, we study this phenomena as a business risk.
A mirror in front of the optical turn and switch between lateral view and front view. In addition it have two pieces working as insulating shield.

Can a Picorover steer in the dust?

Since Picorover1.3degree is designed, more stability is provided for the Picorover. Thrust is achieved by changing the CORE pitch angle respect to the SHIELD. Steering is achieve by controlling the CORE roll angle.

Two stepper motors controls the pitch and the roll angles

Component for the near to space test

Also you can see the dynamic behavior of Picorover in the Dynamic test where forward, turn and climb movements are presented.
The rolling resistance also determines the maximum thrust able to provide over the regolith surface.

Can a PicoRover overcome an obstacle higher than itself?

Picorover uses spikes in order to increase the adherence, which is a function of the Normal Vector as shown before.


Demo1.3degrees in action. 270 grams and 292 spikes	Jump procedure overcoming a higher obstacle than the PicoRover

In order to increase traction in the regolith we have designed a steel hair which increases the effective Picorover diameter when running fast. Theoretically, the PicoRover can overcome an obstacle higher than its normal diameter size. In this respect, testing is required.

Following a list of PicoRover function modes:

Stop: Static stability garateed by the umbalance mass.
Start: Movement starts. Umbalance mass travel up to the maximum angle for maximum acceleration until cruise speed is reached, the pitch angle change only with the slope angle.
Cruise: When cruise speed is reached, the pitch angle change only with the slope angle.
Climb a slope: Pitch angle changes very slow until start the movement and keep small to climb slow and maximize adherence.
Take 360º picture: Make two turns. First is to make a footprint, second to record in lateral view following the same footprint.

Can a PicoRover downlink?

PicoRover group uses a light 2.4 GHz transmitter @ 450 kbps in order to transmitt the near real-time mooncast but has to be in line of sight and near to 50 meters. Beyond these limits, PicoRover is trying to use the same PicoSAR antenna transforming a radar in a high gain antenna for downlink.
The PicoSAR antenna is very directive and can change the lateral direction by the synthetic aperture means. This antenna is a patch array antenna at 10 GHz (X-Band). The longitudinal direction is achieved by the rotation of the shield because the PicoSAR antenna is inside the shield. We can put up to three of these patch antennas in the shield; that way you always have a picoSAR antenna pointing to the sky.

Link to the Lunar Lander

To be done

Link to the Lunar Bus