Breaking New Ground

Global Nominee

Breaking New Ground received a Global Nomination.

Solar System

Develop an approach for characterizing the composition of asteroid for mining potential and a process for mining different compositions. Explore a possible division of labor involving different types of vehicles (e.g. sensor units, drilling units, power gathering and distribution, extracted resources handling and transferring). Consider solutions for moving said asteroids between different orbits and/or consequently make periodical adjustments to keep them in place. Analyze how your idea would cope in some of the given scenarios or outline a scheme of your own.

Explanation

Introduction

Inspired by the near-miss landing of the Philae probe in 2014 [Forbes, 2014], our team has developed a method for not only determining asteroid composition but also for establishing surface density, increasing the safety and minimizing risk for landers that may be ten years and hundreds of millions of dollars in the making.

By training machine learning algorithms to identify surface density, coupled with laser-based spectroscopy, an orbiter can establish a landing zone for the mining robot we have designed. Once landed, the mining robot can gather resources and transport them back to either Lagrange points, Lunar orbit or Earth orbit.

A brief overview of the mission profile we constructed is as follows:

Two heavy-lift launch vehicles (SpaceX Falcon Heavy) are prepped. One contains a solar power array and an orbiter, the other a single large mining robot or multiple smaller ones. The first launch vehicle delivers the solar power array to an orbit closer to the sun than the main asteroid belt, allowing it to generate and provide a steady stream of energy through focused power transfer (microwave or laser, for example) to the orbiter and, later, the mining robot/s.

The orbiter is placed into a close orbit surrounding the target asteroid and begins high-resolution LIDAR mapping of the surface. After completing the mapping, it will decide which sites look the most viable for landing based on known surface composition and certain geographical requirements. It will then begin a controlled bombardment of the surface to confirm the viability of the landing site (section two). Assuming the landing site is viable, the third task of the orbiter is to analyze the surface and subsurface composition using Laser-Induced Breakdown Spectroscopy (section three).

A landing site now successfully chosen and shown to be within the safety parameters of the mission, the second launch vehicle delivers the mining robot/s to the site and they begin excavation of material (section four).

Material return for further processing can be handled in a few different ways (section five), most likely through the use of mass driver systems.

The asteroid we have elected to target for this mission is Anteros. With an estimated value in excess of $1T and an orbit that brings it close enough to Earth to almost make the ‘potentially hazardous objects’ list, it is both accessible and valuable. [Asterank, 2016]

(N.B. the model used during animations is that of Vesta. This was chosen due to the beautifully detailed high-resolution models provided by NASA, to help convey our mission to the local judges)

Section 2 - Landing Site Identification through Controlled Bombardment

To effectively determine the surface density of our prospective landing zones, we drew inspiration from depth-of-penetration ballistics testing and DS-2 Mars microprobe penetration tests [Lorenz et al, 1999], and thought ‘Let’s throw a small rock at the big rock then measure the dent’. To confirm this process would be viable we created an experimental apparatus in the middle of our Hackathon location. Using a golf ball as a projectile, we launched it into a tub containing sand (in the first instance), flour (in the second) and then a mix of the two (in experiments 3 and 4). The resulting impact crater was then scanned using LIDAR technology and converted into 3D models, send in the feature image of the project page.

A Machine Learning algorithm was then passed 80% of the data we gathered, keeping 20% in reserve to test the machine. After training, the algorithm was able to answer with 75% accuracy the surface composition it was analyzing.

In this experiment the sand and flour represent a harder surface and a softer, dustier surface respectively. Ten samples of each were taken, as well as ten each of two parts sand to one part flour, one part sand to two parts flour, with the layer of flour over the sand acting as a coating of ‘dust’ over a rocky surface. The results we gathered seem to demonstrate the effectiveness of the model, with highly believable impact craters, including crater rims and material ejection patterns, matching known patterns. [Melosh, 1989, Osinski et al, 2012]

For this process to be useful in asteroid surface density characterization, the projectiles launched would have a known mass and launch velocity, creating an impact crater that will be distinct for different material densities.

If the Machine Learning algorithm were trained on a larger volume of data, it is certain it would be far more accurate in determining surface density and assisting in pinpointing a landing location. To ensure accurate and useful data is being fed to the algorithm, a high volume of experiments must be undertaken, with potential experimentation in a free-fall environment such as reduced gravity aircraft (‘Weightless Wonders’). This would help establish the pattern ejected material would take during actual asteroid bombardment.

It is also possible the Machine Learning algorithm could be trained on known impact craters. By feeding it the known data (such as depth of crater, theorized impact projectile mass and velocity, density of impacted material), it could gather a still-richer set of variables to analyze during active use and help establish landing sites with even higher accuracy.

Section 3 – Asteroid Composition

Laser-Induced Breakdown Spectroscopy (LIBS) was identified as a potential method for surface and subsurface composition analysis. LIBS utilizes a low-energy laser focused to a tight spot on the material under investigation. The material will then vaporize, forming a plasma that emits with the characteristic lines of the constituent materials as it cools.[Noll, 2012, Cremers et al, 2006]

To apply this technique to asteroid composition analysis, the orbiting spacecraft would be placed in a low orbit and examine the area surrounding the site established as viable by the controlled bombardment surface density testing. A pulsed laser would be used to vaporize the material and a spectrometer used to determine the composition of the resulting plasma. Repeating this process many times would excavate material and allow for sub-surface composition to be studied (though the rate of excavation depends entirely on the intensity of the laser).

As originally envisioned, this process would be performed in an area surrounding the identified landing site, to provide some confines for the mining robot/s to operate within. After excavation from the mining robot/s the process could be repeated again to continue analyzing the composition and ensure useful/target materials are being extracted.

Section 4 – Mining Robots

For the SpaceApps challenge we elected to design our own mining robot, borrowing concepts and ideas from a range of areas.

Originally designed to fit within the fairings of SpaceX’s Falcon Heavy [SpaceX, 2016], we set tight constraints on the mass and diameter of the mining robot. The Falcon Heavy is expected to be able to launch 12 tonne to Mars and has a fairing diameter of 12 meters. To allow for some overhead we designed around 10 tonne and 10 meters. Following input from other individuals in the SpaceApps challenge we ensured the design would be scalable and allow for a robot to be built in a range of sizes.

Asteroids pose a particularly unique challenge for mining. Low surface gravity means there is a very real chance of dislodging when simply drilling to attach anchoring rods. To get around this, the Jet Propulsion Laboratory have developed a system known as ‘Robotic Micro-Spines’[JPL, 2016], which act as gripping ‘claws’ that hold fast onto an irregular surface.

Our design incorporates the JPL Micro-Spines with a quadruped form, designed for foldability during transportation and mobility once unfurled on the asteroid surface. The robot contains a mining column, a mass driver/solenoid and apparatus to collect energy beamed from the solar power arrays in orbit about the sun. An RTG and batteries would also be carried on board, the latter to store energy for mining and the former to provide heat for the electronics and other sensitive components.

To mine, the central mining column descends and, equipped with three interlocking grinding teeth inspired by tunnel-boring machines, begins grinding the surface. The grinding teeth operate in such a way as to funnel the ground material into the body, for later handling (next section). In this process the maximum downwards force that can be applied is determined by the gripping strength of the micro-spines, although it is not necessary to apply a large amount of pressure and more important to apply constant downwards force. Once the surface has been levelled to a desired depth, the mining robot can then retract, release one of its four micro-spine ‘feet’ and reposition itself to continue grinding and mining.

The concept here is akin to ‘peeling a potato’. The mining robot would grind to a set depth then shift its position slightly to continue the operation. After some time it would grind flat the area prescribed during the composition analysis and require further direction.

Section 5 – Material Return

A range of return options exist though time constraints limited our ability to explore these in detail. The ‘easiest’ is that of a mass driver contained within the mining robot itself. A mass driver may take many forms but the standard configuration is that of a coilgun, effectively a solenoid with individually addressable coils. When mining ferrous material, the mined ore could be accelerated by the mass driver into space, the trajectory being controlled through a combination of the current applied to the coils and the position of the legs of the robot. By raising or lowering legs the material could be projected into any area of the sky above the robot, given by some solid angle as a function of the mobility of the legs.

Unfortunately whilst being the easiest this method is also something of the least practical or useful, as ferrous materials are unlikely to be the initial targets of asteroid mining operations. In the future, when orbital construction techniques are developed, this will be required but in the short-term rare earth metals and platinum group metals are considerably more valuable and hence considerably more likely.

The coilgun approach noted above could still be utilized, this time maintaining a fixed ‘projection’ bucket that would launch the material but not be lost during the process. The American physicist Gerard O’Neill, best known for inventing the particle storage ring for high-energy physics experiments, invented and explored the concept of mass drivers (as described here) in great detail.

An alternative to mass drivers is that of in-situ development of LOX/H2 fuel, feasible on any asteroid containing water ice. A highly effective rocket fuel, the robotic miner could contain within itself a mechanism to create and store these components, and a small engine to thrust it off of the asteroid once ‘full’ of mined material. This idea has not been explored in great detail by us but we feel it holds promise, as well as the mass drivers, for recovery of the mined material.

Addendum

A concern was raised regarding the force imparted on the asteroid by the mass-driver concept. As was correctly pointed out, the asteroid will experience a force with each firing of the mass-driver, potentially causing a despinning or destabilization of the asteroid orbit. To that end, we elected to check whether this is something that required consideration.

Anteros is listed as having a diameter of 2.3km. Assuming perfectly spherical (a crude approximation, but sufficient for our purposes), it's volume is then 6.3E9 cubic meters. Asterank.com lists the composition of Anteros as Magnesium Silicate, Aluminium and Iron Silicate. No percentage compositions are available as this asteroid has not been studied in great detail, so an assumption was made that the percentage mass ratio is 30/30/40 for Magnesium Silicate, Aluminium and Iron Silicate respectively. The bulk densities of these three materials are 3110, 2700 and 2400 kg per cubic meter respectively.

Using this percentage composition, the material density and the volume we find a total mass of 17.22E12 kg. As the gravity of Anteros is not known, the mass given by our composition assumption will be used to calculate, offering us a gravity of 8.7E-4 m/s2 or 0.0089% Earth gravity.

With such a small gravitational force, excavated material would only require 0.87N per tonne to achieve an upward velocity of equal magnitude. Then, slightly more than 0.87N per tonne would be required to escape from the asteroid surface.

A more important calculation is the force required to reach a delta-V to match Earth so the material can be deposited into Earth or Lunar orbit. Asterank.com lists Anteros' delta-V as 5440m/s. Acceleration runway becomes important in this situation, with the acceleration runway being dependent on the size of the mining robot.

In any realistic robotic miner design, the available acceleration runway will be small. Our projectile will only experience an acceleration for a time equal to the square root of the acceleration runway (x) over the acceleration (a). Then, t = sqrt(x/a). If the acceleration runway (x) is set at 1m (for ease of calculation), the acceleration time is then proportional to the square root of the acceleration and we encounter a problem. The higher the acceleration, the shorter duration our projectile experiences it for. Time constraints prevented a thorough evaluation of this, and as a result an arbitrary force on our projectile mass was selected at 10kN per kilogram.

If projecting one tonne of material, a force of 10MN would be required, a tremendous amount of force. The original concern was that of affected the asteroid orbital stability, Even with this level of force, the asteroid would only experience an acceleration of 0.000581m/s2, effectively negligible for many dozens of mass projections.

A force of 10MN grossly exceeds the maximum gripping force the JPL-inspired micro-spines can apply to the asteroid. With careful manipulation of the orbital mechanics involved in returning material to a useful location (Earth/Lunar orbit, Lagrange points), the required delta V can be dropped significantly and become far more achievable. This is an area that will be explored in more detail, as it holds promise for a way to return the material with substantially lower accelerations and energy requirements.

***

Please note: the above was quickly explored 'back of the envelope'-style when concerns were raised. It is not robust and will require further investigation and a deeper grasp of orbital mechanics, plus experimentation with coilguns/mass drivers, to fully develop.

Resources Used

Great overview of asteroid mining hypothetical sequence of events:

http://www.nasa.gov/content/goddard/new-nasa-missi...

Inspiration for asteroid cataloguing and astro-prospecting :

http://www.asteroidmission.org/

http://www.jpl.nasa.gov/cubesat/missions/neascout....

https://www.nasa.gov/resource-prospector

http://www.asterank.com/#composition

Potential use of nets to secure and transport mining bulk from asteroid surface to LaGrange point

http://www.nasa.gov/content/wrangler-capture-and-d...

Textbooks and Academic Papers consulted

http://www.lpl.arizona.edu/~rlorenz/ds2.pdf